measuring the impact of lean implementation on ...€¦ · thus, lean philosophy is perceived as...

POLITECNICO DI MILANO

DEPARTMENT OF MANAGEMENT, ECONOMICS AND INDUSTRIALENGINEERING

DOCTORAL PROGRAM IN MANAGEMENT, ECONOMICS AND INDUSTRIAL ENGINEERING

MEASURING THE IMPACT OF LEAN IMPLEMENTATION

ON OCCUPATIONAL HEALTH AND SAFETY THROUGH

LEADING INDICATORS

Doctoral Dissertation of

SEYED SAJAD MOUSAVI

Supervisor and Tutor:

PROF. PAOLO TRUCCO

A.Y 2017-18 – XXX cycle

I

CONTENTS

ACKNOWLEDGEMENTS ................................................................................................................................. 1

ABSTRACT ........................................................................................................................................................... 2

SOMMARIO .......................................................................................................................................................... 3

CHAPTER 1: INTRODUCTION........................................................................................................................ 4

1.1 Research background ........................................................................................................................................ 4

1.2 Problem statement and its relevance ................................................................................................................. 5

1.3 Research purpose and research questions ......................................................................................................... 6

1.4 Research contributions to knowledge and practice ........................................................................................... 6

1.5 Thesis outline .................................................................................................................................................... 9

CHAPTER 2: STATE OF THE ART REVIEW ............................................................................................. 10

2.1.1 The status of employing lean philosophy in industries ................................................................................ 13

2.1.2 Lean maturity ............................................................................................................................................... 14

2.2 Occupational health and safety concepts ........................................................................................................ 16

2.2.1 Measurement of safety performance ............................................................................................................ 17

2.2.2 Safety performance indicators ..................................................................................................................... 18

2.2.3 Antecedents of safety performance .............................................................................................................. 20

2.2.4 Classifying antecedents of safety performance ............................................................................................ 20

2.2.4.1 Workplace environment ............................................................................................................................ 20

2.2.4.2 Workforce characteristics ......................................................................................................................... 21

2.2.4.3 Task characteristics ................................................................................................................................... 22

2.2.4.4 Organizational factors ............................................................................................................................... 22

2.3 Relationship between lean and safety ............................................................................................................. 23

2.4 The lack of a generalized model of the relationship between lean and OHS .................................................. 26

CHAPTER 3: RESEARCH DESIGN ............................................................................................................... 28

3.1 Research model and hypotheses ..................................................................................................................... 28

3.2.1 Introduction to PLS-SEM ............................................................................................................................ 33

3.2. 2 Reasons for using PLS-SEM in the existing study ..................................................................................... 35

3.2.3 Survey design and administration ................................................................................................................ 36

3.2.3.1 Sample design ........................................................................................................................................... 36

3.2.3.2 Developing the questionnaire ................................................................................................................... 37

3.2.3.3 Determination of sample size ................................................................................................................... 37

3.2.3.4 Pilot study ................................................................................................................................................. 38

3.2.3.5 Questionnaire sharing ............................................................................................................................... 38

3.2.3.6 Large-scale study ...................................................................................................................................... 39

3.2.3.7 Data sorting............................................................................................................................................... 40

3.2.3.8 Data encoding ........................................................................................................................................... 40

3.2.3.9 Handling missing data .............................................................................................................................. 43

3.2.3.10 Quality checks of results ......................................................................................................................... 43

3.2.3.11 Data entry ............................................................................................................................................... 43

3.3 Analysis procedures ........................................................................................................................................ 44

3.3.1 Building the inner model ............................................................................................................................. 44

3.3.2 Building the outer model ............................................................................................................................. 44

3.3.3 Formative and reflective measurement ........................................................................................................ 44

3.3.4 Running the path-modeling estimation ........................................................................................................ 45

3.3.4.1 Assessment of the reflective measurement models ................................................................................... 46

3.3.4.1.2 Internal consistency reliability ............................................................................................................... 46

3.3.4.1.3 Convergent validity ............................................................................................................................... 46

3.3.4.1.4 Discriminant validity ............................................................................................................................. 46

3.3.4.2 Assessment of the formative measurement models .................................................................................. 47

II

3.3.4.3 Evaluation of structural model .................................................................................................................. 48

3.3.4.4 Importance-performance matrix analysis ................................................................................................. 49

3.3.4.5 Mediation analysis .................................................................................................................................... 49

3.3.4.5.1 Magnitude of mediation ......................................................................................................................... 50

3.3.4.6 Moderation analysis .................................................................................................................................. 50

CHAPTER 4: FINDINGS .................................................................................................................................. 51

4.1 Reflective measurement analysis .................................................................................................................... 51

4. 2 Analysis of formative measurements ............................................................................................................. 54

4.3 Structural model evaluation ............................................................................................................................ 57

4.5 Mediation effect analysis ................................................................................................................................ 59

4.7 Moderation analysis ........................................................................................................................................ 59

CHAPTER 5: DISCUSSION ............................................................................................................................. 61

CHAPTER 6: CONCLUSIONS AND FUTURE RESEARCH ...................................................................... 68

6.1 Theoretical implications ................................................................................................................................. 68

6.2 Practical/managerial implications ................................................................................................................... 70

6.3 Study limitations and future research ............................................................................................................. 71

PUBLICATIONS OF THESIS RESULTS…................................................................................................... .74 REFERENCES ................................................................................................................................................... ..75

APPENDIES…………………………………………………………..………………………………………....82 Appendix A: OHS leading indicators proposed to lean practices……..…………………………………...........82 Appendix B: The questionnaire……………………………………….…………………………………………87

III

LIST OF TABLES

Table 1: Description of lean tools and techniques (source: http://www.strategosinc.com/lean_tools.htm) ......... 12

Table 2: Antecedents of safety performance extracted from the literature ........................................................... 21

Table 3: Descriptive analysis ................................................................................................................................ 40

Table 4: Latent constructs and corresponding reflective and formative indicators .............................................. 41

Table 5: The outer loadings of the reflective indicators ....................................................................................... 51

Table 6: Results of Fornell-Larcker criterion ....................................................................................................... 52

Table 7: Results of Cross loadings ....................................................................................................................... 53

Table 8: Results summary for reflective measurement models ............................................................................ 54

Table 9: Results of VIF for formative indicators .................................................................................................. 55

Table 10: The outer weights of formative indicators ............................................................................................ 56

Table 11: Collinearity assessment of latent constructs ......................................................................................... 57

Table 12: R2 evaluation of the endogenous variables ........................................................................................... 58

Table 13: Results of total effetcs among constructs ............................................................................................. 58

Table 14: Index values and total effects for the IPMA of OHS performance ....................................................... 59

Table 15: The results of indirect effects ............................................................................................................... 59

Table 16: The mediating effects of antecedents ................................................................................................... 59

Table 17: The moderating effect of sector variable over direct relationships in the model .................................. 60

Table 18: The moderating effect of size variable over direct relationships in the model ..................................... 60

Table 19: Summary of hypotheses testing ............................................................................................................ 64

IV

LIST OF FIGURES

Figure 1: Evolution on lean production ................................................................................................................ 10

Figure 2: Lean Pyramid ....................................................................................................................................... 11

Figure 4: The relationship between lean maturity and performance .................................................................... 15

Figure 5: Lean maturity in enterprise .................................................................................................................. 15

Figure 6: A holistic view of the workplace and the importance of OHS .............................................................. 16

Figure 7: The extended system model showing the feedback from the indicators ............................................... 19

Figure 8: Relationship between leading, lagging indicators, and performance .................................................... 19

Figure 9: Classification of safety performance's antecedents and their relationship to safety performance ......... 23

Figure 10: General process model for safety and lean ......................................................................................... 25

Figure 11: Lean effects on working environment and employee health and well-being ..................................... 25

Figure 12: Relationship between lean practices and safety outcome .................................................................... 25

Figure 13: Research framework ............................................................................................................................ 31

Figure 14: Theory building and fact finding ........................................................................................................ 32

Figure 15: Impact of lean practices on customer satisfaction by using SEM approach ........................................ 35

Figure 16: Proposed sample size .......................................................................................................................... 38

Figure 17: The file extracted from the Qualtric .................................................................................................... 39

Figure 18: The guideline for choosing the measurement model mode ................................................................. 45

Figure 20: The formative measurement models assessment ................................................................................. 47

Figure 21: The structural model assessment procedure ........................................................................................ 48

Figure 22: Mediation analysis .............................................................................................................................. 49

Figure 23: The mediator analysis ......................................................................................................................... 50

Figure 24: The bootstrap sign change options ...................................................................................................... 57

Figure 25: The proposed model for safety concepts............................................................................ ...................65

1

ACKNOWLEDGEMENTS

First and foremost, I would like to thank Almighty Allah (God), the compassionate, the mer-

ciful, for blessing me to accomplish my PhD study.

To my parents, who selflessly stood behind me from childhood up to now. Special thanks for

all that you did for me. This work is lovingly dedicated to you.

I would like to thank all my brothers and sisters for their support and encouragement, espe-

cially my brother, Dr. Farid. Special thanks also to my relatives and my friends for their con-

stant encouragement.

I would like thank Prof. Reza Khani Jazani from Shahid Beheshti University of Medical

Sciences in Iran. Thanks for your support on academic subjects and your advice that greatly

helped me not only in my PhD study, but also in my life.

As I spent some parts of my study at Missouri University of Science and Technology in the

USA, I would firstly like to thank Prof. Elizabeth Cudney for her thoughtful guidance on my

work. I must say it was very helpful during my survey process. Her great patience with me

will always be appreciated. Secondly, thanks to the university staff that provided me with

necessary information and resources that helped me during the research period.

Special thanks to Prof. Antonio Calabrese, discussant of my thesis. The helpful comments,

critique, and suggestions you made during the yearly evaluations, resulted in greatly im-

proved work.

I would like to express my gratitude and admiration for my thesis supervisor, Prof. Paolo

Trucco. Thank you so much for your patience and for trusting me. Your guidance, valuable

advices, and encouragement made me carry out this thesis. You were always available to an-

swer my questions and helped me develop the idea behind this work. Also, thanks for your

continuous support throughout the duration of research.

Lastly, the deepest thanks go to, of course, my wife. Thank you not only for being my life

partner, but also for being my best friend. Your continuous support as usual, especially during

the PhD period made me what I am standing here now. Without your encouragement, I could

not have finished this journey of my study. Part of this thesis is yours. Thank you.

Seyed Sajad Mousavi

2

ABSTRACT Today’s companies are under tremendous pressure to perform at the lowest cost, highest qual-

ity, and fastest pace; therefore, lean has emerged as a popular management philosophy for

companies to attain a competitive edge. The "lean thinking" concept has become more popu-

lar since the beginning of the 21th century in accordance with the advent of the economic cri-

sis. Cost reduction and customer satisfaction are among the primary goals that companies try

to reach when developing their strategies. Thus, lean philosophy is perceived as useful “tool”

by a wide range of organisations, in the manufacturing and service sectors. Along with lean

implementation, however, there is an increasing concern over occupational health and safety

(OHS) at the workplace. There is concern that due to lean implementation, the focus on

productivity may result in health and safety issues being ignored or worsened. The relation-

ship between lean and OHS has not been clearly understood up to now. In literature, many

authors claim that further research is needed to better understand the impact of lean imple-

mentation on occupational health and safety. Moreover, a more suitable approach to measure

these impacts still needs to be proposed and implemented. Previous studies on the relation-

ship between lean implementation and OHS were mostly case studies that focus on some

parts of this relationship. Therefore, a comprehensive study is still lacking in literature.

The aim of this dissertation is to determine how lean implementation influences OHS perfor-

mance, and to suggest the adoption of OHS leading indicators to identify and assess the

strength of different mechanisms that shape the relationship. To do so, a comprehensive liter-

ature review was conducted to identify all the elements (antecedents) that influence on OHS

performance. For each antecedent one or more possible measurement indicators (leading in-

dicators) were then proposed. Finally, again from a literature review, it was possible to extract

three components related to the implementation methods of lean production: fidelity, exten-

siveness and experience. Lastly, in order to validate the overall model, a set of hypotheses on

the relationships between lean components and OHS performance antecedents was tested via

Partial least square-based structural equation modeling (PLS-SEM), based on survey data.

The survey was conducted to gather information from industries across the world. The analy-

sis clearly proved the importance of using OHS leading indicators to forecast and measure the

impact of lean implementation on OHS performance. This thesis contributes to the academic

community and to practitioners by offering a quantitative framework for deploying the rela-

tionship between lean implementation and OHS performance. Moreover, the proposed

framework and OHS leading indicators can be adopted by organizations to design and assess

the expected benefits of implementing lean and OHS improvement programmes simultane-

ously.

Keywords: Lean implementation, health and safety, OHS performance, antecedents, leading

indicators, structural equation modeling (SEM), survey

3

SOMMARIO Le organizzazioni sono oggi sottoposte ad una tremenda pressione per offrire prodotti e servi-

zi al più basso costo, con la massima qualità e velocità di consegna. Pertanto, la produzione

snella (lean production) è emersa come una filosofia di gestione diffusa e vincente per ottene-

re un vantaggio competitivo. Il concetto di "lean thinking" è diventato ancora più popolare

dall'inizio del XXI secolo a seguito della crisi economica. La riduzione dei costi e la soddi-

sfazione del cliente sono tra gli obiettivi primari che le aziende perseguono nel definire le loro

strategie. Pertanto, la filosofia lean è vista come uno strumento utile da una vasta gamma di

organizzazioni in diversi settori, dal manifatturiero ai servizi. A fianco di obiettivi di efficien-

za, vi è tuttavia una crescente preoccupazione per la salute e la sicurezza (OHS) nei luoghi di

lavoro. Si teme che durante l'implementazione snella, l'attenzione alla produttività possa por-

tare a ignorare o aggravare i problemi di salute e sicurezza. La relazione tra lean e OHS non è

stata chiaramente compresa fino ad ora. In letteratura, molti autori sostengono che sono ne-

cessarie ulteriori ricerche per capire meglio l'impatto che l'implementazione lean ha sulle pre-

stazioni aziendali di OHS, e un approccio di misura efficace deve ancora essere proposto ed

attuato. Inoltre, gli studi finora condotti sono stati per lo più casi di studio incentrati su alcuni

aspetti di questa relazione. Pertanto, in letteratura manca uno studio completo sul tema.

Lo scopo principale della tesi è duplice: determinare in che modo l’adozione di una filosofia

lean influenza le prestazioni OHS e proporre l'utilizzo di leading indicators per misurare

l’intensità di questa relazione. A tale scopo, è stata condotta una revisione completa della let-

teratura per identificare tutti gli elementi per i quali è dimostrata una influenza sulle presta-

zioni OHS, sulla cui base è stato sviluppato un modello originale che rappresenta gli antece-

denti delle prestazioni di OHS. Per ciascun antecedente è stato poi proposto uno o più possi-

bili indicatori di misura (leading indicators). Infine, sempre da un’analisi di letteratura, è sta-

to possibile estrarre tre componenti relativi alle modalità di implementazione della lean pro-

duction: fedeltà, estensività ed esperienza. Al fine di validare il modello generale, sono state

testate una serie di ipotesi circa le relazioni tra i componenti lean e gli antecedenti delle pre-

stazioni OHS. L’analisi è stata condotta utilizzando la tecnica PLS-SEM applicata a dati rac-

colti tramite sondaggio. L’indagine on-line ha consentito di raccogliere informazioni su di-

verse industrie in tutto il mondo. I risultati dell’analisi confermano l’importanza degli antece-

denti e dei leading indicators come elemento chiave per comprendere e misurare la relazione

di influenza sussistente tra lean production e prestazioni OHS. Questo lavoro contribuisce sia

alla conoscenza scientifica sia alla pratica industriale mostrando la relazione effettiva tra im-

plementazione della filosofia lean e le prestazioni OHS. Inoltre, il modello e i leading indica-

tors proposti possono essere utilizzati dalle organizzazioni per progettare e valutare i benefici

attesi dall'implementazione simultanea di programmi di miglioramento lean e sicurezza.

Keywords (Italiano): Produzione snella, Salute e sicurezza sul lavoro, leading indicators,

PLS-SEM.

4

CHAPTER1

INTRODUCTION This chapter is an introduction to the present study, which primarily focuses on research

background. Then, the problem statement and its relevance are presented. The research pur-

pose and research questions will be projected in the third section. In the fourth section of this

chapter the contributions of this study to knowledge and practice are discussed. Lastly, the

structure of the thesis is explained.

1.1 Research background

Since the creation of the Lean Production (LP) concept in Japan by Toyota after World War

II (Holweg, 2007), considerable analyses have been conducted about it. The term ‘lean’ is

used by Ghosh (2012) to refer to produce the same output with fewer resources (manpower,

material, space, and machinery). A further definition is given by Ohno who describes lean as

eliminating waste from the production system (Ohno, 1988). Furthermore, to better under-

stand the lean tools and techniques, Shah and Ward (2003) classified them into four sets of

consistent practices: total quality management (TQM), just-in-time (JIT), human resource

management (HRM), and total productive management (TPM).

Nowadays, according to the lean definition (Holweg, 2007; Mi, Park, & Pettersen, 2009)-

manufacturing philosophy for shortening the total time cycle by eliminating wastes from

work processes-enterprises just focus on lean and its results such as quality increase, decrease

time cycle and lower costs. The term "lean" therefore is a philosophy or attitude which tend to

reduce the waste in an organization (Cudney, Furterer, & Dietrich, 2013). The waste in an

organization is the non-value added tasks for which customers would not pay. Therefore, lean

philosophy attempts to identify non-value added tasks through several tools and techniques

and eventually reduce them. In order to implement the lean philosophy and attitude, a set of

tools and techniques have been introduced, such as value stream mapping, 5S, mistake proof-

ing, and kanban. According to the nature and utility of lean, a wide variety of industries are

able to implement these tools and techniques; however the manufacturing sector is the most

common industry for the lean implementation. Commonly reported positive results of using

lean thinking include improved productivity, cost reduction, shortened work cycle time, and

improved quality (Sánchez & Pérez Pérez, 2001; Rahman, Laosirihongthong, & Sohal, 2010).

The "lean thinking" concept has become more popular since the beginning of the 21th century

because of the economic crisis. Thus, most industries employ lean tools and techniques.

Along with the significant results of employing lean techniques for industries, another side of

this issue should be noted. Because of change from traditional mass production to lean pro-

duction, the redesign of production processes, employees' activities, and site lay-out is re-

quired. Moreover, the changing culture is inevitable. In accordance with these broad changes

occurring within the workplace, critical arguments have arisen among researchers and practi-

tioners (Bruno & Jordan, 2002). Occupational health and safety (OHS) issues are one of those

arguments. There is concern about overlooking occupational, health and safety issues while

lean is being implemented at the workplace. Some authors have conducted studies with re-

5

spect to lean implementation impacts on OHS (Longoni, Pagell, Johnston, & Veltri, 2013;

Saurin & Ferreira, 2009); however, there is no agreement on the impact of lean implementa-

tion on OHS performance. For instance, while positive effects such as job autonomy, worker

participation, empowerment, and job enlargement have been reported (Womack, Jones, &

Roos, 1990), negative effects such as occupational stress increase, rise in occupational acci-

dents, and the growth of muscle-skeletal disorders have also been reported (Conti & Angelis,

2006; Hallowell, Veltri, & Johnson, 2009; Landsbergis, Cahill, & Schnall, 1999). Almost all

studies regarding the relationship between lean and safety have not studied this issue in a

comprehensive manner, which means various aspects of both lean and safety have not been

considered as an entire framework so far. Furthermore, mostly the lagging indicators have

been employed within the relationship between lean and safety, and therefore, the importance

of leading indicators has not been represented. In conclusion, a comprehensive study regard-

ing the relationship between lean and safety is needed to overcome the challenges in this re-

gard.

1.2 Problem statement and its relevance

The development from traditional mass production to lean production requires a redesign of

production processes, worker activities, and the site layout, all of which can affect site safety

and health. Thus, there is concern about overlooking occupational, health and safety (OHS)

issues while lean is being implemented at the workplace. Some authors (Anvari, Zulkifli, &

Yusuff, 2011; Brown, O’Rourke, & Rourke, 2013; Conti & Angelis, 2006) have conducted

studies with respect to lean implementation impacts on OHS. There is no consensus among

their results. For example, positive impacts of lean implementation on OHS have been report-

ed in recent literature (Hasle, Bojesen, Jensen, & Bramming, 2012; Nahmens & Ikuma,

2011). In this kind of research stream, authors declare that the cycle time will be reduced by

implementing lean, which leads to easing the work performance for the operators. Generally

speaking, easier means safer. On the other side, negative impacts of lean implementation have

also been reported in some literature (Conti & Angelis, 2006). Occupational stress, musculo-

skeletal disorders, and increasing accident rates are the most common negative impacts of

lean implementation on OHS.

With the disparities in synergy and trade-off impacts of lean on OHS, it can be concluded

that, the relationship between lean and safety is not clearly understood (Cudney, Murray, &

Pai, 2010). In literature, many authors claim that further research is needed to better under-

stand the impact of lean implementation on occupational health and safety (Landsbergis,

Cahill, & Schnall, 1999). Moreover, a more suitable approach to measure these impacts still

needs to be proposed and implemented. All the studies reviewed so far, however, suffer from

the fact that all have used traditional lagging indicators to measure the impacts of lean pro-

duction on OHS performance, and this method of analysis has a number of limitations. For

example, recent evidence (Hubbard, 2004; Sinelnikov, Inouye, & Kerper, 2015) suggests that

solely using lagging indicators is less useful in driving successful and continuous improve-

ment at organizations, but the bases of lean philosophy for enterprises are all about the con-

tinuous improvement of business processes, and one of them in any organisation is safety. A

company's safety program can be broken down into several safety-related processes. On the

other side, leading indicators monitor inputs to the process at advance stages before any ad-

6

verse outcomes have occurred. Therefore, for evaluating the impact of lean implementation

on safety, safety processes should be evaluated not safety outcomes. One key point is to use

leading indicators, which are also potentially useful to measure the strength of mechanisms

for synergy and trade-off between lean and OHS. Although extensive research has been car-

ried out on the relationship between lean production and occupational health and safety, no

single study exists that has systematically used leading indicators to measure OHS perfor-

mance.

Moreover, previous studies measuring lean impacts on OHS have suffered from a lack of

well-grounded theoretical considerations. The focus of most of these studies is using lagging

indicators, whereas a theoretical association between lean production and leading indicators

of OHS has received less research attention. Consequently, a comprehensive framework em-

bodying the full aspects of both safety and lean is still lacking in literature. All the studies re-

garding OHS performance measures in production systems have noted lean implementation

as an input and OHS performance as a direct output. For that reason, the underlying elements

influencing these variables (lean and safety) are lacking in the literature. With regard to this

issue, a complete and proper analysis of the relationship between lean and safety is unreacha-

ble.

Coherently, the motivations of this study are: to develop a comprehensive model representing

the relationship between lean and OHS, and highlight the importance of using leading indica-

tors in explaining and measuring the influence of lean implementation on OHS performance.

1.3 Research purpose and research questions

In order to fully understand the association between lean and OHS, a comprehensive frame-

work involving all influencing elements on both lean and safety is needed. Thus, we first

need to define the antecedents of OHS performance. Although various antecedents of safety

performance were stated in literature, a unified and classified framework for antecedents of

safety performance is lacking. Therefore, the first purpose of this study is to propose a clear

and consistent definition antecedent of OHS performance and then develop a classified

framework of these antecedents. Next, the thesis will define formative elements of maturity in

order to better understand the elements influencing lean implementation maturity in an organ-

ization. Finally, the third purpose is to determine the importance of using leading indicators in

measuring OHS performance.

In order to meet the objectives of this study, the following research questions are put forth:

- What are the antecedents of OHS performance in the workplace?

- How does lean implementation influence the antecedents of OHS performance?

- How does lean implementation affect OHS performance?

- What are the best leading indicators to measure the influence of lean implementation on

OHS performance?

1.4 Research contributions to knowledge and practice

First, regarding the first purpose of this study, a paper based on a literature review was con-

tributed which has important implications for practitioners and policy makers in the field of

occupational health and safety. Because of the increasing number of occupational accidents,

practitioners are experiencing profit-loss. Therefore, determining a framework for the antece-

7

dents of safety and health performance can help them to address this issue. Evidence shows

that determining the antecedents of safety performance can enable organizations to develop a

stronger plan to reduce the number of workplace accidents. However, while considerable re-

search has been devoted to the relationship between safety performance and other operational

activities, less attention has been paid to the antecedents of safety performance themselves.

Furthermore, a consistent definition and conceptualization of the antecedents of safety per-

formance was lacking in the safety literature; there was no clear and widely accepted defini-

tion for the concept of antecedent of safety performance.

Similarly, as described by Danna and Griffin (1999), a unified model or theory is still neces-

sary to develop the main constructs of health and safety in the workplace to better understand

the boundary of these factors and to clearly define the independency and interdependency

among these factors.

In a similar manner to the preceding issues, there is an argument among practitioners about

the synergies and trade-offs between lean initiatives and safety principles, which has resulted

in some challenges in the workplace. While some practitioners report the positive effects of

lean implementation on safety performance, including shortening cycle time and easing work

performance, others declare the negative effects of lean implementation on safety performan-

ce, such as occupational stresses, musculoskeletal disorders, and increasing accident rates.

In order to overcome the preceding challenges, a comprehensive literature review was con-

ducted in the field of occupational health and safety. A wide range of antecedents of safety

performance was listed and provided in a united framework including four certain categories:

working environment, task characteristics, workforce characteristics, and organizational fac-

tors. Because of these findings, four categories can now be treated as the antecedents of safety

performance. Moreover, regarding the challenges around a safety performance concept, a

model is developed in the current study to distinguish the safety performance conceptualiza-

tion.

From an academic perspective, this study provides a unified framework for antecedents of

safety performance, which helps future research to develop research streams regarding the

antecedents of safety performance. From a practical viewpoint, the results and findings of this

study, specifically the proposed framework for antecedents of safety performance, can be use-

ful for organizations to employ this framework while the assessment of safety performance is

being conducted in their systems.

Next, due to the second purpose of this study, the structure of formative elements of lean ma-

turity was lacking in the literature and therefore some challenges arose in this regard. In this

study, by conducting a literature review, three indicators were defined as the reflective ele-

ments of lean maturity including, fidelity, extensiveness, and experience. According to Ansari

et al. (2010), fidelity dimension for lean practices relates to the diffusion of each practice. In

this study, the adoption level of lean practices as an indicator to determine the fidelity of lean

practices was utilized. The extensiveness dimension relates to the extent of implementation of

lean practices. Also, the experience dimension was utilized to show the effectiveness of lean

practices in organizations over time.

The use of these three dimensions for lean maturity can be further employed by scholars to

more clearly illustrate and verify these dimensions for lean maturity.

8

Another contribution to the knowledge and practice of this study is related to the third pur-

pose of this study, highlighting the importance of using leading indicators for steering safety

performance. To capture the third purpose, the results chain model was employed, then sever-

al elements of this model were interrelated with safety concepts. Finally, a new model for il-

lustrating safety concepts was proposed.

The proposed model would have critical implications for both the academic community and

practitioners. The sequence of safety concepts in a holistic framework makes the concepts

explicit, thus helping researchers and practitioners understand the casual logic behind the

safety events. This framework also facilitates the discussion about monitoring and evaluating

safety efforts by showing the effective information of what needs to be monitored and evalu-

ated. A clear definition of both leading and lagging indicators is also understood from this

framework. The proposed structure shows the need to use both leading and lagging indicators

to steer safety performance. While leading indicators could be used for monitoring safety ef-

forts, lagging indicators are used for the evaluation of safety programs. Also, while leading

indicators are used at the shop floor level in organizations, lagging indicators will be useful at

the managerial level to make decisions about OHS policy.

The key findings of this study relate to the developed model for the relationship between lean

implementation and OHS performance. While previous studies address this association to

some degree, the present study represents the main critical elements influencing both lean and

safety variables. This model can serve as a comprehensive model for the interaction between

lean and safety that was lacking in publications. Interactions among lean maturity and ante-

cedents of safety performance indicate the importance of antecedents to measure the impact

of lean implementation on OHS performance, which was overlooked in previous studies.

Now, by realizing the role of antecedents of safety performance, practitioners would regard it

when the assessment of lean impacts on OHS is carried out. The value of each lean maturity's

dimensions affecting the antecedents was also reflected in this study.

Furthermore, the role of company's size and sector to moderate the impact of lean on anteced-

ents of safety performance was discussed.

More importantly, the way that lean implementation effects OHS performance was projected

in this study ,as well as, how various elements affect the relationship between lean and safety.

This comprehensive assessment overcomes the previous arguments about the relationship be-

tween lean and safety. Since the association was not clearly understood, different questions

arose among scholars. By now, it is expected that this part of the findings could answer those

questions.

Finally, by proposing the appropriate OHS leading indicators through which the impacts of

lean implementation on OHS performance would be measured, practitioners will benefit from

this special issue. There are no studies on the subject that assess the impact of lean implemen-

tation on OHS through the leading indicator. Therefore, this study enables practitioners to

monitor the impact of lean implementation before they result in negative effects on the health

and safety of the employees. Also, by using leading indicators, practitioners could reinforce

the possible synergy between lean initiatives and safety efforts.

9

1.5 Thesis outline

The thesis is structured into six chapters and two appended papers. Chapter 1 represents the

overall view of the study, including the research background, problem statement, research

questions, and contributions of the study to knowledge and practice. Chapter 2 introduces de-

tailed theoretical backgrounds of lean implementation, occupational health and safety, and

their relationship. Moreover, the previous literature illustrating the measurement approach of

lean implementation impacts on OHS performance will be discussed. Chapter 3 includes re-

search model, hypotheses, research methodology, and analysis procedures. Chapter 4 repre-

sents the findings of the study. Chapter 5 presents the discussion of the results. Lastly, impli-

cations of the study, study limitations, and suggestions for future studies will be displayed in

the Chapter 6.

Section 2 of the thesis includes two appended papers. The first paper presents the antecedents

of safety performance that were identified through a systematic literature review. This paper

addresses the first purpose of the study and was presented at the 2017 Industrial and Systems

Engineering Conference (IISE) in the USA.

The second paper shows the importance of using leading indicators to measure of safety per-

formance in the workplace. A clear definition for safety concepts is also illustrated in the pa-

per using the results chain model. The second paper was presented at the 8th International

Conference on Applied Human Factors and Ergonomics (AHFE 2017) in the USA.

Lastly, the appendices include appendix A, OHS leading indicators proposed to various types

of lean practices, and appendix B, the questionnaire of the present study, are represented.

10

CHAPTER2

STATE OF THE ART REVIEW

This chapter introduces the state of the art for lean concepts especially lean implementation

and lean maturity, and safety concepts, especially OHS performance, antecedents of safety

performance, and lagging and leading indicators. Next, the state of the art for the relation-

ship between lean and safety is presented. Recent approaches and methods for measuring this

relationship will also be addressed.

2.1 Lean concepts

Today, there is a new management philosophy in manufacturing that has been established in

response to the old failing style of production: mass production. Toyota Corporation has been

known as the father of modern lean movement. In order to be coined the "lean concepts," sig-

nificant steps were taken prior to Toyota's. Figure 1 shows the historic evolution of lean pro-

duction.

Figure 1: Evolution on lean production (source: Elbert, 2012)

As seen from Figure 1, Toyoda and Ohno rebuilt the Toyota Corporation after World War II

in 1950. They studied the Ford Production System (FPS) to constitute the concepts and tools

for the new production system, which was called Toyota Production System (TPS). As de-

scribed by Ohno in his book, Toyota Production System, the primary goal of this new system

is waste elimination from production (Shah & Ward, 2007). His idea was production in the

right amounts, at the time needed, and the unit needed. In 1988, the term "lean" was invented

by Krafcik to illustrate the Toyota production system. In 1990, a great book "The Machine

11

that Changed the World" was published by Womack, Jones, and Roos. In this book, the three

following concepts are presented:

- The origins of lean production

- Elements of lean production

- Diffusing lean production

This book greatly describes lean systems in detail. After this milestone work, numerous con-

tributions were published. In accordance with previous studies, Shah and Ward (2007) have

comprehensively proposed three underlying constructs (supplier related, customer related,

and internally related) and ten operational measures for lean production.

In the interest of employing lean philosophy, Cudney et al. (2013) proposes a lean tool pyra-

mid based on the knowledge needed to implement lean tools. Figure 2 illustrates this pyra-

mid.

Figure 2: Lean Pyramid (source: Cudney et al., 2013)

In order to explain the most common lean tools and techniques in detail, Table 1 is provided.

12

Table 1: Description of lean tools and techniques (source: http://www.strategosinc.com/lean_tools.htm)

Lean tools and

Techniques Purpose Description

5 S Reduce wasted time & motion at micro level.

Organized approach to housekeeping that ensures tools, parts and other objects are in known, optimum locations.

Value Stream Mapping

To visualize macro-level pro-cesses and their conformance to Toyota Production System (TPS) principles.

Uses a wide variety of symbols for many elements of TPS and helps determine how to employ these ele-ments in process improvement.

(SMED) To minimize setup time and cost thereby freeing capacity and en-abling the production of very small lots.

Rapid Setup uses Work Simplification and other con-ventional techniques to analyze each setup as a pro-cess and reduce time and other waste. It also tends to make setups more predictable and improve quality.

Kaizen

To improve work processes in a variety of ways.

Kaizen is a generic Japanese word for improvement or "making things better." In the context of Lean Manufacturing, it can apply to rapid improvement (Blitz) or slow continuous improvement (quick & Easy).

Pokayoke (mistake proof-

ing)

Prevent the occurrence of mis-takes or defects

Uses a wide variety of ingenious devices to prevent mistakes. An example is an automotive gasoline tank cap having an attachment that prevents the cap from being lost.

Process Mapping To visualize and understand the sequence and nature of events in a process at macro and micro levels.

Invented by Frank Gilbreth about 1913, process map-ping visually displays Value-Added and Non-Value Added steps using only a few clear symbols and lines. It lays the foundation for and guides process improvement.

Work Standardi-zation

To ensure that all workers exe-cute their tasks in the same manner and thus reduce varia-tion from differences in work method.

Organized approach to work specifications and in-structions. As practiced at Toyota, work teams care-fully specify the exact manner of performing each task and then adhere to it. Changes are made by the group when that group identifies improvements.

Visual Manage-ment

To provide immediate, visual information that enables people to make correct decisions and manage their work and activi-ties.

Visual Management uses a wide variety of signs, sig-nals and controls to manage people and processes. Traffic signs, lights and curbs are the most familiar examples.

Cellular Manu-facturing

Simplify workflow and concen-trate on a single product or nar-row family. It improves quality, inventory and many other pa-rameters.

Cellular Manufacturing organizes small work units of 3-15 people to build a single product or a narrow product family. Ideally the product is completed without leaving the work cell.

Kanban Schedule production and mini-mize work-in-process while en-couraging improvement in many areas.

Kanban establishes a small stock point (usually at the producing workcenter) that sends a signal when items are withdrawn by a downstream process. The producing work center simply replaces the items re-moved.

One-Piece Flow Reduce inventory internal to a workcell and forces improve-ments and work balance

One-piece flow is the concept of transferring only a single piece between process steps within a work cell with no accumulation of inventory. It forces near-perfect balance and coordination.

Total Productive Maintenance

Ensure uptime, Improve process capability and consistency

A maintenance program that combines predictive and preventive maintenance with problem solving and Total Quality.

http://www.strategosinc.com/lean_tools.htm

http://www.strategosinc.com/value_stream_mapping1.htm

http://www.strategosinc.com/value_stream_mapping1.htm

http://www.strategosinc.com/kaizen.htm

http://www.strategosinc.com/process_map_example.htm

http://www.strategosinc.com/toyota_corporate_culture.htm

http://www.strategosinc.com/toyota_corporate_culture.htm

http://www.strategosinc.com/visual_control.htm

http://www.strategosinc.com/visual_control.htm

http://www.strategosinc.com/cellular_manufacturing.htm

http://www.strategosinc.com/cellular_manufacturing.htm

http://www.strategosinc.com/kanban_1.htm

http://www.strategosinc.com/onepieceflow.htm

http://www.strategosinc.com/maintenance_reliability.htm

http://www.strategosinc.com/maintenance_reliability.htm

13

2.1.1 The status of employing lean philosophy in industries

Although lean philosophy was started from manufacturing sector in Toyota, it is not limited

to this sector (Cua, McKone, & Schroeder, 2001). Nowadays, a wide range of industries uses

lean tools and techniques in their systems across the world. For instance, Lawrence and Hot-

tenstein (1995) studied the relationship between lean implementation on operational perfor-

mance in 124 plants in Mexico. Similarly, Cua et al. (2001) showed a positive association be-

tween lean implementation and manufacturing performances at 163 plants in Italy, USA,

Germany, UK, and Japan. Also, enormous industries in China and India have started to utilize

lean techniques in their industries. As an illustration, in 2007, Taj investigated the application

of lean manufacturing in a wide variety of plants (electronics, pharmaceutical, telecommuni-

cation, etc.) in China and reported significant benefits in connection with lean implementation

(Taj, 2008). Ghosh (2012) conducted a study about lean manufacturing performance in Indian

manufacturing plants.

Moreover, applying lean tools and techniques is not limited to large organizations. Organiza-

tions, both small and large are applying lean philosophy (Anand & Kodali, 2008) . Figure 3,

which is extracted from Cudney et al. (2013), illustrates several case studies involving the

application of lean tools in various kinds of industries.

Figure 3: Lean tools and case studies (source: Cudney et al., 2013)

In accordance with the lean diffusion across different industries around all over the world,

enormous studies have been conducted in multiple research streams of lean systems. For in-

stance, some parts are related to the relationship between lean application and its effects on

operational performance (Dal Pont, Furlan, & Vinelli, 2008; Rahman et al., 2010; Taj, 2008).

Some others investigate the barriers and facilitators of to lean implementation (Aij, Simons,

Widdershoven, & Visse, 2013; de Souza & Pidd, 2011; Dora, Kumar, Van Goubergen,

Molnar, & Gellynck, 2013). Another kind of research team studies the synergies or trade-off

14

effects of lean techniques with concepts such as safety, six sigma, green manufacturing, and

resilience (Birkie, 2016; Cudney et al., 2010; Florida, 1996).

2.1.2 Lean maturity

While the advantages of lean implementation for productivity increase in organizations has

been stated, the time it takes to improve the performance is a challenging topic (Netland &

Ferdows, 2014). Although many companies have been implementing lean programs, each

company is different in size, location, process, culture, policy, and other circumstances.

Moreover, the competitive situations and the underlying expectations are different from one

company to another. These topics become important when managers decide to implement

lean programs. It is worth noting that misplaced expectations of how quickly lean programs

enhance operational activities would compromise the lean efforts. Thus, the managing lean

implementation process is more important than the program itself. In this regard, the lean ma-

turity concept comes up. In order to comprehensively study the impact of lean implementa-

tion on OHS performance, we needed to address the forming variables of lean maturity. Net-

land and Ferdows launched a study in 2007 at VOLVO Corporation. They investigated the

implemented Volvo Production System (VPS) that was based on lean principles in 19 coun-

tries across the world in Volvo factories. In this study, two variables-how widely and how

thoroughly lean is implemented- were proposed as the lean maturity's forming variables. This

study shows that resistance to change in initial stages of lean implementation is subsided by

thoroughly and widely diffusion in later stages. They also found a positive relationship be-

tween lean maturity and plant performance which is illustrated in Figure 4. This shape shows

that as much as lean programs are matured in the organization, the plant performance im-

proves highly. The findings of this study support the previous proposals and model regarding

lean maturity, such as the Lean Enterprise Transformation Maturity Model (Nightingale &

Mize, 2002) , which was developed by Lean Aerospace Initiative (LAI) at the Massachusetts

Institute of Technology in 2001. Figure 5 portrays the enterprise level road map to assist the

organizations to transform their efforts into lean implementation. Different elements affecting

lean maturity are shown in this figure. As seen, the visions experienced in the initial lean im-

plementation stages are different than later stages as lean programs are widely and thoroughly

implemented.

According to Ansari et al. (2010), how thoroughly and widely lean techniques are implement-

ed in an organization is in accordance with fidelity and extensiveness dimensions of a pro-

gram respectively. The fidelity of a program is related to the adoption level. Thus, we em-

ployed the adoption level of lean programs as the fidelity. Similarly, previous studies about

lean implementation (Shah & Ward, 2003; Netland & Ferdows, 2014) have employed the

adoption level as an indicator to better understand the diffusion level of each lean practice in

an organization. Extensiveness of a program is related to the extent of implementation in an

organization. Therefore, for lean programs, we employed the how expanse level as the exten-

siveness level of lean implementation.

15

Figure 4: The relationship between lean maturity and performance (source: Netland & Fer-

dows, 2014)

Figure 5: Lean maturity in enterprise (source: Nightingale & Mize, 2002)

Additionally, previous studies show that experience with lean implementation in an organiza-

tion could increase the effectiveness of lean implementation. For instance, a study conducted

in a health care sector shows that experiences of leaders in lean implementation is a key suc-

cess factor (Aij et al., 2013). As much as a company is experienced with lean implementation,

challenges are overcome regarding knowledge, employees' skills, expertise, information flow,

communication with suppliers, and customer improvement.

In another case study in Turkey in 2003, scholars used the experience level as the parameter

indicating lean maturity in industries. In that research, 17 companies of medium to large size

were investigated. The findings indicate a significant relationship between the time periods of

applying lean techniques and company's performance (Satoğlu & Durmuşoğlu, 2003). In the same vein, according to Ansari et al. (2010), implementing new practice in any organi-

zation faced a not well-understood situation in early phases, which later could be overcome

by capturing greater knowledge about the effectiveness of practice. Furthermore, cultural,

technical, and political fits seem plausible to become more and more common in late stages

compared to early stages of practice implementation (Ansari et al., 2010). Also, according to

the PDCA cycle, the more a company is experienced with lean practices, the more obstacles

that impede lean maturity can be overcome. Thus, we expect an effective lean practice im-

plementation over time. Therefore, in our study the organization experience with lean practic-

es could indicate the effectiveness and maturity of lean practices.

16

In conclusion, fidelity, extensiveness, and experience are identified as the formative con-

structs of lean maturity and were therefore applied in this study.

2.2 Occupational health and safety concepts

Occupational health and safety is described as the science and art of anticipation, recognition,

evaluation, and control of occupational hazards in the workplace. Occupational hazards are

classified in different ways, but the most common category is divided into four categories:

physical hazard, chemical hazards, biological hazards, and ergonomics hazards.

Physical hazards are defined as such factors in the workplace that (without necessarily touch-

ing) can injure the person. Some examples are; noise, heat, radiation, and, electricity.

Chemical hazards are related to the exposure with any chemicals in the workplace. Fumes,

gases, flammable liquids, and pesticides are some kinds of these hazards. Biological hazards

are bacteria, viruses, and other forms of biologic things that might exist in the workplace. Er-

gonomic hazards are related to the job factors that harm the body such as awkward posture,

improper workstation design, repetitive movement, and frequent lifting.

By having recognized these hazards, safety and health professionals enable to evaluate the

workplace conditions and finally control the occupational hazards. Figure 6 portrays a holistic

overview of the workplace and the placement of workers and OHS issues. As can be seen

from the figure, health and safety of the workers is being affected by all programs and pro-

cesses within the workplace. Any changes that happen in the workplace will influence the

health and safety of the workers. Therefore, safety professionals should be aware of hazard

creation in connection with implementation of a new program in the workplace.

Figure 6: A holistic view of the workplace and the importance of OHS (source: Erickson, 1996)

The advantages of safer and healthier workplaces, including productive workforce, improved

financial performance, and lower healthcare costs, have been discussed widely in safety liter-

ature (Vorley, 2008; Nahrgang, Morgeson, & Hofmann, 2011). In contrast to the advantages

of following OHS principles, enormous problems would occur as a result of ignoring those

rules. For example, nearly 6000 deaths and around 4 million work-related injuries and illness-

es have been reported in a given year in the United States (Craig, 2016). These problems af-

fect both the employers and employees. While the organizational cost related with poor safety

at work is incurring, employee's families are also indirectly suffering from overlooking OHS

17

principles in the workplace. As an illustration, the United States Department of Labor has re-

ported an annual cost of more than $53 billion for workers' compensation. Therefore, address-

ing the OHS is a big part of companies attempt that could affect not only companies perfor-

mance but also the society through the influences on employees' families. Equally important,

according to the changes in technology and life style, workplaces conditions are transforming

rapidly. As a result, new hazards have been brought to the employees. For this reason, safety

professionals should modify their approaches to measure safety performance more appropri-

ately. Remarkable progress has already been made to improve the state of occupational safety

in the workplace compared to the past. For example, the number of deaths has dropped from

21,000 in 1912 to 5,000 in 2014 in the workplace in the USA (Craig, 2016). Although this

progress is seemingly striking, there is still a need for establishing new strategies to control

workplace's risks. In conclusion, creating a safer and healthier workplace by establishing pol-

icies and programs would be helpful for individuals, their families, and employers and their

organizations, leading to productive communities.

2.2.1 Measurement of safety performance

The foundations of a business management process is measuring and controlling the perfor-

mance. The gaps between the acceptable level and current level of performance could be de-

termined by measurement (Janicak, 2009). Safety professionals are expected to establish

similar approaches for managing the safety activities and interventions.

In order to achieve a continuous improvement of safety performance in the workplace, cer-

tain strategies are employed. Goals setting, identification of the key activities/interventions to

reach those goals, and performance evaluation are common strategies. The most challenging

and fundamental issue among those strategies is the measurement of safety performance.

There are two common views regarding safety performance: the old view and new view. The

old view refers to the human error blamed for the accidents in the workplaces. By addressing

this view, humans were typically regarded as the only cause of accidents and injuries. As a

result, the underlying indicators for measuring the safety performance within the old view

were included the number of accident and injuries. Human error does not address the influ-

encing factors behind the human activities. Therefore, the reasons that lead to accident and

injuries remained unclear. After two catastrophic accidents, Chernobyl and Bhopal, research-

ers figured out that several other factors attribute to accidents in the workplace (Neal &

Griffin, 2006). It has been shown that the old view is unsuccessful today. The new view be-

lieves that the human error is a symptom not a direct cause of accidents, and regards deeper

root causes such as organizational factors, task characteristics, and working environment.

Compared to the traditional approach that has failed to identify the direct factors influencing

accidents and injuries in the workplace, the current holistic view provides a strong rationale

for recognizing and controlling the causes of accidents. This approach would help organiza-

tions prevent repeated accidents. In the new view, different tools and techniques to measure

safety performance have been developed. The common indicators that are used to measure

safety performance are known as leading indicators. These indicators address the underlying

elements that had been overlooked under the cover of human errors. As an illustration, safety

culture, management commitment, personality, and work design are elements that researchers

are currently working on (Wu, Chen, & Li, 2008; Törner, 2008).

18

2.2.2 Safety performance indicators

Reiman and Pietikäinen (2012) state: “An indicator can be considered any measure-

qualitative or quantitative-that seeks to produce information on an issue of interest. Safety

indicators can play a key role in providing information on organizational performance, moti-

vating people to work on safety and increasing organizational potential for safety” (p.1993).

Agumba et al. (2011) defined that health and safety performance indicators can be broadly

classified into two groups : lagging and leading indicators. Commenting on lagging indica-

tors, Sinelnikov et al. (2015) write: “The vast majority of OHS initiatives are still evaluated

relying primarily on lagging metrics, such as fatality and injury rates, despite the growing ac-

ceptance of the fact that these failure focused measures are less useful in helping organiza-

tions drive continuous improvement efforts. Leading indicators, on the other hand, offer

promise as an improved gauge of OHS activity by providing early warning signs of potential

failure and, thus, enabling organizations to identify and correct deficiencies before they trig-

ger injuries and damage” (p.240).

While lagging measurements can provide data about incidents after the fact, the question re-

mains regarding the value of these metrics as future predictors for safety in the workplace

(Hinze, Thurman, & Wehle, 2013). Mengolini and Debarberis (2008) note that an unbalanced

focus on lagging after-the-fact based measures may convey an unintended message that safety

prevention is less important. In recent years, there has been an increasing amount of literature

on using a combination of leading and lagging indicators for measuring OHS performance.

For example, Hinze et al. (2013), conclude that any firm that truly embraces the zero-injury

philosophy will readily consider using other measures than the traditional lagging indicators

of safety performance. They also note, “While the use of lagging indicators will continue, as

required by safety regulatory agencies and insurance companies, companies that track leading

indicators will be able to maintain a more accurate assessment of the effectiveness of the safe-

ty program or the safety process”. Similarly, in 2010, the American Petroleum Institute (API)

issued a new API standard (ANSI/API RP 754) on process safety performance indicators for

the refining and petrochemical industries (ANSI/API, 2010) where a method for selecting and

calculating leading and lagging indicators is offered. This case confirms the importance of

using a combination of leading and lagging indicators for a more favourable assessment of

the safety performance. This view is also supported by Reiman and Pietikäinen (2012), who

have attempted to draw fine distinctions between leading and lagging indicators based on a

sociotechnical system model. The Reiman and Pietikäinen model helped us to understand the

correlation between leading and lagging indicators while the combination of indicators for

OHS performance measurement is being used (Figure 7).

19

Figure 7: The extended system model showing the feedback from the indicators (source: Reiman

and Pietikäinen, 2012)

Another example of relationship between leading and lagging indicators can be found in the

publication "Step Change in Safety" (2003). These guidelines are based on an extensive anal-

ysis of the UK oil and gas industry. The purpose of the guidelines is to assist health and safety

professionals, advisors, plan developers and anyone wishing to understand lagging and lead-

ing performance indicators. As mentioned in these guidelines, there must be an association

between the inputs that the leading performance indicators are measuring and the desired lag-

ging outputs. There needs to be a reasonable belief that the actions taken to improve the lead-

ing performance indicator will be followed by an improvement in the associated lagging out-

put indicators. Finally, this guidance provides a framework for the exploration of association

between these two indicators, as shown in Figure 8.

tep Ssource: ( lagging indicators, and performanceelationship between leading, R :8 Figure

2003), afetySin hange C

20

2.2.3 Antecedents of safety performance

Evidence shows that determining the antecedents of safety performance can help develop a

stronger plan to reduce the number of workplace accidents. Nevertheless, less attention has

been paid to the antecedents of safety performance (Neal & Griffin, 2002). Further, the exist-

ing safety literature lacks clear and consistent definitions and conceptualizations (Christian,

Bradley, Wallace, & Burke, 2009). There is not a clear and widely accepted definition for the

term of “antecedent of safety performance” (Gibb, Haslam, Gyi, Hide, & Duff, 2006). For

example, personal factors, such as traits and attitudes, were traditionally mentioned in the

safety literature as the antecedents of safety performance, but after two catastrophic accidents

(Chernobyl and Bhopal), researchers were warned of other influencing factors for accidents

such as management practices and work conditions (Neal & Griffin, 2006). Therefore, re-

searchers are now faced with a variety of complex antecedents of safety performance (per-

sonal characteristics, management practices, and work conditions among others), which are

difficult to identify in an integrated framework. In this study, a wide range of literature was

extracted from databases to finally propose a united and comprehensive framework for the

antecedents of safety performance. Table 2 shows a summary of the finding from the litera-

ture review.

As can be seen from Table 2, a wide variety of antecedents of safety performance exists in the

literature. Therefore, it seems that creating a clear and unified framework for classifying the

antecedents of safety performance could be useful.

2.2.4 Classifying antecedents of safety performance

2.2.4.1 Workplace environment

The main factors forming the concept of working environments are related to four factors,

including physical factors (e.g., noise, heat, lighting), chemical factors (e.g., dust, chemical,

smoke), ergonomic factors (e.g., workstation design, chairs), and biological factors (e.g., vi-

rus, bacteria) (Sparks, Faragher, & Cooper, 2001). The effect of each of these factors on OHS

performance has been widely reported in the literature. For instance, Shikdar and Sawaqed

(2003) show the importance of working environments factors on the rate of occupational ac-

cidents and injuries in the workplace. In another study, Dann and Griffin (2002) highlight the

role of working environments with biological factors and chemical factors that influence

health and safety performance. In the same vein, the significance of physical factors on pre-

venting occupational accidents at construction sites is shown by Wu et al. (2010). Also, nu-

merous studies were carried out to investigate how ergonomic factors affect OHS perfor-

mance. As an illustration, Marek Dźwiarek (2004) analyzes the accidents caused by improper

functioning of control systems, which consist of the errors made by designers. In summary,

these four elements are kept together as one unit noted as working environment in the present

study.

21

Table 2: Antecedents of safety performance extracted from the literature

2.2.4.2 Workforce characteristics

In regards to domino theory, which was developed by Heinrich in 1930, humans are the key

reason behind accidents. Although this definition has been criticized by other authors, such as

Peterson, the human factor is still being discussed as the main cause of accidents (McClay,

1989; Norman, 1981; Recht, 1966). By searching the literature, we also found several papers

that mention the importance of people's role as an antecedent for safety performance (Chris-

tian et al., 2009; Neal et al., 2000). On the other side, by referring to the theory of individual

differences in task and contextual performance, Motowidlo et al. (1997) state, "Individual dif-

ferences in personality and cognitive ability variables, in combination with learning experi-

Reference Antecedents of Safety Performance Embrey et al. (1994) Operating environment, task characteristics, operator characteristics, organ-

izational and social factors

Hofmann et al. (1995) Individual factors, micro, and macro organizational factors Kraus (1995) culture, management system, exposure

Manuele (1997) Culture, management system, task performance practices

Sawacha et al. (1999) Historical factors, economical factors, psychological factors, procedural

factors, organizational factors, environmental factors

Dana and Griffin (1999) Work setting, personality traits, occupational stress Griffin and Neal (2000) Individual-level factor, group and organizational factors Goldenhar et al. (2003) Job-task demands, organizational factors, physical/chemical stressors Ai lin Teo et al. (2005) Policy, process, personnel, incentive

Haslam et al. (2005) Worker (work team), workplace, materials, equipment, originating influ-ences (safety culture, management)

Gibb et al. (2006)

Work team, workplace, equipment , material

Griffin and Neal (2006) Organization factors, individual factors

Nahrgang et al. (2007)

Job demands, Job resources

Wu et al. (2009)

workplace, work team , equipment , material

Clarke (2010) Job characteristics, work group, leader , organizational structure

Hansez and Chmiel (2010)

Job demands, job resources, management commitment

Fernández-Muniz et al. (2011)

Management's commitment, incentives, work pressure, communication

Clarke(2013) Leadership styles, organizational climate

Card (2013) Person, organization, technologies and tools, process, environment, tasks El-nagar et al. (2015)

Worker factors, environmental factors, organizational factors

22

ences, lead to variability in knowledge, skills, and work habits that mediate effects of person-

ality and cognitive ability on job performance." For example, people with type A behaviour

patterns are "hard-driving, competitive, job involved, and hostile." Complementary to this,

several studies have been conducted on the relationship between personality differences and

safety issues (Friedman & Rosenman, 1974; Orpen, 1982). The items extracted from litera-

ture that are related to the workforce characteristics are as following: motivation, emotional

control, risk-taking, extraversion, neuroticism, physical condition, and age.

2.2.4.3 Task characteristics

The study conducted by Parker et al. (2001) on investigating the direct and indirect effects of

work characteristics on workplace safety suggests that work characteristics are an important

antecedent for safety performance in the workplace. The result of this study is consistent with

Clarke (2010) who concludes that job characteristics, such as job control, autonomy, and

challenge, have a strong influence on perceived safety climate and safety outcomes. Addi-

tionally, in the Barling and Zacharatos model (1999), they propose ten practices for enhanc-

ing safety performance. Some of them are related to the work characteristics, such as job au-

tonomy and high-quality jobs. In the same way, work characteristics have been reported

(Betcherman, Mcmullen, Leckie, & Caron, 1994) as a critical factor to lower accident rate at

the organizational level. Besides, the effect of ergonomic factors such as fatigue, shift work,

equipment design, and workload on safety performance is inevitable. Numerous works have

been conducted to investigate the relationship between ergonomic factors and safety perfor-

mance (Sagot, Gouin, & Gomes, 2003; Hofmann, Jacobs, & Landy, 1995; Yeow & Sen,

2003). In our work, we also found many authors who address the work characteristics as an

antecedent for safety performance (Haslam et al., 2005).

2.2.4.4 Organizational factors

In order to address the importance of organizational factors in safety performance, Hofman et

al. (1995) state, "Although individual safety-related attitudes and behaviours are certainly im-

portant and no doubt to be addressed by the organizations, there are clearly larger organiza-

tional variables that impact safety performance." Therefore, the interest in knowing the ef-

fects of management and organizational factors on safety performance is rising (Andel,

Hutchinson, & Spector, 2015) . For example, in the model developed by Embrey (1992), or-

ganizational factors have been introduced as latent factors that induce unsafe systems and

human errors. In the same vein, Paté-Cornell (1990) argues that organizational factors are the

root of failures of the critical engineering system. Likewise, for demonstrating the significant

role of leadership, as an organizational factor, Zohar and Luria (2003) argue that the leader of

organizations through their supports for safety can be the major source of employee climate.

Similarly, a number of researches indicate the influence of leadership practices on safety-

related behaviours of employees (Kapp, 2012; Hofmann, Morgeson, & Gerras, 2003; Zohar,

2003). By searching the literature, we also found a number of organizational factors that are

proposed as the antecedents of safety performance. Management issues, culture, and commu-

nication are the most common organizational factors extracted from the literature.

In accordance with the other objectives of this study, a new framework for antecedents of

safety performance was proposed (Figure 9). As illustrated, four elements (working environ-

ment, task characteristics, workforce characteristics, and organizational factors) construct the

23

blocks of the framework. Therefore, these four elements are treated as the antecedents of

safety performance.

Figure 9: Classification of safety performance's antecedents and their relationship to safety

performance

This model provides a unified framework for antecedents of safety performance, which helps

future research develop the research streams of the antecedent of safety performance. From a

practical viewpoint, this model can be useful for organizations to employ while the assess-

ment of safety performance is being conducted in their systems.

2.3 Relationship between lean and safety

The transition from traditional to lean production requires a redesign of production processes,

worker activities, and site layout, all of which can affect site safety and health.

The issue of occupational health and safety (OHS) has been a controversial and much disput-

ed subject when it comes with the investigation of benefits and impacts of lean implementa-

tion at shop floor level. A few studies have investigated the association between lean and

OHS, and a systematic understanding of how lean contributes to or impairs OHS is still lack-

ing.

Numerous research and lines of thought exist in literature regarding the association between

safety and lean. The two sections below discuss about the current body of theoretical contri-

butions. A first stream claims that lean production has negative effects on OHS, while the

second tries to understand and assess mechanisms of positive effects of lean on OHS.

Brown et al. (2013) in their paper provide evidence of the negative effects of lean manufac-

turing on workplace health and safety in Chinese industry. They adopted a case-study ap-

proach to obtain further in-depth information on the association between lean manufacturing

and occupational health safety. This study analyzed the data from a 13000 worker factory in

Northern Province. Following the implementation of the lean approach, a significant increase

24

in the occupational hazards was recorded. The findings observed in this study confirm the

negative impact of LP on OHS.

This view is also supported by Landsbergis et al. (1999), who conducted a literature review

on the relationship between LP and OHS from 1976 to 1998. They reviewed 38 papers, of

which 13 report evidences from the automotive industry, 11 from the health care industry, 1

from telecommunications, and 13 from other manufacturing industries. The review highlights

a significant positive correlation between LP and high levels of stress on workers. Authors of

reviewed papers offer several possible explanations for this result: increased workload, an

increase in repetitive work, and a decrease of rest breaks in lean manufacturing systems.

In the same vein, Conti and Angelis (2006), reported the effects of lean production on worker

job stress. The analysis was based on the conceptual framework proposed by Karasek (1989)

about job stress. This study uses qualitative analysis in order to gain insights into the impacts

of lean production on worker job stress. Data were gathered from multiple sources at various

organizations. A semi-structured interview was conducted with management. Also, a ques-

tionnaire was completed by 1,391 workers. Data were gathered from 21 sites of four UK in-

dustry fields. The results of this study confirm the association between negative impacts of

lean implementation on worker job stress as an indicator of occupational health and safety

outcome.

Other authors concentrated on collecting and discussing experimental evidences on the posi-

tive impacts of LP on working conditions and OHS, as discussed in the following section.

Womack et al. (1990), in their book "The Machine that Changed the World: The Story of

Lean Production." note that there are some positive results about the relationship between

lean and working conditions, such as job autonomy, worker participation, empowerment, and

job enlargement.

To determine the effects of lean production on working conditions, Berggren (1993) de-

scribed some positive impacts, such as job security, its egalitarian character, management

considerations to worker proposals, attentive selection, and highly qualified workers.

To better understand the mechanisms of lean and its effects on safety, Cudney et al. (2010)

conducted an online survey to check the impacts of lean approach on safety. The lean areas

that they have mentioned are value stream mapping (VSM), one-piece flow, material han-

dling, and single minute exchange of dies (SMED). Interestingly, 88% of those who were

surveyed indicated that they had observed a positive impact of their lean activities on the

health and safety performance of their workers.

Overall, the current body of literature highlights the complexity of lean impacts on OHS, and

many studies indicate the need for a better understanding of the mechanisms that drive the

relationship between the two.

In this direction, only in recent years, few authors have begun to provide some explanatory

models for the relationship between lean and safety. For example, an important study address-

ing the integration of lean and safety was released in ANSI (2007). The aim of this report was

to provide guidelines to industries who wish to concurrently address lean and safety concerns

when using machinery. The report proposes a risk assessment framework to address lean and

safety concerns (Figure 10).

25

ANSI, 2007) (source: eanl process model for safety and lGenera :10 Figure

In the same vein, the relationships between LP and OHS may partly be explained by consid-

ering not only the lean practices but also implementation and context of lean (Figure 11), as

suggested by Hasele et al. (2012).

source: ( being-Lean effects on working environment and employee health and well :11 Figure

Hasele et al., 2012)

This view is supported by Longoni et al. (2013) who wrote about the effects of lean practices

on safety climate, which eventually results in safety outcomes. Although in this paper they

also discuss operational outcomes, according to Zohar (2003), they claimed that safety cli-

mate is a predictor of future safety outcomes (Figure 12).

2013)Longoni et al., (source: Relationship between lean practices and safety outcome: 12 Figure

26

2.4 The lack of a generalized model of the relationship between lean and OHS

Lean manufacturing works as a double-edged sword; despite its benefits on improving

productivity and profitability in the workplace, its downsides might jeopardize employees

health and safety. Although many studies have been conducted on the positive side of lean

manufacturing, less attention has been devoted to the drawbacks of this new system of work

organization. Therefore, the causal association between lean and safety has remained unclear.

Within conducted studies, the disparities in synergies and trade-offs of lean on OHS addition-

ally conclude that the relationship between lean and safety is not clearly understood yet. Recently, a serious challenge has arisen on the costs borne by society because of lean imple-

mentation, such as, occupational injuries and diseases. Therefore, the association between

these two concepts needs to be forcefully addressed. By knowing the robust association be-

tween lean and safety, negative effects of lean implementation on OHS performance could be

minimized and, furthermore, positive impacts of this relationship could be maximized. By

knowing the association, organizations could have great improvements in work conditions by

refining the lean tools and methods without ignoring the basic principles of lean manufactur-

ing. Consequently, lean and safety goals are addressed at the same time. Also, having a de-

veloped framework for this association could result in making lean manufacturing more hu-

mane. A developed framework could also help health and safety administrations, such as

OSHA, NIOSH, to regulate enforcements to prevent impairing employees health and safety.

Furthermore, integrating safety concepts to lean principles can be done through the developed

model and drives an additional motivation to link lean and safety theories.

Previous models lack in explicit relationship between lean and safety. There is no agreement

on the formative variables of both lean and safety. Because of this reason, the results of

measuring the impact of lean implementation on safety performance are disparate. Therefore,

in literature, many authors claim that further research is needed to better understand the im-

pact of lean implementation on occupational health and safety.

Moreover, a suitable approach to measure these impacts still needs to be proposed and im-

plemented. Previous studies of measuring lean impacts on OHS have suffered from a lack of

well-grounded theoretical considerations. The focus of most of these studies has been lagging

indicators, whereas a theoretical association between lean production and leading indicators

of OHS has received less research attention.

The existing studies regarding the relationship between lean and safety have addressed the

lean implementation as an input and OHS performance as the output. In this respect, lean has

been retained as one single concept without regard to associated factors. The forming ele-

ments of lean maturity, such as how wide and how thorough, therefore remain ambiguous.

This situation could result in confusing assessment of lean impacts on OHS performance.

When there is no comprehensive information on all aspects of lean implementation, the im-

pacts of lean could not be properly addressed and, consequently, the risk of false interpreta-

tion of lean effects on workers health and safety could exist.

27

Furthermore, antecedents of safety performance are not taken into account in previous studies

in connection with lean impacts. Most studies regarding lean and safety association have con-

sidered some parts of safety performance. Not all aspects of safety performance have been

examined while the measurement of lean implementation effects is being undertaken. This

situation also leads to inappropriate conclusions for the association between lean and safety.

Given the above notes, the need for the development of a comprehensive model covering all

formative elements of both lean and safety is highly significant. This reason was one of the

main motivations for this study. In order to develop the model, first, the antecedents of safety

performance were extracted from the literature and then placed in a new model illustrating the

relationship between the antecedents and safety performance. Next, three formative con-

structs for lean maturity were identified by searching within lean literatures. Fidelity, exten-

siveness, and experience are the three main elements constructing the lean maturity variable.

Finally, through utilizing SmartPLS software the model was developed. We expect that future

practitioners and academic communities employ this developed model to appropriately meas-

ure the impact of lean implementation on OHS performance. Also, this model highlights the

importance of leading OHS indicators to be utilized within the measurement approach.

Moreover, this study is the first study that attempts to include all associated factors influenc-

ing both lean and safety initiatives. The results of this study could be important for both re-

searchers and practitioners.

28

CHAPTER3

RESEARCH DESIGN

This chapter, first, presents the research model and its underlying hypotheses. Second, the

research methodology will be discussed along with the procedures adopted for data pro-

cessing and analysis.

3.1 Research model and hypotheses

Bacharach (1989) defines theory as "a system of constructs and variables in which the con-

structs are related to each other by propositions and the variables are related to each other by

hypotheses. The whole system is bounded by the theorist's assumptions." (p. 510). The im-

portance of having theory for researches is stated by Wacker (1998) as follows:

….."(1) It provides a framework for analysis; (2) it provides an efficient method for field de-

velopment; and (3) it provides clear explanations for the pragmatic world." (p. 362). Conse-

quently, to fully understand the association between lean implementation and OHS perfor-

mance in this study, we first need a theoretical theory illustrating the relationship in great de-

tail. Although previous studies in this regard have proposed some theoretical frameworks,

almost none have been entirely convincing. Most of them address lean implementation as an

input and OHS performance as an output. That is, the influencing elements of lean implemen-

tation and OHS performance are not provided. In the interest of capturing an appropriate

judgment on the impacts of lean implementation on OHS performance, according to Gertler

et al. (2011) we analyze three items: how, where, and when in connection with lean imple-

mentation. This issue has also been supported by several scholars in lean literature. Recently

a new concept has been introduced by Netland and Ferdows (2014) to illustrate the quality of

lean implementation in organizations. They propose the forming elements how thoroughly

and how widely in measuring lean maturity. According to Ansari et al. (2010) fidelity's di-

mension is related to the degree of completeness of each practice as it is currently implement-

ed by the organization. Therefore, we combine two concepts together and propose fidelity as

a concept illustrating how thoroughly lean practices have been implemented in an organiza-

tion.

As explained for the where item, Netland and Ferdows (2014) have utilized the "how widely"

item to illustrate the degree or extent of implementation of the lean practice in an organiza-

tion. This definition is similar to extensiveness as described by Ansari et al. (2010). There-

fore, we again combine two items, where and how widely, into one item: extensiveness. As a

result, extensiveness relates to the degree or extent of implementation of lean practices in an

organization (from small areas to an entire organization).

For the when item, which is associated with the length of time that lean practices have been

implemented in an organization, we take into account this item as the experience to imple-

ment lean practices in an organization. Prior studies show that experience with lean imple-

mentation in an organization could increase the effectiveness of lean implementation. For in-

stance, a study conducted in a health care sector shows that experience of leaders in lean im-

plementation is a key success factor (Aij et al., 2013). If a company is experienced with lean

http://www.sciencedirect.com/science/article/pii/S0272696398000199#BIB1

http://www.sciencedirect.com/science/article/pii/S0272696398000199

29

implementation, challenges are overcome due to employees' knowledge and skills, expertise,

information flow, communication with suppliers, and customers improvement (Kovacheva,

2010). In the same vein, according to Ansari et al. (2010), implementing new practice in any

organization faced with a not well- understood situation in early phases which later could be

overcame by capturing greater knowledge about the effectiveness of practice. Furthermore,

cultural, technical, and political fit seems plausible to become more and more common in late

stages compared to early stages of practice implementation (Ansari et al., 2010). Additional-

ly, according to the PDCA cycle, the more a company is experienced with lean practices, the

more obstacles that impede lean maturity can be overcome. Thus, we expect an effective lean

practice implementation over time. Therefore, in our study the organization's experience with

lean practices could indicate the effectiveness and maturity of lean.

Given the above consideration, three elements, fidelity, extensiveness, and experience, con-

struct the forming indicators of lean maturity, which as shown in the research model.

An additional two specific factors that should be considered in adopting new work systems

like lean practices are the size of the business unit and the business sector. Both of them are

controversial topics.

In the next step of developing research model, the OHS performance variable was addressed.

Previous studies in connection with the relationship between lean and safety mostly have kept

this variable solely. That is, in spite of the importance of the antecedents of the OHS perfor-

mance, they are lacking in analyses. Consequently, this study goes through the antecedents

and highlights its position. After conducting a literature review, four main antecedents of

safety performance were determined, which are described in more detail in Chapter 2. They

are working environment, task characteristics, workforce characteristics, and organizational

factors. To define mediator, Reuben and Kenny (1986) state, "In general, a given variable

may be said to function as a mediator to the extent that it accounts for the relation between

the predictor and the criterion. A variable functions as a mediator when it meets the following

conditions: (a) variations in levels of the independent variable significantly account for varia-

tions in the presumed mediator, (b) variations in the mediator significantly account for varia-

tions in the dependent variable" (p. 1176), which is consistent with the role of antecedents in

safety performance. We consequently have taken the antecedents of safety performance into

account as the mediators between lean implementation and safety performance, as is shown

graphically in the research model. It should be mentioned here that each of these four ante-

cedents has its own underlying aspects that have been extracted from the literature. The work-

ing environment variables include physical, chemical, biological, and ergonomic aspects. The

task characteristics contain type of task, time, job demands, and equipment. The workforce

characteristics include four aspects; risk taking, safety knowledge, safety motivation, and lo-

cus of control. Lastly, organizational factors are policy, communication, management, and

culture.

Additionally, four traditional indicators demonstrating the status of safety performance in an

organization are employed in this study: recordable injuries, worker's compensation cost, ac-

cident records, and lost working days.

30

Extant literature has shown that application of lean practices is not equal to large and small

firms (Matt & Rauch, 2013). While several scholars declare the difficulties of implementing

new operational practices in large firms due to complication of the process and administrative

tasks in this kind of firms, others show a positive relationship between the size of firms and

the success of new woks system's implementation. Shah and Ward (2003) state" Large firms

are more likely to implement lean practices than their smaller counterparts" (p. 133), and the

difficulties of implementing lean practices in small enterprises have been reported by Matt

and Rauch (2013) in northern Italy.

To summarize, the role of company size needed to be addressed while reviewing the impacts

of lean implementation on other subjects like OHS performance.

The business sector also brings various views on the success or failure of lean implementa-

tion. Although lean manufacturing has originated from Toyota, a manufacturing company in

Japan, nowadays various business sectors employ lean tools and techniques across the world

(Hallowell et al., 2009). For instance, Poksinska, (2010) has studied the current state of lean

implementation in the healthcare sector. The barriers, challenges, and outcomes of imple-

menting lean practices have been analyzed in this study. In another example, Kim (2002) as-

sesses the implementation of lean practices within construction sites. The results of this study

show the importance of lean implementation in improving associated factors with project ac-

complishment. Finally, the author recommends using lean tools and techniques for construc-

tion sites. Additionally, although the utilization of lean techniques in service companies is in

its early stages, several studies (Piercy & Rich, 2009; Portioli-Staudacher, 2009) show the

benefits of this management system employing in the service sector.

Therefore, given the reasoning above, the significance of two factors (size and sector) could

not be ignored in studying the success or failure of lean tools and techniques. In order to ad-

dress this issue, the present study has taken into account the role of company size and sector

in measuring the impacts of lean implementation on OHS performance.

On the nature of moderators, Reuben and Kenny (1986) state " In general terms, a moderator

is a qualitative (e.g., sex, race, class) or quantitative (e.g., level of reward) variable that af-

fects the direction and/or strength of the relation between an independent or predictor variable

and a dependent or criterion variable. Specifically within a correlational analysis framework,

a moderator is a third variable that affects the zero-order correlation between two other varia-

bles" (p. 1174). Size and sector of firms are consistent with this definition and are therefore

taken into account in the existing study as moderators of the relationship between lean im-

plementation and OHS performance.

By considering all these factors, the complete research framework is depicted as follows:

(Figure 13)

https://repositories.lib.utexas.edu/discover?filtertype=author&filter_relational_operator=equals&filter=Kim%2C+Daeyoung

31

Figure 13: Research framework

The next step of the study is the hypothesis formulation. Eleven hypotheses are constructed

for this study in response to the research questions and research model. Four hypotheses re-

late to the relationships between lean implementation and the four antecedents of safety per-

formance, one hypothesis is linked to the relationship between lean implementation and OHS

performance, four hypotheses are in connection with the mediation effects of the antecedents,

and two hypotheses are linked to the moderation effects of company size and sector.

Hypotheses:

H1: Lean implementation significantly influences OHS performance.

H2: Lean implementation significantly influences working environment.

H3: Lean implementation significantly influences task characteristics.

H4: Lean implementation significantly influences workforce characteristics.

H5: Lean implementation significantly influences organizational factors.

………………………………………………

Mediation hypotheses:

H6: Working environment significantly mediates the relationship between lean implementa-

tion and OHS performance.

H7: Task characteristics significantly mediate the relationship between lean implementation

and OHS performance.

H8: Workforce characteristics significantly mediate the relationship between lean implemen-

tation and OHS performance.

H9: Organizational factors significantly mediate the relationship between lean implementa-

tion and OHS performance.

…………………………………………………………………. Moderation hypotheses:

H10: There is a significant categorical moderating effect of business sector on the relation-

ship among model constructs.

H11: There is a significant categorical moderating effect of business size on the relationship

among model constructs.

32

3.2 Research methodology

With respect to Figure 14, by having formulated the model of the relationship between lean

implementation and OHS performance, the next step is to go through a theory-testing process.

Research methodology uses information from real entities to build theories on relationships or

to test them in the real world (Kumar & Phrommathed, 2005).

Figure 14: Theory building and fact finding (source: Kumar & Phrommathed, 2005)

Statistical empirical research (panel study, focus group, survey) contributes to the testing of

theories and hypotheses on statistical basis in wide samples. According to the definition of

survey, it is an instrument for gathering qualitative or numeric information in a wide group of

subjects (Bartlett, 2005), where structured information is asked directly to people in a provid-

ed sample of population.

Three certain types of survey are provided: exploratory, descriptive, and explanatory

(Pinsonneault & Kraemer, 1993). Exploratory studies try to figure out new topics and con-

cepts. Indeed, the purpose of an exploratory approach is collecting various viewpoints from a

population in order to design a more effective survey in the future. Descriptive survey is

about figuring out the situation or events that are happening in a population. The main pur-

pose of this survey is finding the status of an event or situation's distribution in a population.

Descriptive questions are constructed to find the actual facts, not theory testing. Explanatory

survey, which is also called confirmatory study or theory testing, tries to explain the relation-

ship between variables. So, first, a theoretical framework needs to be developed about the

form of how and why the variable should be related. Contextual theory within the explanatory

survey not only quantifies the cause and effects situation between variables, but also deter-

mines the positive or negative effects of one variable over the other variables. As a result,

questions in a survey instrument (questionnaire) are constructed in a manner not only to quan-

tify the casual relationship between variables, but also to explain the reasoning of the rela-

tionships.

According to the above considerations, this study is consistent with the concepts of explana-

tory research. That is, within the lean implementation and occupational health and safety

fields there are adequate information and research studies that have been conducted on the

33

impact of lean on worker health and safety (Brenner, Fairris, & Ruser, 2002; Lewchuck,

Stewart, & Yates, 2001). Therefore, several theoretical models exist in this area (e.g,

ANSI/API 2010; Longoni et al., 2013). Moreover, several empirical research studies have

focused on the methods of measuring the lean implementation impact on OHS (Conti &

Angelis, 2006; Saurin & Ferreira, 2009). Consequently, this study, attempts to test the devel-

oped theory of the relationship between lean implementation and OHS performance in great

detail compared to previous studies through gathering information from lean industries. Since

prior studies have shown various positive and negative effects of lean implementation on

OHS performance, this study attempts to confirm these effects in comprehensive detail

through addressing the antecedents of safety performance. Moreover, existing studies lack the

lean maturity concept, which plays an important role in the quantification of lean impact on

the OHS performance.

In short, this study specifies the following conditions through explanatory survey:

-How lean implementation affects the antecedents of safety performance?

-How lean implementation relates to the OHS performance?

-Why antecedents of safety performance are important for measuring the impacts of lean im-

plementation on OHS performance?

-Why the lean maturity's elements are important for measuring the impacts of lean on OHS?

-How the effects of lean implementation on OHS performance can be measured in a more ap-

propriate approach?

-How company size is significant to moderate the lean implementation effects on OHS per-

formance?

-How company sector is significant to moderate the lean implementation effects on OHS per-

formance?

3.2.1 Introduction to PLS-SEM

Following the progress in research on statistics, Structural Equation Modeling (SEM) has

been introduced as a second-generation method for multivariate data analysis (Chin, 1998).

This method has several distinct advantages compared to the statistical first-generation tech-

niques, such as factor analysis, discriminant analysis, or multiple regression. While the tradi-

tional and old methods were only able to analyze one level of the association between inde-

pendent and dependent variables, SEM methods enable researchers to analyze multiple de-

pendent and independent variables simultaneously. Within the modern version of data analy-

sis researchers have more flexibility to interplay the data and theory in comparison with tradi-

tional methods (Wong, 2013). In the traditional methods, researchers need a strong theoretical

background to build the research model, but in SEM, less confident theories also could be

used to structure the research model. Moreover, normal distribution of the data is one of the

main requirements of the first-generation methods. Because of this, researchers experience

some problems during data analysis. In contrast, SEM methods are not grounded on the nor-

mal distribution of the data. Therefore, the academic community is more interested in these

kinds of statistical methods. Another advantage of SEM relates to its ability to evaluate the

34

measurement model in addition to the structural model assessment. Chin (1998) presents the

advantages of SEM methods as follows:

In general, SEM- based approaches provide the researchers with the flexibility to perform the

following: (a) model relationships among multiple predictor and criterion variables, (b) con-

struct unobservable latent variables, (c) model errors in measurements for observable varia-

bles, and (d) statistically test a priori substantive/theoretical and measurement assumptions

against empirical data (i.e., confirmatory analysis) (p.297).

Within SEM framework, there are two submodels: inner and outer models. While the former

refers to the association between independent and dependent variables, the latter specifies the

relationship between each variable and its observed indicators (Wong, 2013). Moreover, two

types of variables are defined in the SEM method. Exogenous variables are those that arrows

point outward and endogenous variables, in contrast, have a path leading to it. Also, SEM in-

cludes a type of measurement scales: formative and reflective. For the definition of these

scales, we refer to Wong (2013): "If the indicators cause the latent variable and are not inter-

changeable among themselves, they are formative. In general, these formative indicators can

have positive, negative, or even no correlations among each other" (p.14), and "If the indica-

tors are highly correlated and interchangeable, they are reflective and their reliability and va-

lidity should be thoroughly examined" (p.15).

There are four distinct approaches to SEM. The first and most widespread is the covariance-

based approach called CB-SEM, where several softwares are utilized such as LISREL,

AMOS, and EQS. The second approach is based on the analysis of variance and SmartPLS,

PLS Graph software packages are used in this context. The third one known as GSCA, which

is component-based, and VisualGSCA is the main software for this approach. Lastly, the

fourth approach relates to non linear structural modelling, and the NEUSREL package is used

for this approach.

Among these approaches the CB-SEM is widespreadly used. However, meeting the require-

ments for using this approach is often difficult. A large sample size, normal distribution of

data, and a strong model are three main problems that researchers are faced with. Additional-

ly, since we know there is insufficient information for the relationship among variables in ex-

ploratory studies, CB-SEM would not be an effective approach for analysis. As a result, re-

searchers currently use the second approach, SEM-PLS.

The four main logical reasons why PLS a good alternative to CB-SEM is stated by Wong

(2013) as follows:

1. Sample size is small.

2. Applications have little available theory. 3. Predictive accuracy is paramount. 4. Correct model specification cannot be ensured (p.3) In spite of these advantages, we should also consider the limitations of the SEM-PLS, which is again stated by Wong (2013): 1. High-valued structural path coefficients are needed if the sample size is small. 2. Problem of multicollinearity if not handled well.

35

3. Since arrows are always single headed, it cannot model undirected correlation. 4. A potential lack of complete consistency in scores on latent variables may result in biased component estimation, loadings and path coefficients. 5. It may create large mean square errors in the estimation of path coefficient loading (p.3) Using SEM-PLS is becoming gradually popular among the academic community. Applied

research projects are part of a common research area that use the PLS.

3.2. 2 Reasons for using PLS-SEM in the existing study

Several studies within lean manufacturing literature have utilized the SEM approach. For in-

stance, Braunscheidel and Hamister (2012) studied the impact of lean practices on customer

satisfaction by using SEM approach. Figure 15 depicts the research model of this study.

Figure 15: Impact of lean practices on customer satisfaction by using SEM approach (source:

Braunscheidel & Hamister, 2012)

In another study, Monge et al. (2014) compare the SEM approach and multiple regression in

interpreting the study's results of manufacturing and continuous improvement areas while the

sample size was small (n=40). As they stated, the obtained results from the SEM approach are

highly significant. Also, the results of validity and reliability analysis from the SEM are quite

similar with using multiple regression. Therefore, they recommend using the SEM approach

in diversified research fields. Several other authors have also used the SEM approach in their

lean studies and have confirmed the ability of this approach for data analysis in a significant

manner (Moori, Pescarmona, & Kimura, 2013; Russell & Millar, 2014; Todorova, 2013).

Thus, the existing lean literature associated with using SEM, led us to first utilize this ap-

proach in our study. Those literatures helped us to better understand the concepts of SEM ap-

36

proach such as formative or reflective indicators, and we also employed them as a benchmark

for data analysis process.

Secondly, since, there is not sufficient information regarding the relationship between lean

and safety, the SEM approach could help us largely to explore the actual nature of the associ-

ation between lean and safety. As explained before, the proposed model is the first model that

has been developed to predict the relationship between lean and safety, and therefore the cor-

rectness of the model is not ensured. With regard to this issue, SEM is proposed for the analy-

sis of these non-strong theoretical models (Wong, 2013).

Third, in order to generalize the study's results, we needed to conduct an international survey.

It has been shown that the response rate of the surveys in the operations management field is

not well-satisfied (7.47%, (Nahm, Vonderembse, & Koufteros, 2003) and 6.3% , Li et al.

(2005). Thus, the SEM approach could help us: as explained by Wong (2013), SEM is a good

solution to predict the relationships among variables for a small sample size.

Fourthly, due to strict requirements about normal distribution of the data for analysis, SEM

again has been proposed as a productive alternative. In this study, because of time-limitation,

we were not able to distribute the questionnaire on a large scale to obtain the minimum re-

quirements of normal distribution of the data.

In conclusion, these four reasons led us to utilize the SEM approach and its software

SmartPLS for this study.

3.2.3 Survey design and administration

3.2.3.1 Sample design

Since this research attempts to generalize the results of this study, the population is the vari-

ous types of industries that use lean tools across the world. In order to include multiple types

of industries, the North American Industry Classification System (NAICS) was employed,

which is used as a standard for industries' classification. A twenty-category list is provided for

industries in this standard. By having information from different types of lean industries, the

final results will be more reliable to express the impact of lean implementation on OHS per-

formance. Moreover, global distribution of the questionnaire helps in figuring out the effects

of different cultural contexts on considering the impacts of lean implementation on OHS per-

formance. To find common lean practices for investigation, a literature review was conduct-

ed. Several contributions in this field were researched. A popular study by Shad and Ward in

2003 classified the lean practice bundle into four certain sections: total quality management

(TQM), human resource management (HRM), just-in-time (JIT), and total productive mainte-

nance (TPM). In 2007, they defined three underlying constructs for lean practices: supplier

related, customer related, and internally related. According to this category, the internally re-

lated lean practices are related to the objective of this study. That is, the main purpose of this

study is to investigate the impact of lean implementation on workers health and safety inside

of the firm. Consequently, supplier- and customer-related constructs are out of the objective

of this study. Cudney et al. (2013) introduce major lean practices as follows:

5S/visual workplace, quick changeover, mistake proofing, kanban, cell design and one-

piece flow, load leveling (heijunka), kaizen events, standard work, SMED, value stream map-

ping, poka-yoke, and 3P for product and process design.

37

Lastly, Brikie (2016) recently has classified lean practices from extant literature into seven

certain bundles; total quality management, just-in-time, lean purchasing, total productive

maintenance, human resource management, active involvement customer, and supplier col-

laboration and relationship.

Overall, 16 major lean practices, which are all related to the shop floor level, are investigated

for this study.

3.2.3.2 Developing the questionnaire

The first part of the questionnaire includes demographic questions such as age, gender, educa-

tion level, location, and job function. Next, industry size is asked, which is essential due to

study objectives. As previously mentioned, classification of NAICS is employed for this in-

formation. Additionally, in the interest of having information of business size, two metrics;

number of employee and annual revenue are employed. Then, the questions associated with

lean maturity are asked. Fidelity, extensiveness, and experience are the three elements that

form the lean maturity level, as previously described. After that, the main part of question-

naire is provided to discuss the impact of lean implementation on the antecedents of safety

performance. To do this, we asked respondents what effects they have experienced by imple-

menting lean practices. How have antecedents of OHS performance been affected directly

because of lean implementation? The answers include worse, same, and better. Thirty eight

items linked to the antecedents were asked which involve 11 items of working environment, 7

items of workforce characteristics, 10 items of task characteristics, and 10 items of organiza-

tional factors. These items are the main factors linked to antecedents of safety performance.

They have been extracted from relevant literature.

Lastly, the status of OHS performance with respect to lean implementation is questioned.

Four questions were defined to illustrate the status of OHS performance in connection with

lean implementation. The main lagging indicators are utilized in this context, which includes

lost working days, accidents records, and workers compensation. The final version of the

questionnaire is provided in the Appendix.

3.2.3.3 Determination of sample size

According to the published guidelines in SEM-PLS literature, there are several general rules

determining sample size that need to be followed when performing this approach. For in-

stance, Wong (2013) points to the four influencing factors in determining sample size in the

SEM method from the proposal of Hair et al. (2016) as follows:

1-The significance level

2- The statistical power

3-The minimum coefficient of determination (R2 values) used in the model

4-The maximum number of arrows pointing at a latent variable (p. 5)

Generally, for operations management research, the following criteria are pursued.

Significance level= 5%

Statistical power= 80%

R2 values= 0.25

In accordance with these criteria, Marcoulides and Saunders (2013) propose the volume of

sample size through the table below, which depends on the maximum arrows pointed out to a

latent variable in the model.

38

Figure 16: Proposed sample size (source: Wong, 2013)

With respect to this approach, the minimum required sample size for the present study is 110;

the working environment variable has eleven arrows pointing to it.

Additionally, Barclay et al. (1995) suggest the "10-time rule" for determining sample size in

structural equation modeling. That is, sample size should be 10 times bigger than the number

of arrows pointed out to a latent variable anywhere in the SEM-PLS method.

Regarding this suggestion, the minimum required sample size is 110; the working environ-

ment variable has eleven arrows pointing to it.

3.2.3.4 Pilot study

After constructing the questionnaire, it was sent to two professors and two industry experts to

review the items and provide their feedbacks. After receiving the feedbacks, a few changes

were implemented. The main concerns were related to the question wording.

Moreover, Dillman study (2000), which is commonly used in the operations management for

survey distribution, was employed to conduct the pilot study. Initial contacts with the re-

spondents were made for the pilot study. After sending 20 emails with a questioner link to

respondents, 14 responses were collected. Then, initial reliability was performed in order to

conduct the final large-scale study. Respondents from the pilot study were not included into

the final large-scale study. The result of initial reliability of the questionnaire was satisfacto-

ry; Crobnach's alpha= 0.78.

3.2.3.5 Questionnaire sharing

The final version of questionnaire was provided in 18 questions. Because it was needed to

distribute the questionnaire globally, the Qualtrics software package was utilized. This web-

based software has been developed by a private research software company in the United

States in 2002. The Qualtrics software provides an online platform to help researchers collect

data. A number of professional and academic journals has cited the utilization of Qualtrics by

scholars in studies (Albaum & Smith, 2006; Colombo, Bucher, & Sprenger, 2017; D’Mello,

Turkeltaub, & Stoodley, 2017; Strutz, 2008). Therefore, an account was created on its official

website (https://www.qualtrics.com) and questions were transferred into this platform. In or-

der to design the questionnaire, different plans were formulated. To design a user-friendly

questionnaire, all possible methods were taken into account and two academic scholars re-

https://www.qualtrics.com/

39

viewed the design of questionnaire. Finally, it was constructed and its underlying link gener-

ated.

In order to distribute the questionnaire, data was gathered from multiple sources at various

time points from April to July 2017. First, data source that Shah and Ward (2007) used in

their study was employed, which consists of a contact list from Productivity Inc. A firm in-

volved with the consulting, training, and implementing of lean systems. As described on the

website of Productivity Inc" Productivity Inc. has worked with the Global 1000 for more than

35 years. We pioneered the implementation of lean and TPM methodologies in manufacturing

in the late 1970s."(http://www.productivityinc.com/about/). Nowadays, they are working with

not only manufacturing industries, but also healthcare, finance, and other services. For this

research, the data source Productivity Inc. was invaluable because they are in connection with

a set of industries that are at various stages of lean implementation programs. For example, in

the Shah and Ward study, 2616 records were used for their sample. Therefore, this data

source was worthwhile enough for the present study. In addition, we used the social media of

American Society for Quality (ASQ), which is an international group involving individual

and organizational members. In more than 140 countries, people and organizations are in

connection with ASQ, So, it is a worthwhile source for gathering the information. Additional-

ly, social media of the Institute of Industrial & Systems Engineers (IISE) was another valua-

ble source for distributing the questionnaires. To obtain more responses, personal contacts

with scholars across the world were asked to distribute the questionnaire to various industries.

3.2.3.6 Large-scale study

Following the pilot study, we went into the large-scale study of questionnaire testing. Again,

the main steps of conducting this stage were in accordance with Dillman study. By sharing

the questionnaire link and using the data sources, 146 responses were received. According to

the two rules mentioned in the previous sections, we received the required response from re-

spondents for data analysis.

Figure 17: The file extracted from the Qualtric

http://www.productivityinc.com/about/

40

3.2.3.7 Data sorting

After receiving the responses a csv file was extracted from the Qualtrics platform. As Hair et

al. (2016) declare "When the amount of missing data on a questionnaire exceeds 15%, the

observation is typically removed from the data file."(p. 51). We utilized this guideline in the

interest of sorting data and, therefore, the observations which contained greater than 15%

missing data were excluded from the final list. Subsequently, 112 cases were included in the

final list, and therefore the analysis was conducted to these cases. Figure 17 shows the final

file extracted from Qualtrics.

Table 3 presents descriptive summary of the dataset. The classification of the business size in

table 3 is based on the European and American standards, which has been divided into two

main categories: small and medium sized enterprises (SME) and large enterprises. Further,

the business sector is classified into two main categories: manufacturing and services indus-

tries.

Table 3: Descriptive analysis

Frequency Education level Frequency Sex Frequency Age

5 High School 88 Male 6 18-24

34 BS 24 Female 27 25-34

56 MS 112 Total 30 35-44

17 PhD 27 45-54

112 Total 16 55-64

6 +65

112 Total

Frequency Education level Frequency Sex Frequency Business size

Small and Medium

Frequency Business sector Frequency Large

53 Manufacturing 57 Total

59 Services 55

112 Total 112 Country

Europe

Frequency Job category Frequency Asia

39 Health and Safety 28 Australia

19 Engineering 22 South America

18 Operation / Produc-

tion

8 North America

12 Human resources 12 Africa

8 Consulting 29 Total

9 Education 13

7 Other 112

112 Total

3.2.3.8 Data encoding

According to the study objectives, latent variables in the model include lean maturity, work-

ing environment, task characteristics, workforce characteristics, organizational factors, and

OHS performance. The indicators of each latent construct are referring on the concepts intro-

duced in Chapter 1. Accordingly, they are shown in Table 4.

41

Table 4: Latent constructs and corresponding reflective and formative indicators

Latent constructs Indicators Lean maturity Reflective indicators

1. Fidelity (Fid) 2. Extensiveness (Ext) 3. Experience (Exp)

Working environment

Formative indicators 1. Awkward/strained positions (WE_1) 2. Exposure to biological hazards (WE_2) 3. Exposure to dust and/or smoke (WE_3) 4. Exposure to flammable explosive chemicals (WE_4) 5. Exposure to poisonous chemicals (WE_5) 6. Exposure to vibration (WE_6) 7. Exposure to workplace noise (WE_7) 8. Extensive and frequent force (WE_8) 9. Frequent lifting (WE_9) 10. Repetitive motion (WE_10) 11. Status of workplace illumination/lighting (WE_11)

Task characteristics

Formative indicators 1. Breaks (TC_1) 2. Job autonomy (TC_2) 3. Job safety (TC_3) 4. Job satisfaction (TC_4) 5. Job stress (TC_5) 6. Machinery and tool safety (TC_6) 7. Time pressure (e.g. deadlines) (TC_7) 8. Work intensity (e.g. cognitive demands) (TC_8) 9. Workload and pressure (TC_9) 10. Work pace (TC_10)

Workforce characteristics

Formative indicators 1. Skills utilization (WC_1) 2. Risk-taking behavior (WC_2) 3. Motivation for safe working (WC_3) 4. Knowledge about safety issues (WC_4) 5. Defined/clear job functions (WC_5) 6. Employee involvement (overall) (WC_6) 7. Employee involvement in creating a safe environment (WC_7)

Organizational factors

Formative indicators 1. Employee involvement in improving work methods (OF_1) 2. Labor management (OF_2) 3. Management commitment to safety issues (OF_3) 4. Organization's policies on safety issues (OF_4) 5. Reward systems for safety (OF_5) 6. Safety culture (OF_6) 7. Safety systems (e.g. lock-out, tag-out) (OF_7) 8. Teamwork and communication (OF_8) 9. Training on safety and health principles (OF_9) 10. Workplace health promotion programs (OF_10)

OHS performance

Reflective indicators 1. Recordable injuries (OHS_1) 2. Worker’s compensation cost (OHS_2) 3. Accident records (OHS_3) 4. Total lost working days (OHS_4)

In order to encode the latent variables in the models, the Likert scale was employed. For the

16 questions related to the fidelity item, the following values were utilized.

42

If the company implements lean practices to a great extent, the value of 4, for somewhat im-

plementation the value of 3, for very little implementation the value of 2, for not at all imple-

mentation the value of 1 and lastly the value of 0 for the do not know were used.

The extensiveness coding, this is related to how wide lean practices have been implemented

in the organization, is as follows:

All departments: 3

Some departments: 2

No department: 1

Do not know: 0

To encode the experience indicator, this is concerning how long lean practices have been im-

plemented in the organization, the following values are used:

More than 5 years: 4

Between 2 to 5 years: 3

Less than 2 years: 2

Never: 1

Do not know: 0

In order to encode the antecedents (i.e. Working environment, workforce characteristics, task

characteristics, and organizational factors), the following values are used.

If the organization has experienced a worse situation following the lean implementation, the

code of 1 is used and the code of 2 and 3 are used in same and better situations respectively.

More, for the "do not know", the code of 0 is used.

Since 11 items are related to the working environment, the abbreviation of "WE" numbering

from 1 to 11 is used to define the working environment variable. Further, 7 items are related

to workforce characteristics; therefore, the "WC" numbering from 1 to 7 is used to it. Also, 10

items construct task characteristics variable and the "TC" numbering from 1 to 10 are used to

it. Lastly, 10 items construct organizational factors and the "OF" numbering from 1 to 10 are

used to specify them.

To encode the OHS performance, three conditions are utilized. If the OHS performance expe-

riences increased level of accident and injury following the lean implementation, the code of

1 is used and the code of 2 and 3 are used to specify the stable and decreasing level respec-

tively. Further, with regard to four items forming the OHS performance construct, the abbre-

viation of "OHS" numbering from 1 to 4 is used.

More, the size and sector of the industry are two separate variables that are analyzed in this

study, which have moderating effects between lean and safety. In order to encoding the size

of industry, the recommendation of Europe union (European Union, 2003) and American

standards have been utilized. The followings are the descriptions:

Under 500 SME

More than 500 Large

First, the number of employees was checked, if they were less than 500 employees, catego-

rized as SME. And, more than 500 employees categorized as large. Second, the annual reve-

nue was checked, if they were less than 50 million dollars, categorized as SME, otherwise

Large.

Under 50 million dollars: SME

More than 50 million dollars: Large

43

Finally, the two columns were compared. They were the same, meaning that the number of

employees was consistent with the amount of annual revenue. After combining the two col-

umns, one column was set for the size of companies. Subsequently, the code of 1 specifies

small and medium industries and the code of 2 specifies large industries.

To encode the business sector, two common categories were utilized: manufacturing and ser-

vices. The code of 1 was designated for manufacturing industries and the code of 2 for ser-

vices industries.

3.2.3.9 Handling missing data

After coding the variables, in the interest of managing missing data, they were transferred in-

to SPSS 15.In order to manage the missing data, the expectation maximization (EM) method

was utilized to replace the missing data in the SPSS. To do so, the little's missing completely

at random (MCAR) test is needed to be conducted for all variables to ensure that missing data

are in a random manner. According to the definition of this test, if the EM means is not statis-

tically significant, the data are probably missing in random. Then, by having failed to reject

the null hypothesis, it is a good opportunity to do some imputation techniques to replace

missing values to complete the data set

3.2.3.10 Quality checks of results

The next step to prepare the data is the computation of the degree of internal consistency

among questionnaire items. This process should be conducted for all variables. In the

Minitab, the assessment of Cronbach's alpha is utilized to explain the degree of internal con-

sistency. By having run the item analysis under multivariate item, the overall Cronbach's al-

pha and each item's Cronbach's alpha are determined. The analysis of Cronbach's alpha indi-

cates what items are relevant. Those with Cronbach's alpha equal to or greater than 0.70 are

relevant; otherwise, they are omitted from the final list of items.

The computation of Cronbach's alpha was performed for all variables. All the underlying

items of all variables were relevant and therefore all retained in the model for further analysis.

After finalizing the data we additionally checked the data distribution. Although the PLS-

SEM is a nonparametric method, it has been suggested to verify the normality of data to en-

sure that data are not too far from normal, because non-normal data cause some problems

during data analysis as Hair et al. (2016) write "extremely non-normal data inflate standard

errors obtained from bootstrapping and thus decrease the likelihood some relationships will

be assessed as significant" (p. 54). In the interest of examining the data normality two

measures have been proposed by Hair et al. (2016): skewness and kurtosis. The former exam-

ines the symmetry of the variable's distribution and the latter examines the tailedness of the

variable's distribution. The acceptance rate of both kurtosis and skewness is within the -1 and

+ 1. Thus, the distributions outside this range take into account as non- normal data. In this

regard, the skewness and kurtosis were examined in the current study. The results indicate

that all variable exhibit an acceptable degree of normality (within -1, +1).

3.2.3.11 Data entry

The next step of data analysis is calculating the mean value for three underlying indicators of

lean maturity; fidelity, extensiveness and experience. Since 16 items were linked to those var-

44

iables, an adjusted value needs to import into final data sheet in the SmartPLS. Thus, the

mean value was computed for the three variables in each case (company).

Other variables in the model (i.e., working environment, workforce characteristics, task char-

acteristics, organizational factors, and OHS performance) have their own indicators separate-

ly. Therefore, they were prepared to form the final values without any changes to them.

Lastly, the resultant excel file was converted to a .cvs file format and imported into

SmartPLS.

3.3 Analysis procedures

3.3.1 Building the inner model

In regards to conceptual framework that was illustrated in chapter two, first, it needs to build

the inner (structural model). The structural model displays the constructs and their interrela-

tionships. In this regard, Table 4 illustrates the constructs and their indicators.

In order to estimate the PLS model, first, the data should be transferred from the question-

naire into Microsoft Excel. All data were transferred attentively. In the interest of avoiding

errors, two colleagues also reviewed the data-transfer process. Indicators are placed in the

first row of the Excel file. Then, each row contains an individual response from cases.

3.3.2 Building the outer model

The next step to analyze the data is building the outer model. To do this, indicators should be

linked to latent variables.

As stated by Wong (2016), in order to reduce the model complexity, make the theoretical

model more parsimony, and eliminate discriminant validity, hierarchical component model

(HCM) can be designed in the PLS-SEM. The HCM includes two underlying components:

while the first called observable lower-order components (LOCs), the second refers to unob-

servable higher-order components (HOCs). In this study, lean maturity is a higher-order con-

struct. It is identified by evaluating three underlying indicators. It means lean maturity holds a

reflective relationship with its lower-order components (fidelity, extensiveness, and experi-

ence). Therefore, the three underlying items of lean maturity (fidelity, extensiveness, and ex-

perience) are deployed again for the lean maturity naming Lean_1, Lean_2, and Lean_3.

By having all indicators defined for each latent variable, the indicators are linked to each la-

tent variable in the model. The colors of latent variable, now, change from red to blue.

3.3.3 Formative and reflective measurement

The Figure 18 captured from Hair et al. (2016) was utilized as the main guideline for deter-

mining formative and reflective measurement in the model. In this regard, the underlying var-

iables linked to lean maturity designated as reflective type. Moreover, the previous studies

confirm the nature of three underlying indicators as reflective measurement for the lean ma-

turity.

45

Figure 18: The guideline for choosing the measurement model mode

(source: Hair et al., 2016) Due to this fact that these three indicators are highly correlated with each other, the causality

direction goes from lean maturity to these indicators. In similar fashion the indicators associ-

ated with the OHS performance are assumed as reflective measurement.

On the other side, the underlying indicators for other variables in the model (working envi-

ronment, workforce characteristics, task characteristics, and organizational factors) are mod-

eled as formative measurement because their indicators cause the latent variables and previ-

ous studies also confirm the nature of formative measurement for the antecedent variables

Overall, Table 4 in proceeding section shows latent constructs and their reflective or forma-

tive indicators.

3.3.4 Running the path-modeling estimation

In order to systematically evaluate the results of PLS-SEM, the guidelines from Hair et al.

(2016) are utilized. Figure 19 portrays the steps that should be followed for the data analysis.

Figure 19: The systematic evaluation of the PLS-SEM results (source: Hair et al., 2016)

46

As seen in Figure, prior to examine the structural model the reliability and validity of con-

structs should be established. If the reliability and validity of the constructs are acceptable,

then the estimates of the structural model will be undergone.

3.3.4.1 Assessment of the reflective measurement models

In order to appropriately assess the reflective measurement model, firstly, the internal con-

sistency reliability and validity should be examined. To examine the internal consistency reli-

ability, the evaluation of composite reliability is utilized and to examine the validity, the con-

vergent validity and discriminant validity are checked. In the SmartPLS the average variance

extracted (AVE) is performed to evaluate the convergent validity and the Fornell-Locker

standard and cross loading are utilized to evaluate the discriminant validity. These criteria are

not applicable to single-item constructs (Hair et al., 2016). In this regard, the reliability and

validity for single-item construct is measured based on the various forms of validity assess-

ment. In the following sections these criteria are addressed.

3.3.4.1.2 Internal consistency reliability

Traditionally, the Cronbach's alpha is the criterion of the internal consistency reliability. This

criterion based on the correlations between indicators estimates the reliability (Hair et al.,

2016). With regard to the limitations of Cronbach's alpha, the composite reliability, as a re-

placement, is utilized to measure the internal consistency reliability. The value of the compo-

site reliability is categorized into three grades: between 0.60 and 0.90 is satisfactory, below

0.60 indicating lack of reliability and above 0.90 is not desirable.

3.3.4.1.3 Convergent validity

Hair et al. (2016) explain the convergent validity as "the extent to which a measure correlates

positively with alternative measures of the same construct" which is measured through the

outer loading of the indicators and the average variance extracted (AVE). The acceptable val-

ue for the outer loadings is 0.708 or higher which indicates the underlying indicators have

much in common on a construct. On the other side, the satisfactory value for the AVE is 0.50

or higher, which indicates that the latent variable explains more than half of its indicators'

variance. It is important to note that the AVE is not applicable for single-item construct in the

model since the outer loading of the indicators is fixed at 1.00.

3.3.4.1.4 Discriminant validity

Hair et al. (2016) explain the discriminant validity as" is the extent to which a construct is tru-

ly distinct from other constructs by empirical standards" which indicates that a latent variable

is unique and not represented by other latent variables in the model. The cross loading of the

indicators is used to assess the discriminant validity. The outer loading of indicators linked to

a construct should be higher than the cross loadings on other constructs. The second tech-

nique that assesses the discriminant validity is the Fornell-Larcker criterion. For this criterion,

as stated by Hair et al. (2016), "the square root of the AVE of each construct should be higher

than its highest correlation with any other construct" (p. 107).

We go to calculate button on the right side of the main page of the SmartPLS environment

and select the PLS algorithm. By selecting a path weighting scheme, maximum iterations at

300, stop criterion at 10E-7, and initial weights by 1, it is prepared to start the calculation. Be-

47

fore analyzing the results, we should check the algorithm convergence. The number of itera-

tions should be lower than the maximum iterations, which is 300 in this case.

The existing model has two latent variables with reflective measurement models (i.e., lean

maturity and OHS performance) and three single-item constructs (i.e., fidelity, extensiveness,

experience). Therefore, we need to estimate the relationships between reflective constructs

and their indicators.

3.3.4.2 Assessment of the formative measurement models

Unlike the reflective measurement, the internal consistency approach is not applicable for the

formative measurements since the formative models do not play the role of predictor in the

model. Therefore, different techniques are employed to assess the quality of the formative

measurements. Figure 20 outlines the assessment procedure to formative measurement mod-

els, captured from Hair et al. (2016).

Figure 20: The formative measurement models assessment (source: Hair et al., 2016)

Considering the convergent validity of the formative constructs ensures that the formative

construct and its relevant facets are covered correctly with the selected indicators. The collin-

earity technique assesses the relationships between formative indicators and the contributions

of indicators to constructs are assessed by examining the indicators' significance and rele-

vance. The redundancy analysis is used to assess the convergent validity, whereas the vari-

ance inflation factor (VIF) is utilized to measure the collinearity and t value is calculated to

assess the contributions of each indicator in formative constructs.

Since we did not provide a question representing the global measure for each formative con-

struct in the original questionnaire, it is not applicable to perform the redundancy analysis to

show the convergent validity. Therefore, other evaluations related to formative construct are

performed.

In order to check the collinearity of indicators within formative constructs, the results of PLS

algorithm are employed. The collinearity statistics (VIF) in the quality criteria section is em-

ployed.

The next step is related to check the outer weights of formative indicators to ensure their sig-

nificance and relevance. First, the significance of outer weights is performed by bootstrapping

48

routine. The number of 5000 (suggested by the software itself) was selected as subsamples in

bootstrapping process. Moreover, the two-tailed test and the significance level of 0.05 were

selected for the analysis. After running the procedure, t values are provided for both the

measurement model and structural model.

3.3.4.3 Evaluation of structural model

The assessment of the structural model is performed to decide whether empirical data support

the hypotheses and to make a decision on the empirical confirmation of the theory. In this re-

gard, three steps are followed: First, assessing the path coefficients and R2 values, second,

reviewing the goodness-of-fit criterion, third, addressing the heterogeneity issue in the esti-

mating path model. Figure 21, captured from Hari et al. (2016), shows the systematic ap-

proach to assess the structural model in the existing study.

Figure 21: The structural model assessment procedure (source: Hair et al., 2016)

To assess the collinearity issues, the tolerance value and VIF criteria are utilized like forma-

tive indicators in the previous section. In step two, to assess the significance and relevance of

the relationships between latent variable, the path coefficients extracted by the PLS-SEM al-

gorithm is employed. The path coefficients close to +1 show a strong positive relationship

between the variables, whereas the adjacent value to -1 shows negative relationships. Moreo-

ver, the path coefficient values close to 0 indicate a weak relationship between variables.

Lastly, to determine the significance of the path coefficients, the standard error obtained by

bootstrapping means is utilized. By comparing the empirical t value to critical t value, the ul-

timate decision on the significance of path coefficient is undertaken. Additionally, the p val-

ue reported by bootstrapping means can be employed to make decision on the significance of

path coefficients. In step three, the coefficient of determination (R2) is assessed through the

PLS algorithm. Hair et al. (2016) explain this coefficient as "a measure of the model's predic-

tive accuracy and is calculated as the squared correlation between a specific endogenous con-

struct's actual and predicted values. The coefficient represents the exogenous latent variables'

combined effects on the endogenous latent variable"(p. 174). The R2 values of 0.75, 0.50, and

0.25 are specified respectively, for substantial, moderate, and weak for levels of predictive

accuracy in endogenous constructs. In step four, the f2 effect size is calculated through the

49

PLS algorithm to evaluate the effect of each exogenous construct on endogenous construct.

The values of 0.02, 0.15, and 0.35 are the guidelines for assessing the effect size of exoge-

nous constructs on endogenous constructs as small, medium and large respectively.

3.3.4.4 Importance-performance matrix analysis

In the SmartPLS to explain the importance and performance of each construct on other con-

structs the importance-performance matrix analysis (IPMA) is employed. The results of this

analysis are significant for improving the managerial decisions. The importance of each vari-

able is shown through total effect to target construct and the performance of each variable is

shown through average construct score.

3.3.4.5 Mediation analysis

The analysis of the PLS model is not always straightforward. Sometimes it is needed to eval-

uate the effects of mediators in the model. According to the definition described by Reuben

and Kenny (1986) mediators account for the relation between independent and dependent var-

iables. The observed variations in the mediators are the result of variations in the independent

variable, which finally cause variations in the dependent variable. In regards to the above def-

inition, in this study, the antecedents of OHS performance play the role of mediators in the

model, because, they typically mediate the relation between lean maturity and OHS perfor-

mance. The variations in the level of lean maturity cause the variations in the antecedents, and

finally these variations reflect on the OHS performance. Therefore, to measure the effects of

antecedents on the relationship between lean maturity and OHS performance, the Preacher

and Hayes procedure (2008) was followed. Figure 22 illustrates an example of mediation

analysis of the antecedents.

Figure 22: Mediation analysis

In regards to the procedure by Preacher and Hayes, two steps are followed in the Smart PLS

by using the bootstrapping technique.

First, the significance of direct effect should be evaluated. If the significance of the direct ef-

fect between variables could be established, then, the mediating effect of mediators is possi-

bly measurable. In order to evaluate the direct effect of lean maturity on the OHS perfor-

mance, the bootstrapping procedure is utilized without the presence of antecedents. Moreover

the procedure illustrated in Figure 23, extracted from Hair et al. (2016), was employed to

study the effects of antecedents as the mediators in the model.

50

Figure 23: The mediator analysis (source: Hair et al., 2016)

3.3.4.5.1 Magnitude of mediation

Following mediation analysis, the magnitude of mediation can be examined. According to

Wong (2016), two items are utilized to check the magnitude of mediation in the SmartPLS

platform: the total effect and the variance account for (VAF). The total effect equals direct

effect plus indirect effect. The VAF equals indirect effect divided by total effect. The thresh-

old of 20% has been proposed for the VAF (Hair et al., 2016). When the VAF is higher than

20%, partial mediation is established and when the VAF is higher than 80%, the full media-

tion is achieved.

3.3.4.6 Moderation analysis

With respect to previous studies about lean implementation, the size and sector of organiza-

tions, possibly affect the final impacts of lean implementation. Therefore, in this study, the

size and sector of organizations have been considered as moderators. The effects of size and

sector, possibly show differences in relationships of the model. In order to evaluate the mod-

erators' effects, the multi-group analysis (PLS-MGA) is performed. To do so, a parametric

approach involving two independent-sample t test is utilized to compare the path coefficient

between groups of data. In the SmartPLS, the standard deviation of path coefficient is per-

formed via the bootstrapping procedure. By having defined the standard deviation, the mod-

erating effects of size and sector are explored. By determining the variance of parameters in

the PLS, the differences between categorical moderators are assessed.

The next chapter presents all findings of the above mentioned analysis procedures.

51

CHAPTER4

FINDINGS

This chapter presents the accumulative findings of this research in an interlinked way. To do

so, first the results linked to reflective and formative indicators are presented. Then, the cas-

ual relationships among latent variables are explored. Lastly, mediation effects of OHS ante-

cedents and moderation effects of sector and size are shown respectively.

4.1 Reflective measurement analysis

By running the PLS algorithm, the output result is provided for reflective indicators. Table 5

shows the results of reflective measurement.

Table 5: The outer loadings of the reflective indicators

All outer loadings of lean maturity variable are acceptable; above the threshold level (i.e.,

0.708). The fidelity has the highest outer loading (i.e., 0.891) and its indicator reliability is

0.793 (0.8912) and the extensiveness has the lowest indicator reliability with the value of

0.519 (0.7212). Additionally, four underlying indicators of OHS performance have also the

satisfactory level of outer loadings, including 0.777, 0.775, 0.716, and 0.712 to OHS_1,

OHS_2, OHS_3, and OHS_4 respectively.

The composite reliability for lean maturity and OHS performance are 0.851 and 0.824 respec-

tively, indicating an acceptable level of internal consistency reliability. For the three reflec-

tive indicators (fidelity, extensiveness, and experience) the composite reliability is 1.00 since

they are single-item construct.

In order to evaluate the convergent validity, the AVE value is utilized. In the existing model

the AVE value for the lean maturity is 0.657 and for the OHS performance is 0.541. Since the

required minimum level for the AVE is 0.50, the values for lean maturity and OHS perfor-

mance are above this level and therefore indicating these variables have high levels of con-

vergent validity.

Lastly, to measure the discriminant validity, the cross loadings and the Fornell-Larcker crite-

ria are used. Table 6 shows the matrix of Fornell-Larcker of the model.

Latent constructs Reflective indicators

OHS performance

Recordable injuries

0.777

Compensation costs

0.775

Accident records

0.716

Lost Working days

0.712

Lean maturity

Fidelity

0.891

Extensiveness

0.721

Experience

0.811

52

Table 6: Results of Fornell-Larcker criterion

With respect to Hair et al. (2016) the values are in accordance with the Fornell-Larcker crite-

rion and confirm the discriminant validity for the lean maturity and OHS performance as the

reflective constructs in the model. That is, square root of the AVE of lean maturity is higher

than its correlation with any other construct and additionally the square root of the AVE of

OHS performance is higher than its correlation with any other construct in the model.

In order to establish the discriminant validity by the cross loadings criterion, the loadings of

an indicator linked to a construct should be higher than its cross loadings with other variables.

Table 7 shows the cross loadings of constructs, where reported data confirms the discriminant

validity of the selected constructs. That is, all indicators' outer loadings on the associated con-

struct are greater than all of their loadings on other constructs.

Constructs Lean

maturity

OHS per-

formance

Organizational

factors

Task charac-

teristics

Workforce

characteristics

Working

environment

Lean maturity 0.820

OHS perfor-

mance 0.256 0.780

Organizational

factors 0.385 0.481

Formative

measurement

model

Task character-

istics 0.488 0.333 0.689

Formative

measurement

model

Workforce char-

acteristics 0.410 0.291 0.675 0.708

Formative

measurement

model

Working envi-

ronment 0.482 0.468 0.645 0.758 0.759

Formative

measurement

model

53

Table 7: Results of Cross loadings

Working environment


Task charac-teristics


OHS per-formance

Lean maturity

0.410 0.398 0.420 0.357 0.226 0.880 Fid

0.255 0.241 0.217 0.156 0.152 0.708 Ext

0.474 0.343 0.494 0.375 0.235 0.860 Exp

0.558 0.394 0.472 0.497 0.709 0.242 OHS_1

0.143 0.024 0.025 0.204 0.740 0.091 OHS_2

0.304 0.159 0.166 0.287 0.839 0.197 OHS_3

0.312 0.238 0.222 0.404 0.823 0.217 OHS_4

0.483 0.433 0.441 0.508 0.132 0.271 OF_1

0.495 0.348 0.485 0.559 0.125 0.187 OF_2

0.568 0.621 0.454 0.506 0.196 0.254 OF_3

0.510 0.512 0.427 0.548 0.164 0.298 OF_4

0.403 0.344 0.522 0.557 0.010 0.327 OF_5

0.578 0.656 0.572 0.751 0.347 0.306 OF_6

0.568 0.606 0.704 0.861 0.345 0.418 OF_7

0.182 0.134 0.118 0.178 0.013 0.093 OF_8

0.472 0.620 0.624 0.641 0.269 0.280 OF_9

0.533 0.617 0.572 0.707 0.240 0.308 OF_10

0.566 0.579 0.673 0.429 0.243 0.315 TC_1

0.573 0.556 0.643 0.362 0.190 0.280 TC_2

0.535 0.582 0.666 0.615 0.330 0.214 TC_3

0.421 0.400 0.539 0.322 0.240 0.137 TC_4

0.661 0.510 0.802 0.468 0.250 0.403 TC_5

0.558 0.619 0.637 0.548 0.228 0.285 TC_6

0.361 0.276 0.461 0.290 0.021 0.315 TC_7

0.494 0.540 0.669 0.502 0.208 0.336 TC_8

0.585 0.627 0.723 0.550 0.134 0.425 TC_9

0.601 0.513 0.640 0.533 0.164 0.346 TC_10

0.181 0.287 0.266 0.257 0.007 0.135 WC_1

0.053 0. 320 0.190 0.245 0.116 0.102 WC_2

0.482 0.609 0.552 0.601 0.082 0.317 WC_3

0.564 0.717 0.637 0.502 0.203 0.298 WC_4

0.609 0.798 0.548 0.579 0.276 0.295 WC_5

0.449 0.592 0.507 0.447 0.314 0.270 WC_6

0.549 0.740 0.414 0.483 0.195 0.318 WC_7

0.825 0.674 0.614 0.430 0.375 0.409 WE_1

0.465 0.402 0.363 0.392 0.224 0.219 WE_2

0.737 0.625 0.627 0.488 0.205 0.304 WE_3

54

Working environment


Task charac-teristics


OHS per-formance

Lean maturity

0.626 0.540 0.562 0.500 0.169 0.235 WE_4

0.582 0.526 0.578 0.531 0.236 0.308 WE_5

0.661 0.446 0.598 0.505 0.265 0.269 WE_6

0.626 0.425 0.507 0.407 0.181 0.224 WE_7

0.613 0.565 0.481 0.463 0.344 0.240 WE_8

0.693 0.543 0.428 0.313 0.182 0.192 WE_9

0.706 0.513 0.645 0.490 0.336 0.335 WE_10

0.770 0.677 0.589 0.537 0.250 0.385 WE_11

In summary, according to the reliability results of reflective indicators linked to lean maturity

it discloses that three reflective indicators of lean maturity (i.e. Fidelity, extensiveness, and

experience) are highly related to each other and significantly explain the lean maturity con-

struct. This in support of the Ansari et al. (2010) study that introduces the fidelity and exten-

siveness as the forming items of lean maturity variable and the study of Satoğlu &

Durmuşoğlu (2000) that introduce the experience level as a parameter showing the lean ma-

turity in industries.

Further, the results of OHS performance construct show a high reliability among its reflective

indicators (i.e. number of accidents, number of injuries, number of lost working days, and

compensation cost). This in support of previous studies (Hinze et al., 2013; Qien, Utne, &

Herrera, 2011) that establish OHS lagging indicators to measure the performance of OHS in

the workplace. Table 8 shows the results summary for reflective constructs

Table 8: Results summary for reflective measurement models

Latent con-struct

Indicators Loading Indicator reliability

Composite reliability

AVE Discriminat validity?

Lean maturity

Fid 0.891 0.793

0.851

0.657

Yes

Ext 0.721 0.519

Exp 0.811 0.657

OHS performance

OHS_1 0.777 0.603

0.824

0.541

Yes

OHS_2 0.775 0.600

OHS_3 0.716 0.512

OHS_4 0.712 0.506

4. 2 Analysis of formative measurements

Firstly, the colliniarity of formative indicators is assessed. The variance inflation factor (VIF)

is a related measure of collinearity. Table 9 shows the results of VIF for formative constructs,

including working environment, workforce characteristics, task characteristics and organiza-

tional factors. A VIF value of 5 or higher indicates a collinearity problem within formative

indicators (Hair et al., 2016).

55

Table 9: Results of VIF for formative indicators

Formative

Indicators

VIF

OF_1 2.560

OF_2 1.703

OF_3 2.039

OF_4 2.576

OF_5 2.506

OF_6 2.604

OF_7 2.307

OF_8 1.498

OF_9 2.432

OF_10 3.686

TC_1 1.890

TC_2 1.627

TC_3 1.708

TC_4 3.102

TC_5 3.308

TC_6 1.537

TC_7 1.796

TC_8 3.345

TC_9 3.236

TC_10 2.976

WC_1 1.692

WC_2 1.740

WC_3 1.965

WC_4 1.651

WC_5 2.745

WC_6 1.598

WC_7 1.334

WE_1 2.310

WE_2 2.326

WE_3 3.216

WE_4 2.578

WE_5 3.114

WE_6 3.604

WE_7 2.667

WE_8 2.426

WE_9 1.991

WE_10 1.781

WE_11 1.494

As seen, all formative indicators are below the maximum value of 5 indicating the collinearity

is not an issue for formative measurements assessment.

The next step is related to check the outer weights of formative indicators to ensure their sig-

nificance and relevance. Table 10 shows the results of outer weights of formative indicators.

In order to establish the significance of outer weights of formative indicators, the value of 5%

(α = 0.05) and its probability of error 1.96 were chosen for this analysis. Regarding these re-

sults, we retain the formative indicators of each construct in the model. For seven indicators

56

that were not significant (NS), the procedure shown in Figure 24 was utilized. The final re-

sults confirm their significance.

Table 10: The outer weights of formative indicators

Formative constructs

Formative indicators

Outer weights (outer loadings)

t Value Significance level

P Value Confidence in-tervals

WE

WE_1 0.133 (0.440) 2.631 *** 0.009 [-0.168, 0.506] WE_2 0.170 (0.269) 1.105 NS 0.269 [-0.124, 0.473] WE_3 0.099 (0.541) 1.70 * 0.092 [-0.254, 0.399] WE_4 -0.068 (-0.658) 2.014 ** 0.046 [-0.372, 0.253] WE_5 0.039 ( 0.797) 2.051 ** 0.042 [-0.265, 0.348] WE_6 0.142 (0.298) 1.732 * 0.089 [-0.135, 0.416] WE_7 -0.097 (-0.495) 1.981 ** 0.050 [-0.370, 0.193] WE_8 0.259 (0.088) 1.711 * 0.087 [0.105, 0.495] WE_9 0.089 (0.573) 0.600 NS 0.548 [-0.216, 0.373] WE_10 0.327 (0.029) 2.249 ** 0.025 [-0.004, 0.556] WE_11 0.403 (0.003) 2.965 *** 0.003 [0.117, 0.650]

WC

WC_1 0.067 (0.061) 2.292 *** 0.024 [-0.473, 0.591] WC_2 -0.283 (-0.170) 1.991 ** 0.049 [-0.747, 0.598] WC_3 -0.045 (-0.012) 0.158 NS 0.874 [-0.583, 0.582] WC_4 0.295 (0.247) 1.832 * 0.069 [-0.425, 0.732] WC_5 -0.288 (-0.230) 1.661 * 0.438 [-0.876, 0.663] WC_6 0.461 (0.363) 2.632 *** 0.009 [-0.532, 0.793] WC_7 0.835 (0.614) 1.790 * 0.074 [-0.815, 0.993]

TC

TC_1 0.339 (0.321) 2.170 * 0.030 [0.011, 0.614] TC_2 0.133 (0.123) 1.642 * 0.103 [-0.188,0.424] TC_3 0.337 (0.358) 1.899 * 0.058 [-0.038, 0.705] TC_4 0.099 (0.124) 2.71 *** 0.007 [-0.236, 0.375] TC_5 0.056 (0.037) 0.196 NS 0.844 [-0.316, 0.427] TC_6 0.298 (0.315) 2.162 ** 0.031 [-0.007, 0.557] TC_7 0.043 (0.029) 0.148 NS 0.883 [-0.331, 0.446] TC_8 0.096 (0.112) 2.014 ** 0.046 [-0.253, 0.395] TC_9 -0.027 (-0.32) 2.921 *** 0.004 [-0.369, 0.281] TC_10 0.087 (0.066) 1.893 * 0.612 [-0.290, 0.397]

OF

OF_1 0.360 ( 0.339) 1.821 * 0.071 [-0.057, 0.712] OF_2 -0.130 (-0.143) 2.012 ** 0.046 [-0.507, 0.214] OF_3 0.040 (0.047) 2.613 *** 0.010 [-0.228, 0.311] OF_4 -0.002 (-0.001) 0.011 NS 0.991 [-0.373, 0.317] OF_5 0.055 (0.038) 1.652 * 0.101 [-0.345, 0.381] OF_6 0.132 (0.173) 1.729 * 0.086 [-0.177, 0.578] OF_7 0.258 (0.261) 1.251 * 0.213 [-0.157, 0.652] OF_8 0.111 (0.114) 0.696 NS 0.486 [-0.198, 0.444] OF_9 0.403 (0.363) 2.187 ** 0.031 [-0.010, 0.703] OF_10 0.195 (0.150) 1.694 * 0.093 [-0.252, 0.497]

Note: NS = not significant. a. Bootstrap confidence intervals for 5% probability of error (α= 0.05). *p < .10. **p < .05. ***p < .01.

In summary, the validity of all formative indicators linked to antecedents is in an acceptable

range. Further, the results of t value show that formative indicators in the model significantly

contribute to their constructs (i.e. four antecedents). These findings support the results of the

literature review conducted in the current study (Paper A) to identify the formative indicators

of each antecedent. Regarding these results we retained all the formative indicators of each

antecedent in the model.

57

Figure 24: The bootstrap sign change options (source: Hair et al., 2016)

4.3 Structural model evaluation

First, by running the PLS algorithm the results of VIF are presented. Table 11 shows the VIF

results of predictors in two sets. The first set is in connection with the lean maturity as the

predictor for OHS performance and the four antecedents of OHS performance. The second set

is in connection with the antecedents (working environment, workforce characteristics, task

characteristics, and organizational factors) as the predictors of OHS performance.

Table 11: Collinearity assessment of latent constructs

As seen from the table above, the collinearity is not an issue among predictors constructs in

the structural model.

On the other hand, the R2 is the most commonly used measure to evaluate the relationship

among constructs in a model. That is, Hair et al. (2016) explain R2 as "a measure of the mod-

el's predictive accuracy and is calculated as the squared correlation between a specific endog-

enous construct's actual and predicted values. The coefficient represents the exogenous latent

First set Second set

Constructs VIF Constructs VIF

Lean maturity

1.57 OHS performance

Working environ-ment


1.00 Working envi-

ronment

Workforce character-istics


1.00 Workforce char-

acteristics Task characteristics


1.00 Task characteris-

tics

Organizational fac-tors


1.00 Organizational

factors

58

variables' combined effects on the endogenous latent variable"(p. 174). Accordingly, R2 crite-

rion shows how much variance of the latent variable is being explained by the latent varia-

bles. Table 12 depicts the amount of variance of endogenous constructs - i.e. Working envi-

ronment, Task characteristics, Workforce characteristics, Organizational factors, and OHS

performance - that is explained by the lean maturity as exogenous variable linked to them.

Table 12: R2 evaluation of the endogenous variables

Among antecedents, the working environment has the highest R2 value (i.e., 0.321) and the

organizational factors has the lowest value (i.e., 0.298) .The R2 value of the direct influence

of lean maturity on OHS performance is 0.227.

Next, the total effect of lean maturity on target constructs is evaluated. As shown in Table 13,

the total effect of lean maturity on OHS performance is 0.304 while the total effects of lean

maturity on the antecedents are higher (i.e., 0.567, 0.546, 0.528, 0.423). Additionally, the

total effects of antecedents on OHS performance are shown in this table.

Table 13: Results of total effetcs among constructs

Constructs Lean

maturity

OHS

perfor-

mance

Organiza-

tional

factors

Task

character-

istics

Workforce

Characteristics

Working

environment

Lean maturity 0.304 0.423 0.567 0.528 0.546

OHS

performance

Organizational

factors

0.460

Task characteristics 0.188

Workforce

Characteristics

0.298

Working

environment

0.520

As seen, lean maturity has the strongest influence on antecedents compared to OHS perfor-

mance indicating the importance of antecedents for taking into account in measuring the im-

pact of lean implementation on OHS performance.

4. 4 Importance-performance matrix analysis (IPMA)

The IPMA is utilized to explain the importance and performance of each construct in the

model on the target construct (OHS performance). Table 14 depicts the results.

R Square Constructs

0.321 Working environment

0.279 Task characteristics

0.271 Workforce characteristics

0.298 Organizational factors


59

Table 14: Index values and total effects for the IPMA of OHS performance

Constructs Importance (Total effects) Performance

(Index values)

Lean maturity 0.30 45%

Working environment 0.52 62%

Workforce characteristics 0.29 43%

Task characteristics 0.18 37%

Organizational factors 0.46 58%

The results show that among antecedents working environment is the main construct to estab-

lish OHS performance. This issue should be taken into account in managerial decision.

4.5 Mediation effect analysis

As previously mentioned, the relationship between lean maturity and OHS performance is

significant (p <0.05), therefore, it is possible studying the meditating effects of antecedents.

The mediation variables (antecedents) were then included in the PSL bootstrapping procedure

to analyze whether the indirect effect of lean maturity on OHS performance via the anteced-

ents is significant as well. The path coefficient between lean maturity and the working envi-

ronment is found to be 0.56, between working environment and OHS performance is 0.34;

therefore, the indirect effect of lean maturity on OHS performance via working environment

is 0.19 (i.e. 0.56*0.34). Table 15 shows the results of all the indirect effects.

Table 15: The results of indirect effects

Direct effect

Lean Antecedents

Direct effect

Antecedents OHS

Indirect-path effect

Lean OHS

Lean →WE 0.56 WE→OHS 0.34 0.19

Lean →WC 0.42 WC→OHS 0.24 0.10

Lean →TC 0.52 TC→OHS 0.26 0.13

Lean →OF 0.54 OF→OHS 0.27 0.14

Thus, the total effects are captured (direct effects + indirect effects) as shown in Table 16.

Finally, the VAF is computed (it equals the indirect effects divided by the total effects).

Table 16: The mediating effects of antecedents

Indirect effect

Direct effect Lean→ OHS

Total effect VAF

Lean→ WE 0.19 0.30 0.49 38 %

Lean→ WC 0.10 0.30 0.37 27 %

Lean→ TC 0.13 0.30 0.40 32 %

Lean→ OF 0.14 0.30 0.36 38 %

WE: Working environment, WC: Workforce characteristics, TC: Task characteristics, OF: Organizational fac-

tors, Lean: Lean maturity, VAF: Variance account for = Indirect effect/Total effect

4.7 Moderation analysis

To study the moderating effect of different contextual variables, i.e. size of the organization

and its sector, the corresponding data was separately analyzed. Table 17 and 18 show the re-

sults of analyses.

60

Table 17: The moderating effect of sector variable over direct relationships in the model

Sector1:

Manufacturing

Sector2:

Services

Manufacturing vs. Services

p(1)

se(p(1))

p(2)

se(p(2) )

|p(1)-p(2) |

t value

Significance

level

p value

Lean→Fid 0.884 0.029 0.881 0.037 0.003 0.064 0.949

Lean→Ext 0.610 0.166 0.755 0.065 0.145 0.853 0.395

Lean→Exp 0.862 0.027 0.727 0.088 0.135 1.479 0.144

Lean→WE 0.651 0.072 0.443 0.122 0.208 1.481 0.142

Lean→WC 0.469 0.419 0.340 0.347 0.129 0.241 0.810

Lean→TC 0.730 0.061 0.583 0.153 0.147 1.125 0.370

Lean→OF 0.634 0.071 0.469 0.107 0.165 1.296 0.198

Lean→OHS 0.246 0.281 -0.088 0.172 0.158 0.495 0.622

Note: p (l) and p (2) are path coefficients of sector 1 and sector 2, respectively; se (p (1)) and se (p (2)) are the stand-ard error of p (l) and p (2), respectively.

Table 18: The moderating effect of size variable over direct relationships in the model

Size1: Small and Medium

Size 2:

large

Small and medium vs. large

p(1)

se(p(1))

p(2)

se(p(2) )

|p(1)-p(2) |

t value

Significance

level

p value

Lean→Fid 0.852 0.047 0.914 0.022 0.062 1.191 0.236

Lean→Ext 0.740 0.078 0.736 0.101 0.004 0.032 0.975

Lean→Exp 0.760 0.070 0.830 0.033 0.07 0.902 0.369

Lean→WE 0.490 0.132 0.612 0.070 0.122 0.815 0.417

Lean→WC -0.160 0.311 0.415 0.312 0.575 0.584 0.560

Lean→TC 0.488 0.124 0.609 0.081 0.121 0.824 0.412

Lean→OF 0.471 0.113 0.597 0.068 0.126 0.956 0.341

Lean→OHS -0.106 0.171 0.150 0.161 0.256 0.189 0.851

Note: p(l) and p(2) are path coefficients of size 1 and size 2, respectively; se(p(1) ) and se(p(2) ) are the standard error of p(l) and p(2) , respectively.

As seen in tables 17 and 18, based on the results of p values, the moderating effects of sector and size variables are not supported in this study.

61

CHAPTER5

DISCUSSION The discussion of the main results will be remarked to conclude with some propositions sum-

marizing the main theoretical contributions of the thesis.

5.1 Discussion of model hypotheses

This study strives to empirically examine the theory of using OHS leading indicators to

measure the impact of lean implementation on OHS performance. Therefore, 11 hypotheses

were formulated to be empirically tested. The hypotheses cover:

The direct effects of lean implementation on OHS performance (H1);

The direct effects of lean implementation on antecedents of OHS performance (H2, H3,

H4, and H5);

The mediating effects of antecedents between lean implementation and OHS performance

(H6, H7, H8, and H9); and

The moderating influence of business size and sector on the relationships between lean

implementation, antecedents and OHS performance (H10, H11).

Previous studies have revealed a significant association between lean implementation and

OHS performance. While some studies report positive effects of lean implementation on OHS

performance (Womack, Jones, & Roos, 1990), other show negative effects of lean implemen-

tation on OHS performance (Conti & Angelis, 2006; Hallowell, Veltri, & Johnson, 2009).

The first hypothesis in the current study was also formulated to empirically examine this sig-

nificance. Our results support the previous studies that report both positive and negative ef-

fects of lean implementation on OHS performance. We found that lean maturity significantly

predicts OHS performance and there is not a collinearity issue in this context (i.e. VIF=1.57).

The result of path coefficient between lean maturity and OHS performance reveals a signifi-

cant relationship between these two variables in support to hypothesis H1. Overall, the find-

ings related to the first hypothesis confirm and provide a deeper understanding to the signifi-

cant relationship between lean implementation and OHS performance.

The second part of the current study related to the investigation of the mechanisms of influ-

ence, i.e. the direct effects of lean implementation on OHS antecedents. To study the signifi-

cance of path coefficients between lean maturity and OHS antecedents in the structural mod-

el, the bootstrapping routine was employed. The findings show that lean maturity significant-

ly influences OHS antecedents: the path coefficients are 0.567, 0.423, 0.528, and 0.546 for

working environment, workforce characteristics, task characteristics, and organizational fac-

tors respectively. As seen, lean maturity has the strongest influence on the working environ-

ment (0.567), whereas the workforce characteristics receives the lowest effect (0.423). When

these results are compared to the path coefficient between lean maturity and OHS perfor-

mance directly (0.304), the importance of OHS antecedents becomes even more apparent.

This is in support of the four hypotheses (H2 to H5).

62

To evaluate the significance of the combined effects of lean maturity and antecedents, the R2

values were analyzed through the PLS-SEM algorithm. The findings show that the working

environment has the highest value among the antecedents (0.321) and workforce characteris-

tics, with the value of 0.271, has the lowest significance.

Overall, the results of path coefficients, R2 level, and total effects between lean maturity and

OHS antecedents all confirm the formulated theory of the present study. Results are also con-

sistent with findings of several studies that present the positive impact of lean implementation

on working conditions (Saurin & Ferreira, 2009), negative impact on ergonomics situations

(Brown et al., 2013), on job characteristics (Seppälä & Klemola, 2004), and on organizational

factors (Gelei, Losonci, & Matyusz, 2015). Accordingly, also our proposal of using OHS

leading indicators to measure the impact of lean implementation on OHS performance holds

for future research.

The next part of the study intended to examine the mediation effect of antecedents between

lean maturity and OHS performance (H6 to H9). Previous studies have highlighted the key

role of antecedents as mediators to OHS performance. For instance, DeJoy et al. (2004) dis-

cuss the role of management and organizational factors as mediators to OHS performance,

whereas Neal et al. (2000) highlight the mediation effect of safety climate on OHS perfor-

mance. Accordingly, through hypotheses from H6 to H9 we were striving to test the role of

four antecedents as mediators between lean maturity and OHS performance. By conducting

bootstrapping routine, the significance of the direct relationship between lean maturity and

OHS performance was grasped (p < 0.05). Next, the indirect effects of lean maturity on OHS

performance were analyzed by keeping the antecedents in the analysis. All the mediating ef-

fects of antecedents resulted statistically significant. Lastly, the magnitude of mediating ef-

fects was analyzed. The final results disclosed that the working environment and organiza-

tional factors have the highest mediating effects on OHS performance (i.e. 38%) meaning

38% of the lean maturity effect on OHS performance are mediated by WE (working envi-

ronment) and OF (organizational factors), since, the VAF is larger than 20%, but smaller

than 80% it is concluded that working environment and organizational factors have partial

mediating effect between lean maturity and OHS performance. The results of mediation ef-

fects for task characteristics, workforce characteristics are 32%, and 27% respectively. These

findings verify the proposed hypotheses (H6 to H9) regarding the mediating effects of ante-

cedents between lean implementation and OHS performance. By having identified the signif-

icance of the mediating effects of the antecedents, it can be concluded that to measure appro-

priately the impact of lean implementation on OHS performance the role of antecedents

should be taken into account. Further, the results of importance-performance matrix analyses

(IPMA) reveal that among the antecedents, working environment is the main construct to es-

tablish OHS performance. Consequently, to improve managerial activities, focusing on ante-

cedents especially on working environment is significant to measure the impact of lean ma-

turity on OHS performance.

Therefore, the predicted mediating effects of four antecedents (i.e. working environment,

workforce characteristics, task characteristics, and organizational factors) between lean ma-

turity and OHS performance is in support with previous studies that show the mediating ef-

fects of antecedents to OHS performance (Clarke, 2006; Siu, Phillips, & Leung, 2004; Zohar,

2002).

63

From the other point of view, the proposed model to steer OHS performance in the present

study (Paper B) exhibits the association between the antecedents and OHS leading indicators.

By combining the results of PLS-SEM and the theoretical results of the proposed model, we

are going to propose specific OHS leading indicators linked to lean practices.

Lastly, the two last hypotheses (H10, H11) in the present study are related to the effects of

business size and sector on the relationship between lean maturity and OHS performance.

Previous studies show that the implementation of lean practices is not similar in large and

small enterprises with respect to different contexts in these enterprises (Matt & Rauch, 2013).

More, the impact of lean implementation in different business sectors (i.e. manufacturing and

services) is a challenging topic among scholars. For instance, while some authors (Poksinska,

2010) show a limited success of lean implementation in services industries, others (Kim,

2002) report a fair success of lean implementation in manufacturing industries. Therefore, it

is reasonable to study the effects of business size and sector in connection with lean imple-

mentation. As described in preceding chapters, size and sector have a moderating effect on

the relationship between lean maturity and OHS performance. The multi-group analysis

(PLS-MGA) was performed in the SmartPLS to compare the path coefficient between groups

of data. By doing this kind of analysis, the moderating effects of business size and sector

were revealed. As shown in the results section, the sector of business does not have a signifi-

cant moderating effect between lean maturity and OHS performance. More, there is not a sig-

nificant difference between the effects of lean maturity on OHS performance in manufactur-

ing and services sectors. With respect to these findings, the hypothesis H10 is rejected in this

study. That is not in support of the studies by Poksinska (2010) and Kim (2002), where they

report different impacts of lean implementation in manufacturing and services sectors.

The moderating effects of business size also present no significant difference between small

and medium sized enterprises (SME) and large one while the impact of lean is measured on

the OHS performance. In other words, business size has not moderating effect on the relation-

ship between lean maturity and other constructs (i.e. antecedents and OHS performance).

Therefore, the hypothesis H11 is rejected in this study. This is not in support of the study of

Shah and Ward (2003), where they disclose a significant difference in impact of lean imple-

mentation between small and medium sized enterprises and large enterprises.

Summary of hypotheses testing

As eleven hypotheses were articulated for the present study based on the theoretical model,

through the PLS-SEM software they were tested. All hypotheses except two were accepted in

the current study. Table 19 summarizes the results of them.

https://repositories.lib.utexas.edu/discover?filtertype=author&filter_relational_operator=equals&filter=Kim%2C+Daeyoung

64

Table 19: Summary of hypotheses testing

Hypotheses Accepted?

(Yes/No)

H1: Lean implementation significantly influences OHS performance Yes

H2: Lean implementation significantly influences working environment Yes

H3: Lean implementation significantly influences task characteristics Yes

H4: Lean implementation significantly influences workforce characteristics Yes

H5: Lean implementation significantly influences organizational factors Yes

H6: Working environment significantly mediates the relationship between lean implementa-

tion and OHS performance. Yes

H7: Task characteristics significantly mediate the relationship between lean implementation

and OHS performance. Yes

H8: Workforce characteristics significantly mediate the relationship between lean imple-

mentation and OHS performance. Yes

H9: Organizational factors significantly mediate the relationship between lean implementa-

tion and OHS performance. Yes

H10: There is a significant categorical moderating effect of sector on the relationship among

model constructs. No

H11: There is a significant categorical moderating effect of business size on the relationship

among model constructs No

5.2 Proposal of a new causal model for predicting OHS performance via leading indicators The statistical significance of the relationship between lean maturity and antecedents shows

the key role of antecedents to predict the impact of lean implementation on OHS perfor-

mance. Therefore, to linking the OHS antecedents with OHS leading indicators a casual mod-

el is proposed. To do so, the results chain model, which has been introduced by several re-

searchers (i.e. Gertler et al., 2011; Jahanmehr et al., 2015), for outlining the program devel-

opment, is employed.

The proposed model illustrates how the sequence of events links the safety concepts from an-

tecedents to accidents and injuries defined as the final outcomes of safety efforts. Concerning

Craig (2016) the antecedent is defined as an input to safety efforts. The second item in the

proposed model relates to activities, which are defined as any action undertaken to inputs to

construct the outputs, that is, any undertaken action linked to antecedents. Since the base of

safety performance refers to safety activities, measuring the safety activities in organizations

will result in identifying early inconsistency with safety goals. In other words, analyzing and

evaluating the safety activities depicts the overall state of safety efforts in an organization. In

case any contradiction appears, corrective measures can be undertaken prior to accident and

injury. Therefore, in the interest of accident prevention, we should provide such indicators

that measure the status of safety activities. To do so, leading indicators have been introduced

by scholars. For instance, Mengolini and Debarberis (2008) introduce the leading indicators

65

to quantify the effectiveness of safety activities. The third item in the model is in connection

with the outputs produced by safety activates, which scholars call safety behaviors (Craig,

2016). For clear understanding, safety behavior has been categorized into two items: safety

participation and safety compliance, of which the former refers to employee involvement in

safety activities and the latter, refers to following OHS rules in the workplace. The next item

in the proposed model is linked to short-term result of activities, which are called outcomes.

In safety, this item is consistent with the definition of near-misses. Further, near-misses are

the transitional link between safety behaviors and final outcomes of safety efforts. Therefore,

by investigating the nature and reason of near-misses, corrective action can be implemented

to prevent accidents in the future. The last item placed in the model nominates the final out-

comes, which has long-term impacts. This definition is similar to the definition of accidents

and injuries in safety. In order to evaluate the safety interventions, the final outcomes of safe-

ty efforts are measured through lagging indicators like number of accidents in a year and the

amount of compensation paid to injured employees. By having defined all items, the proposed

model to predict OHS performance via leading indicators is shown in Figure 25.

Figure 25: The proposed model for safety concepts

By having defined the safety concepts through a clear framework, the placement of indicators

to measure the safety efforts are required for clear illustration. Two kinds of indicators have

already been proposed for safety and health programs: leading and lagging indicators. In re-

gards to this proposed framework, lagging indicators measure the final outcomes of safety

activities or events (Reiman & Pietikäinen, 2012). Therefore, the lagging indicators are con-

sidered as the after-the-fact indicators (Zwetsloot, Drupsteen, & de Vroome, 2014). Tradi-

tionally, various types of lagging indicators are utilized in the academic community and in-

dustries such as number of accidents, the rate of recordable injury, the amount of compensa-

tion to employee, and days away from workplace. These indicators represent the outcomes of

safety and health interventions. In other words, decision-making on the acceptance or rejec-

tion of a safety intervention is defined by utilizing lagging indicators. If the effectiveness of

the existing safety intervention is acceptable as measured by lagging indicators, managers can

maintain the program; otherwise, they should make changes to the program or replace it with

other programs.

In the present study, therefore, it is shown that to measure the impact of lean implementation

on occupational health and safety, the use of lagging indicators cannot be an appropriate op-

tion because they only provide information about the number of accidents, injuries, and the

amount of compensation caused by lean implementation. In other words, the after-the-fact

66

events have been measured and there in no way to conduct corrective actions and amend-

ments. Therefore, the final outcomes of lean implementation are only measured through lag-

ging indicators, and, do not provide appropriate knowledge on how and why these results

have happened. It seems various drawbacks to use lagging indicators to measure the impact of

lean implementation on occupational health and safety exist in this context. Hence, using such

indicators to measure the before-the-fact events are needed when it tried to measure the lean

implementation impacts.

In the present study, the usage of leading indicators has been proposed. The leading indicators

also have known as activities indicators that monitor the safety activities in the organization.

In a definition by step change in safety (2003), the leading indicators have been defined as

"something that provides information that helps the user respond to changing circumstances

and take actions to achieve desired outcomes or avoid unwanted outcomes” (p. 3). A new def-

inition for leading indicators is also proposed in this study as "a measure that provides infor-

mation on activities undertaken on the antecedents of safety performance" (p. 5). According

to this definition, the role of antecedents is considerably highlighted. By having defined the

antecedents of safety performance, related activities can be identified and finally monitored

and measured through leading indicators. In the present study, as described in previous sec-

tions, antecedents of safety performance are categorized into four items: working environ-

ment, workforce characteristics, task characteristics, and organizational factors. Therefore,

measuring any activity related to these antecedents is undertaken through leading indicators.

In regards to the objectives of this study, various leading indicators to measure the impact of

lean implementation are proposed.

According to definition of value-added work in lean context (Womack, Armstrong, & Liker,

2009), "the time that a worker spends physically transforming the product over the total cycle

time"(p. 283), if a suitable workplace is provided in regards to working environment, task

characteristics, workforce characteristics, and organizational factors, workers can conduct

their job conveniently, effectively, and efficiently. The final goal of lean implementation is

producing the product in right amount of time, right amounts, right quality level, and right

place. Non value-added works are related to an unsuitable workplace, including problematic

working environments, workforce characteristics, task characteristics, and organizational fac-

tors, resulting in time spent waiting and stopping over the cycle time. Therefore, the more

suitable the workplace, the higher the value added ratio and the lower the non value-added

ratio and finally the leaner job.

Since the leading indicators present the status of processes in the workplace and lean ap-

proach focuses on continuous improvement of working processes, from this perspective, the

relationship between lean and leading indicators is perceptible because both focus on activi-

ties. That is, lean approach strives to improve the activities throughout workplace, and OHS

leading indicators strive to monitor the undertaken activities throughout workplace to quanti-

fy the status of safety and health. These definitions clarify that a close association exists be-

tween lean practices and OHS leading indicators. By having investigated the activities

changed by lean practices from the safety perspective, the impact of lean implementation will

be recognized. This investigation can be undertaken through OHS leading indicators, which is

the main goal of this study that was articulated in the proposed model.

67

By using leading indicators to evaluate the status of health and safety of changed activities,

positive and negative effects of lean practices on employee safety and health are determined.

After having determined the activities from a safety perspective, integrating safety and lean

programs can also be addressed. Since lean professionals strive to improve the processes, any

information that can help figure out the status of processes will result in process advance-

ment. The OHS leading indicators provide information on the status of safety of workplace

processes. In case that problems arise in the processes from the safety perspective, employees

should increase their effort in order to complete the job and maintain the efficiency, which in

this condition can result in an increase of the number of mistakes and accidents. As described

earlier, mistakes and accidents endanger the lean flow of the processes. Therefore, OHS lead-

ing indicators help identify the latent failures of the antecedents of safety performance in ear-

ly steps, which help managers to perform preventive, proactive, and predictive actions. If the

changed processes undertaken by lean practices have safety issues, employees cannot perform

their task perfectly and, therefore, time will be wasted, efficiency impaired, and quality de-

graded. Generally speaking, employees prefer safety to lean efforts.

One of the main challenges in the success of lean initiatives refers to employee involvement

in the lean implementation process. As an employee is involved in the problem solving of the

lean implementation process, the success of lean practices is more assured. From this perspec-

tive, considering the safety and health of employees in the processes changed by lean practic-

es highlights the role of employees in the success of lean initiatives. Further, some challenges

arisen on the role of human factors in lean implementation. For instance, Yang et al. (2012)

note that, lean professionals mostly emphasize the technical practices of the lean initiatives

and, therefore, neglect the role of human factors. More, Shoaf et al. (2004) note that lean pro-

fessionals mostly focus on process optimization and, therefore, ignore the impact of new

work practices on employees. Employees are very important and have a central role in lean

efforts, and it is essential to guarantee that they feel well and that their health and safety are

assured. By considering the OHS leading indicators and the involvement of employees, a sus-

tainable lean system is achievable in the workplace. If lean implementation endangers the

safety and health of employees, employees experience problems in their job that jeopardize

the sustainability of lean initiatives.

Since there are various types of lean practices, the OHS leading indicators linked to them are

different. In appendix, the OHS leading indicators are proposed to various types of lean prac-

tices.

68

CHAPTER6

CONCLUSIONS AND FUTURE RESEARCH

This chapter represents the concluding remarks, including theoretical implications and practi-

cal/managerial implications, then study limitations and finally the propositions for future re-

search.

This study was conducted to shine new light on the usage of the OHS leading indicators to

measure the impact of lean implementation on occupational health and safety performance in

the workplace. Since companies are striving to improve their performance and increase the

profitability aligned with lesser material and resources, the lean approach is widely imple-

mented in different sectors of industries. In spite of gaining advantages due to lean implemen-

tation, some troubles have been reported in relation to the employee safety and health. For

instance, occupational stress, musculoskeletal disorders, and accidents are issues that oc-

curred due to lean implementation in the workplace. Although the issue of employee safety

and health had been overlooked in the past during lean implementation, fortunately, address-

ing safety and health is becoming more common among lean professionals. Aligned with this

trend, academic communities also work on the relationship between lean and safety. Numer-

ous studies analyze the impact of lean implementation on occupational health and safety.

However, most studies utilize the OHS lagging indicators to evaluate the impact of lean im-

plementation. Nevertheless, safety professionals state several downsides of using lagging in-

dicators to measure the safety and health performance. Therefore, the OHS leading indicators

have been introduced to overcome the drawbacks of lagging indicators, yet, there is a lack of

using leading indicators in the lean manufacturing subjects. Accordingly, the current study is

the first study that proposes the usage of OHS leading indicators to measure the impact of

lean implementation on occupational health and safety. To sum up, the following paragraphs

present the concluding remarks of the study.

6.1 Theoretical implications

The topic of the antecedents of safety performance was arisen in the early stages of research

development. There was not a clear and consistent definition for the antecedent of safety per-

formance, which has created some arguments among safety professionals. By searching rele-

vant literature within workplace safety subjects a list of contributions was provided. Then,

through contributions analysis, a robust and clear definition for the antecedent was provided

when it applied to safety performance. The antecedents extracted from literature were catego-

rized into four groups: working environment, workforce characteristics, task characteristics,

and organizational factors. Finally, a model distinguishing the safety performance conceptual-

ization was developed in regards to results of the literature review. The interesting findings of

the study were articulated in the form of a paper. Therefore, the paper itself provides im-

portant implications for the academic community. The robust and clear definition of the ante-

cedents as well as the developed framework for antecedents exhibit theoretical implications

of the current study. More, the formative indicators linked to each antecedent were lacking in

the extant studies that the findings of this study theoretically provide these indicators to OHS

69

antecedents. The resultant framework fills the gap of lacking attention to antecedents of safe-

ty performance in previous studies.

In order to reach the understanding of association between lean and OHS leading indicators,

the relationship between the antecedents of safety performance and the OHS leading indica-

tors was studied in this study. To successfully provide a relationship between the antecedents

of safety performance and the leading indicators, it was important to link several safety con-

ceptualizations in a unified framework. The results chain was utilized to provide a robust and

consistent framework for the safety conceptualizations and to highlight the importance of us-

ing leading indicators to measure the safety performance in the present study. The findings of

this part of the current study were also results in the second paper. The results include a novel

framework for the safety conceptualizations. Further, the unclear ideas and concepts related

to OHS leading indicators seem to be withdrawn. By having defined the safety concepts in a

sequence, the holistic framework was provided to explain the causal logic behind the safety

issues. Also, the proposed framework in this part facilitates the discussions on the subjects

linked to safety monitoring and evaluating. The importance of using leading indicators to

measure safety efforts was also recognized through the proposed framework. Furthermore,

the association between leading and lagging indicators was illustrated within the framework

to be used for measuring safety behaviors and safety outcomes. The findings showed that the

antecedents of safety performance are the inputs of safety efforts, and therefore their role is

significant to select the relevant OHS leading indicators. In other words, the antecedents are

the entrance of choosing OHS leading indicators. Overall, the theoretical implications of the

second paper are related to the following perspectives:

1. The importance of the antecedents of safety performance in achieving safety goals.

2. The role of safety activities in achieving safety goals.

3. The position of leading and lagging indicators among safety concepts.

4. The association between a near-miss and an accident.

5. The function of safety behaviors among safety concepts

The developed model to examine the relationship between lean maturity and OHS perfor-

mance also provides several theoretical implications. For instance, a novel concept linked to

lean was utilized in the current study; lean maturity. Although several studies have been con-

ducted about the lean maturity, no study utilizes the term "lean maturity" in the field of safety.

Through data analysis within SmartPLS software, a strong relationship was revealed between

lean maturity and its three forming items (i.e. Fidelity, extensiveness, and experience). This

finding is new in lean manufacturing context that help academic community working on this

issue in future research. Also, the proposed model to illustrate the relation between lean and

safety is significant to the academic community since there are limited models linked to the

relationship between lean and safety. Lastly, the outputs of SmartPLS confirm the key role of

the antecedents to be addressed within the relationship between lean and safety. Further, the

outputs show the importance of the mediating effects of antecedents between lean maturity

and OHS performance. The study results exhibit significant meditation effects of antecedents

on the relationship between lean maturity and OHS performance. All these findings assist the

academic community to understand deeply the relationship between lean and safety.

Finally, the introduction of specific OHS leading indicators, concerning the four antecedents,

linked to each lean practice provides an opportunity for researchers who would like to work

70

in this research area. The constructed questionnaire (shown in the Appendix) also provides a

pattern to the academic community to work in the research stream of lean and safety.

6.2 Practical/managerial implications

Different parts of the present study provide practical implications for manager and organiza-

tions who are interested in promoting the relationship between lean and safety. Since there are

some synergy and trade-off relations between lean and safety, the findings of this study help

managers and organizations to enforce the positive effects of lean implementation and reduce

the negative effects in connection with OHS situations.

The first part of the study, where develops a unified framework to antecedents of OHS per-

formance, can be utilized by safety professionals to consider the formative indicators of ante-

cedents when they are measuring the OHS performance in the workplace. The approved

formative indicators of antecedents can be taken into account to improve the status of the em-

ployee safety and health in the workplace. Also, managers will be enabled to monitor the sta-

tus of formative indicator related to the antecedents to measure safety activities and safety

behaviors in the workplace. More, the framework developed for antecedents assist managers

to select KPIs for safety monitoring and evaluating in the workplace. The second part of the

study, which is related to OHS leading indicators issues, also provide practical implications

for practitioners. For instance, by introducing leading indicators linked to lean practices,

managers can utilize them to monitor the impact of lean implementation on employee health

and safety prior to any adverse results. In other words, if any adverse impact from lean im-

plementation is detected through leading indicators, managers can modify the lean practice

itself or in severe situations can cease the implemented lean practice. According to the lead-

ing indicators definition, this detection helps to change the circumstances and avoid the un-

wanted outcomes. For lean implementation, leading indicators help to avoid the unwanted

outcomes on employee health and safety. By monitoring the activities related to lean imple-

mentation, positive effects of lean practice can be reinforced and negative effects can be re-

duced by various procedures such as employee training and practice modification. Therefore,

by eliminating the negative effects of lean practices on employee safety and health such as

long working hours, job stress, and musculoskeletal disorders, advantages of lean implemen-

tation for organizations including quality and productivity improvement will be significantly

promoted. By making consistent the safety and lean, employee safety and health from one

side and advantages of lean implementation from the other side will be provided concurrent-

ly. Using the OHS leading indicators induces safety professionals to engage in lean imple-

mentation processes in the workplace, which facilitates the relationship between safety and

lean professionals.

The third section of the study that analyzes the relationship between lean and safety addition-

ally provides some practical implications for managers and organizations. For example, the

lean maturity concept developed in the present study provides invaluable information to man-

agers who are interested in promoting the process of lean implementation in the organization.

To have a complete and perfect lean implementation, fidelity, extensiveness and experience

factors should be taken into account. As much as a lean practice being widely and thoroughly

implemented in an organization the success of lean implementation is more achievable.

71

6.3 Study limitations and future research

The proposed hypotheses regarding the moderating effects of business size and sector (H10

and H11) have not been supported. Possibly, the relative limited size of the sample has in-

duced this situation. Therefore, there are possible avenues for future research grounded on a

larger dataset. Therefore, there are possible avenues for future research, including maximize

the number of samples and study again the moderating effects of size and sector on the rela-

tionship between lean implementation and OHS performance. The contribution of lean ma-

turity constituents has been briefly explored in this study. Therefore, to establish a deeper un-

derstanding of lean maturity and its forming items, further investigation is needed. Although

there are some limitations and criticisms about using PLS-SEM, the findings of the study and

the outweighing benefits of using PLS-SEM has been identified in the present study to exam-

ine the relationship between lean implementation and OHS performance.

The findings of the present study provide more opportunity for scholars to conduct further

research in the research stream of lean and safety.

With respect to the first paper extracted from the present study, since there is a robust and

well-defined framework for the antecedents (which was lacking in previous studies), case

studies, theories development, research proposals, observations, and policy considerations can

be formulated in future research. More, the independency and interdependency among the

antecedents could be analyzed. The verification of the proposed model could also be ad-

dressed beside other attempts.

Also, the findings of the second paper provide important implications for future research. For

instance, the verification of the proposed framework linked to OHS leading indicators could

be addressed in future research. The testing of independency and interdependency among

safety conceptualization provides an opportunity for scholars who are interested in working in

this area. The developed framework provides a research stream to study the relationship be-

tween leading and lagging indicators in future works. The findings also create a space for re-

searchers to further analyze on the safety ideas and concepts illustrated through the final

framework. Experimental analyses can also be conducted to test the approach of measuring

safety performance through leading indicators or lagging indicators. More studies can be car-

ried out on the two concepts of observable actives and outcomes based on the topics illustrat-

ed in the framework. The interrelating safety concepts and the results chain model were uti-

lized for the first time in the current study. Therefore, it creates an opportunity for researchers

to develop other theories in the field of safety in their future studies. The approach testing the

interrelating safety concepts into the results chain model can also be conducted in an inde-

pendent study by researchers. Future research can utilize the proposed framework to develop

specific and more consistent methods for measuring OHS performance in different operation-

al contexts according to different priorities. In order to steer safety performance more appro-

priately, the existing study shines a light on this area, and based on the findings, requires ad-

ditional research to be conducted in the future. Therefore, research proposals, observations,

case studies, and policy considerations should be articulated in future research based on the

findings of the second paper.

The findings from the proposed model concerning the relationship between lean implementa-

tion and OHS performance additionally provide an important implication for future studies

72

where the term of lean maturity needs to be investigated more. Furthermore, the underlying

items for each indicator need more analysis. The model proposed for the relationship between

lean and safety could be additionally analyzed to verify the various variables within it and

finally proposed as a standard model for illustrating the relationship between lean and safety.

Further analysis also can be carried out in regards to the relationship between each indicator

of lean maturity and other variables in the model such as the relationship between fidelity and

OHS performance, the relationship between the experience and OHS performance, the rela-

tionship between the experience and its effects on the antecedents, and the relationship be-

tween fidelity, extensiveness and each antecedent independently.

Next, the four underlying indicators to measure the OHS performance level could also be ex-

amined itself to verify them and test their interrelation. Future research is proposed to consid-

er the status of OHS management maturity in companies since this item may affect the subse-

quent impacts of lean implementation on occupational health and safety.

The usage of the SmartPLS for data analysis in the field of safety seems to be new and cer-

tainly is interesting. Therefore, scholars could review the outputs of the present study and

then decide to utilize them within their research. However, some weaknesses of SmartPLS are

necessary to be addressed, as stated by Wong (2013):

"1. High-valued structural path coefficients are needed if the sample size is small.

2. Problem of multicollinearity if not handled well.

3. Since arrows are always single headed, it cannot model undirected correlation.

4. A potential lack of complete consistency in scores on latent variables may result in biased

component estimation, loadings and path coefficients.

5. It may create large mean square errors in the estimation of path coefficient loading."(p.3)

Therefore, future research can determine the appropriate usage of SmartPLS for analysis of

the relationship between lean implementation and OHS performance.

The last part of this study is the main part which proposes the dedicated OHS leading indica-

tors for common lean practices in the industries. Future works can be dedicated to more anal-

ysis on the leading indicators proposed for each lean practice. The case studies could be con-

ducted to investigate the relevance and usefulness of proposed leading indicators of lean prac-

tices in order to measure their impact on occupational health and safety. Additional OHS

leading indicators could be proposed for lean practices. Since the proposed OHS leading indi-

cators do not include all available lean practices, researchers can take the notes from this

study to propose OHS leading indicators to other lean practices. Further research can be dedi-

cated to investigating the relationship between OHS leading indicators and the lagging indica-

tors to measure the impact of lean implementation. Since the effect of size and sector is sig-

nificant to measure the impact of lean implementation, it needs to address this point while

proposing OHS leading indicators to various sectors and sizes of industries in the future re-

search. That is, based on the size and sector of industries specific OHS leading indicators

could be proposed in further analyses. By having known the OHS leading indicators through

this study for lean practices, further works can be carried out to propose new strategies and

approaches for minimizing the negative effects of lean implementation on occupational health

and safety in one hand and maximize the positive effects of lean implementation on health

and safety in the workplace. Therefore, specific safety behaviors and safety activities based

on each lean practice could be proposed and measured in the future research. Corrective ac-

73

tions and revisions could be undertaken by practitioners based on OHS leading indicators for

lean practices.

Also, future research can modify various parts of the questionnaire due to additional study

objectives. Some researchers can verify the different items existing in the questionnaire to

subsequently propose that as a standard questionnaire in the relationship between lean and

safety.

Overall, this study paves the way to a new stream of research where the systematic use of

leading indicators is leveraged for achieving a better understanding and measurement of the

multifaceted relationship between lean implementation and OHS performance.

74

Publications of thesis results

Preliminary and partial results of the thesis have been already published or submitted for publications as follows: - Mousavi, S.S., Jazani, R.K, Cudney, E., Trucco, P., “Linking lean implementation and occupa-

tional health and safety through leading indicators", International Journal of Lean Six Sigma. Un-der Review.

- Mousavi S., Cudney E., Trucco P., “Towards a framework for steering safety performance: a re-view of the literature on leading indicators”, in Advances in Safety Management and Human Fac-tors, Arezes P. (Ed), pp. 195-204, 2018. Part of the Advances in Intelligent Systems and Compu-ting book series (AISC, volume 604). Springer, Cham. DOI: 10.1007/978-3-319-60525-8_21

- Mousavi S., Cudney E., Trucco P., “What are the antecedents of safety performance in the work-place? A critical review of literature”, Proceedings of the 67th Annual Conference and Expo of the Institute of Industrial Engineers, K. Coperich, E. Cudney, H. Nembhard, (eds), pp. 1-6., May 20-23, 2017, Pittsburgh, USA.

75

References

Aij, K. H., Simons, F. E., Widdershoven, G. A. M., & Visse, M. (2013). Experiences of leaders in the implementation of Lean in a teaching hospital—barriers and facilitators in clinical practices: a qualitative study. BMJ Open, 3(10), e003605.

Albaum, G. S., & Smith, S. M. (2006). Handbook of marketing research. London, United. Anand, G., & Kodali, R. (2008). Selection of lean manufacturing systems using the

Promethee. Journal of Modelling in Management, 3(1), 40–70. Andel, S. A., Hutchinson, D. M., & Spector, P. E. (2015). Safety at Work: Individual and

Organizational Factors in Workplace Accidents and Mistreatment. In Research in

Personnel and Human Resources Management (pp. 235–277). Emerald Group Publishing Limited.

Ansari, S. M., Fiss, P. C., & Zajac, E. J. (2010). Made to fit: How practices vary as they diffuse. Academy of Management Review, 35(1), 67–92.

ANSI/API. (2010). Recommended Practice 754: Process safety performance indicators for

the refining and petrochemical industries. Anvari, A., Zulkifli, N., & Yusuff, R. M. (2011). Evaluation of approaches to safety in lean

manufacturing and safety management systems and clarification of the relationship between them. World Applied Sciences Journal, 15(1), 19–26.

Bacharach, S. B. (1989). Organizational theories: Some criteria for evaluation. Academy of

Management Review, 14(4), 496–515. Barclay, D., Higgins, C., & Thompson, R. (1995). The partial least squares (PLS) approach to

causal modeling: Personal computer adoption and use as an illustration. Technology

Studies, 2(2), 285–309. Barling, J., & Zacharatos, A. (1999). High performance safety systems: Ten management

practices for creating safe organizations. In SK Parker & MA Griffin (Chairs), Managing safety at work: Beyond blam ing the individual. Symposium conducted at the Acad—emy

of Management Conference, Chicago. Baron, R. M., & Kenny, D. A. (1986). The moderator–mediator variable distinction in social

psychological research: Conceptual, strategic, and statistical considerations. Journal of

Personality and Social Psychology, 51(6), 1173. Bartlett, K. R. (2005). Survey research in organizations. Research in Organizations:

Foundations and Methods of Inquiry, 97–113. Berggren, C. (1993). Alternatives to lean production: Work organization in the Swedish auto

industry. Cornell University Press. Betcherman, G., Mcmullen, K., Leckie, N., & Caron, C. (1994). The Canadian Workplace in

Transition, IRC Press, 142p, 1994.gston, Ontario, Canada. Birkie, S. E. (2016). Operational resilience and lean: in search of synergies and trade-offs.

Journal of Manufacturing Technology Management, 27(2), 185–207. Borman, W., & Motowidlo, S. (1997). A Theory of Individual Differences in Task and

Contextual Performance. Human Performance, 10(2), 99–109. Braunscheidel, M. J., & Hamister, J. W. (2012). The Impact of Lean Practices on Value to

Customer. The BRC Academy Journal of Business, 2(1), 51–86. Brenner, M. D., Fairris, D., & Ruser, J. (2002). Flexible Work Practices and Occupational

Safety and Health : Exploring the Relationship. Political Economy Research Institute, 43(30), 242–266.

Brown, G. D., O’Rourke, D., & Rourke, D. O. (2013). Lean manufacturing comes to China: a case study of its impact on workplace health and safety. International Journal of

Occupational and Environmental Health, 13(3), 249–57. Bruno, R., & Jordan, L. (2002). Lean Production and the Discourse of Dissent. Radicalizing

76

the Shop Floor at Mitsubishi Motors? WorkingUSA, 6(1), 108–134. Chin, W. W. (1998). The partial least squares approach to structural equation modeling.

Modern Methods for Business Research, 295(2), 295–336. Christian, M. S., Bradley, J. C., Wallace, J. C., & Burke, M. J. (2009). Workplace safety: A

meta-analysis of the roles of person and situation factors. Journal of Applied

Psychology, 94(5), 1103–1127. Clarke, S. (2006). The relationship between safety climate and safety performance: a meta-

analytic review. Educational Publishing Foundation. Clarke, S. (2010). An integrative model of safety climate: Linking psychological climate and

work attitudes to individual safety outcomes using meta-analysis. Journal of

Occupational & Organizational Psychology, 83(3), 553–578. Colombo, M., Bucher, L., & Sprenger, J. (2017). Determinants of judgments of explanatory

power: Credibility, Generality, and Statistical Relevance. Conti, R., & Angelis, J. (2006). The effects of lean production on worker job stress.

International Journal of Operations & Production Management, 26(9), 1013–1038. Craig Wallace, J. (2016). Creating a safety conscious organization and workforce.

Organizational Dynamics. Cua, K. O., McKone, K. E., & Schroeder, R. G. (2001). Relationships between

implementation of TQM, JIT, and TPM and manufacturing performance. Journal of

Operations Management, 19(6), 675–694. Cudney, E. A., Furterer, S., & Dietrich, D. (2013). Lean Systems: Applications and Case

Studies in Manufacturing, Service, and Healthcare. CRC Press. Cudney, E. A., Murray, S. L., & Pai, P. M. (2010). An Analysis of the Impact of Lean and

Safety. D’Mello, A. M., Turkeltaub, P. E., & Stoodley, C. J. (2017). Cerebellar tDCS modulates

neural circuits during semantic prediction: A combined tDCS-fMRI study. Journal of

Neuroscience, 2816–2818. Dal Pont, G., Furlan, A., & Vinelli, A. (2008). Interrelationships among lean bundles and

their effects on operational performance. Operations Management Research, 1(2), 150–158.

Danna, K., & Griffin, R. W. (1999). Health and Well-Being in the Workplace : A Review and Synthesis of the Literature. Journal of Management, 25(3), 357–384.

De Souza, L. B., & Pidd, M. (2011). Exploring the barriers to lean health care implementation. Public Money & Management, 31(1), 59–66.

DeJoy, D. M., Schaffer, B. S., Wilson, M. G., Vandenberg, R. J., & Butts, M. M. (2004). Creating safer workplaces: assessing the determinants and role of safety climate. Journal

of Safety Research, 35(1), 81–90. Dora, M., Kumar, M., Van Goubergen, D., Molnar, A., & Gellynck, X. (2013). Operational

performance and critical success factors of lean manufacturing in European food processing SMEs. Trends in Food Science & Technology, 31(2), 156–164.

Dźwiarek, M. (2004). An analysis of accidents caused by improper functioning of machine control systems. International Journal of Occupational Safety and Ergonomics, 10(2), 129–136.

Elbert, M. (2012). Lean production for the small company. CRC Press. Embrey, D. E. (1992). Incorporating management and organisational factors into probabilistic

safety assessment. Reliability Engineering & System Safety, 38(1), 199–208. Erickson, P. A. (1996). Practical guide to occupational health and safety. Elsevier. European Union. (2003). Commission Recommendation of 6 May 2003 concerning the

definition of micro, small and medium-sized enterprises. Official Journal of the

European Union, 36–41.

77

Florida, R. (1996). Lean and green: the move to environmentally conscious manufacturing. California Management Review, 39(1), 80–105.

Friedman, M., & Rosenman, R. H. (1974). Type A Behavior and Your Heart. New York, Alfred A Knopf. Inc.

Gelei, A., Losonci, D., & Matyusz, Z. (2015). Lean production and leadership attributes–the case of Hungarian production managers. Journal of Manufacturing Technology

Management, 26(4), 477–500. Gertler, P. J., Martinez, S., Premand, P., Rawlings, L. B., & Vermeersch, C. M. J. (2011).

Impact evaluation in practice. World Bank Publications. GHosh, M. (2012). lean manufacturing performance in indian manufacturing plants. Gibb, A. G. F., Haslam, R., Gyi, D. E., Hide, S., & Duff, R. (2006). What causes accidents? Hair Jr, J. F., Hult, G. T. M., Ringle, C., & Sarstedt, M. (2016). A primer on partial least

squares structural equation modeling (PLS-SEM). Sage Publications. Hallowell, M. R., Veltri, A., & Johnson, S. (2009). Safety & Lean: One manufacturer’s

lessons learned and best practices, (November), 22–27. Haslam, R. A., Hide, S. A., Gibb, A. G. F., Gyi, D. E., Pavitt, T., Atkinson, S., & Duff, A. R.

(2005). Contributing factors in construction accidents. Applied Ergonomics, 36(4 SPEC. ISS.), 401–415.

Hasle, P., Bojesen, A., Jensen, P. L., & Bramming, P. (2012). Lean and the working environment.

Hinze, J., Thurman, S., & Wehle, A. (2013). Leading indicators of construction safety performance. Safety Science, 51(1), 23–28.

Hofmann, D. A., Jacobs, R., & Landy, F. (1995). High reliability process industries: Individual, micro, and macro organizational influences on safety performance. Journal

of Safety Research, 26(3), 131–149. Hofmann, D. A., Morgeson, F. P., & Gerras, S. J. (2003). Climate as a moderator of the

relationship between leader-member exchange and content specific citizenship: Safety climate as an exemplar. The Journal of Applied Psychology, 88(1), 170–178.

Holweg, M. (2007). The genealogy of lean production. Journal of Operations Management, 25(2), 420–437.

Hubbard, E. . (2004). The Diversity Scorecard: Evaluating the Impact of Diversity on

Organizational Performance. Elsevier Butterworth-Heinemann, Burlington, MA. Jahanmehr, N., Rashidian, A., Khosravi, A., Farzadfar, F., Shariati, M., Majdzadeh, R., Sari, A.A., Mesdaghinia, A.: A conceptual framework for evaluation of public health and primary

care system performance in Iran. Glob. J. Health Sci. 7(4), 341 (2015) Janicak, C. A. (2009). Safety metrics: Tools and techniques for measuring safety

performance. Government Institutes. Kapp, E. A. (2012). The influence of supervisor leadership practices and perceived group

safety climate on employee safety performance. Safety Science, 50(4), 1119–1124. Kim, D. (2002). Exploratory study of lean construction: Assessment of lean implementation. Kovacheva, A. V. (2010). Challenges in Lean implementation: successful transformation

towards lean enterprise. Aarhus: University of Aarhus. Kumar, S., & Phrommathed, P. (2005). Research methodology. Springer. Landsbergis, Paul; Cahill, Janet; Schnall, P. (1999). The impact of lean production and

related new systems of work organisation on worker health .pdf. Journal of

Occupational Health Psychology, 4(2), 108–130. Lawrence, J. J., & Hottenstein, M. P. (1995). The relationship between JIT manufacturing and

performance in Mexican plants affiliated with US companies. Journal of Operations

Management, 13(1), 3–18. Lewchuck, W., Stewart, P., & Yates, C. (2001). Quality of working life in the automobile

78

industry: a canada-UK comparative study. New Technology, Work and Employment, 16(2), 72–87.

Li, S., Rao, S. S., Ragu-Nathan, T. S., & Ragu-Nathan, B. (2005). Development and validation of a measurement instrument for studying supply chain management practices. Journal of Operations Management, 23(6), 618–641.

Longoni, A., Pagell, M., Johnston, D., & Veltri, A. (2013). When does lean hurt? – an exploration of lean practices and worker health and safety outcomes. International

Journal of Production Research, 51(11), 3300–3320. M. Goldenhar, L., Williams, L. J., G. Swanson, N., Goldenhar, L. M., Williams, L. J., &

Naomi, G. (2003). Modelling relationships between job stressors and injury and near-miss outcomes for construction labourers. Work & Stress, 17(April 2013), 218–240.

Marcoulides, G. A., & Chin, W. W. (2013). You write, but others read: Common methodological misunderstandings in PLS and related methods. In New perspectives in

partial least squares and related methods (pp. 31–64), Springer. Martínez Sánchez, A., & Pérez Pérez, M. (2001). Lean indicators and manufacturing

strategies. International Journal of Operations & Production Management, 21(11), 1433–1452.

Matt, D. T., & Rauch, E. (2013). Implementation of lean production in small sized enterprises. Procedia CIRP, 12, 420–425.

McClay, R. E. (1989). Toward a more universal model of loss incident causation. Professional Safety, 34(1), 15–20.

Mengolini, A., & Debarberis, L. (2008). Effectiveness evaluation methodology for safety processes to enhance organisational culture in hazardous installations. Journal of

Hazardous Materials, 155(1–2), 243–252. Mi, S., Park, D., & Pettersen, J. (2009). Defining lean production.

Monge, C., Cruz, J., & López, F. (2014). Manufacturing and continuous improvement areas using partial least squares path modeling with multiple regression comparison. In Proceedings of CBU International Conference on Innovation, Technology Transfer and

Education (2014), February (pp. 3–5). Moori, R. G., Pescarmona, A., & Kimura, H. (2013). Lean manufacturing and business

performance in Brazilian firms. JOSCM: Journal of Operations and Supply Chain

Management, 6(1), 91. Mousavi, S. S., Cudney, E. A., & Trucco, P. (2017). Towards a Framework for Steering

Safety Performance: A Review of the Literature on Leading Indicators. In International

Conference on Applied Human Factors and Ergonomics (pp. 195–204). Springer. N.Agumba, J., Thwala, W., & Haupt, T. (2011). Identification of health and safety

performance improvement measuring indicators: A literature review. Nahm, A. Y., Vonderembse, M. A., & Koufteros, X. A. (2003). The impact of organizational

structure on time-based manufacturing and plant performance. Journal of Operations

Management, 21(3), 281–306. Nahmens, I., & Ikuma, L. H. (2011). Effects of lean construction on sustainability of modular

homebuilding. Journal of Architectural Engineering, 18(2), 155–163. Nahrgang, J. D., Morgeson, F. P., & Hofmann, D. A. (2011). Safety at work: A meta-analytic

investigation of the link between job demands, job resources, burnout, engagement, and safety outcomes. Journal of Applied Psychology, 96(1), 71–94.

Neal, A., Griffin, M. . A., & Hart, P. . M. (2000a). The impact of organizational climate on safety climate and individual behavior. Safety Science, 34(1–3), 99–109.

Neal, A., & Griffin, M. A. (2002). Safety Climate and Safety Behaviour. Australian Journal

of Management, 27(1 suppl), 67–75. Neal, A., & Griffin, M. A. (2006). A study of the lagged relationships among safety climate,

79

safety motivation, safety behavior, and accidents at the individual and group levels. Journal of Applied Psychology, 91(4), 946–953.

Neal, A., Griffin, M. A., & Hart, P. M. (2000b). The impact of organizational climate on safety climate and individual behavior. Safety Science, 34(1–3), 99–109.

Netland, T., & Ferdows, K. (2014). What to Expect From a Corporate Lean Program What to Expect From a Corporate Lean Program. MIT Sloan Management Review, Summer, 83–89.

Nightingale, D. J., & Mize, J. H. (2002). Development of a lean enterprise transformation maturity model. Information Knowledge Systems Management, 3(1), 15–30.

Norman, D. A. (1981). Categorization of action slips. Psychological Review, 88(1), 1. Ohno, T. (1988). (1988). Toyota Production System. Productivity Press, 4(1), 3–11. Orpen,

C. (1982). Type A personality as a moderator of the effects of role conflict, role ambiguity and role overload on individual strain. Journal of Human Stress, 8(2), 8–14.

Parker, S. K., Axtell, C. M., & Turner, N. (2001). Designing a safer workplace: importance of job autonomy, communication quality, and supportive supervisors. Journal of

Occupational Health Psychology, 6(3), 211. Paté-Cornell, M. E. (1990). Organizational aspects of engineering system safety: The case of

offshore platforms. Science, 250(4985), 1210–1217. Piercy, N., & Rich, N. (2009). Lean transformation in the pure service environment: the case

of the call service centre. International Journal of Operations & Production

Management, 29(1), 54–76. Pinsonneault, A., & Kraemer, K. (1993). Survey research methodology in management

information systems: an assessment. Journal of Management Information Systems, 10(2), 75–105.

Poksinska, B. (2010). The current state of Lean implementation in health care: literature review. Quality Management in Healthcare, 19(4), 319–329.

Portioli-Staudacher, A. (2009). Lean implementation in service companies. In IFIP

International Conference on Advances in Production Management Systems (pp. 652–659). Springer.

Preacher, K. J., & Hayes, A. F. (2008). Asymptotic and resampling strategies for assessing and comparing indirect effects in multiple mediator models. Behavior Research

Methods, 40(3), 879–891. Qien, K., Utne, I. B., & Herrera, I. A. (2011). Building Safety indicators: Part 1 - Theoretical

foundation. Safety Science, 49(2), 148–161. Rahman, S., Laosirihongthong, T., & Sohal, A. S. (2010). Impact of lean strategy on

operational performance: a study of Thai manufacturing companies. Journal of

Manufacturing Technology Management, 21(7), 839–852. Recht, J. L. (1966). Systems safety analysis: Error rates and costs. National Safety News, 20–

23. Reiman, T., & Pietikäinen, E. (2012a). Leading indicators of system safety–monitoring and

driving the organizational safety potential. Safety Science, 50(10), 1993–2000. Reiman, T., & Pietikäinen, E. (2012b). Leading indicators of system safety - Monitoring and

driving the organizational safety potential. Safety Science, 50(10), 1993–2000. Robert Karasek, T. T. (1989). Healthy Work: Stress, Productivity, and the Reconstruction of

Working Life. Rose, N. L. (2016). Profitability and Product Quality : Economic Determinants of Airline

Safety Performance Author ( s ): Nancy L . Rose Published by : University of Chicago. Russell, S. N., & Millar, H. H. (2014). Exploring the relationships among sustainable

manufacturing practices, business performance and competitive advantage: Perspectives from a developing economy. Journal of Management and Sustainability, 4(3), 37.

80

Sagot, J.-C., Gouin, V., & Gomes, S. (2003). Ergonomics in product design: safety factor. Safety Science, 41(2), 137–154.

Salminen, S., & Klen, T. (1994). Accident locus of control and risk taking aming forestry and construction workers. Perceptual and Motor Skills, 78(3), 852–854.

Satoğlu, Ş. I., & Durmuşoğlu, M. B. (2003). A field study on measuring the lean maturity level in manufacturing firms in Turkey. Endüstri Mühendisliği Dergisi, 14(3).

Saurin, T. A., & Ferreira, C. F. (2009). The impacts of lean production on working conditions: A case study of a harvester assembly line in Brazil. International Journal of

Industrial Ergonomics, 39(2), 403–412. Seppälä, P., & Klemola, S. (2004). How do employees perceive their organization and job

when companies adopt principles of lean production? Human Factors and Ergonomics

in Manufacturing & Service Industries, 14(2), 157–180. Shah, R., & Ward, P. T. (2003). Lean manufacturing : context , practice bundles , and

performance, 21, 129–149. Shah, R., & Ward, P. T. (2007). Defining and developing measures of lean production.

Journal of Operations Management, 25(4), 785–805. Shikdar, A. A., & Sawaqed, N. M. (2003). Worker productivity, and occupational health and

safety issues in selected industries. Computers & Industrial Engineering, 45(4), 563–572.

Shoaf, C., Genaidy, A., Karwowski, W., & Huang, S. H. (2004). Improving performance and quality of working life: A model for organizational health assessment in emerging enterprises. Human Factors and Ergonomics in Manufacturing & Service Industries, 14(1), 81–95.

Sinelnikov, S., Inouye, J., & Kerper, S. (2015). Using leading indicators to measure occupational health and safety performance. Safety Science, 72, 240–248.

Siu, O., Phillips, D. R., & Leung, T. (2004). Safety climate and safety performance among construction workers in Hong Kong: The role of psychological strains as mediators. Accident Analysis & Prevention, 36(3), 359–366.

Sparks, K., Faragher, B., & Cooper, C. L. (2001). Well‐being and occupational health in the 21st century workplace. Journal of Occupational and Organizational Psychology, 74(4), 489–509.

Step Change in Safety. (2003). Leading Performance Indicators: Guidance for Effective Use. Step Change in Safety, Aberdeen, UK. Step Change in Safety, 2003. Retrieved from http://scholar.google.com

Strutz, M. (2008). A Retrospective Study of Skills, Traits, Influences, and School Experiences of Talented Engineers. In ASEE IL/IN Section Conference, Terre Haute.

Taj, S. (2008). Lean manufacturing performance in China: assessment of 65 manufacturing plants. Journal of Manufacturing Technology Management, 19(2), 217–234.

Todorova, D. (2013). Exploring Lean Implementation Success Factors in Job Shop, Batch Shop, and Assembly Line Manufacturing Settings.

Törner, M. (2008). Safety climate in a broad context: what is it, how does it work, and can it be managed? Scandinavian Journal of Work, Environment & Health (5), 5–8.

Vorley, G. (2008). Mini Guide To Root Cause Analysis, 7–12. Wacker, J. G. (1998). A definition of theory: research guidelines for different theory-building

research methods in operations management. Journal of Operations Management, 16(4), 361–385.

Womack, J. P., Jones, D. T., & Roos, D. (1990). The Machine that Changed the World: The

Story of Lean Production. World. Womack, S. K., Armstrong, T. J., & Liker, J. K. (2009). Lean job design and musculoskeletal

disorder risk: A two plant comparison. Human Factors and Ergonomics in

81

Manufacturing, 19(4), 279–293. Wong, K. K.-K. (2013). Partial least squares structural equation modeling (PLS-SEM)

techniques using SmartPLS. Marketing Bulletin, 24(1), 1–32. Wu, T.-C. C., Chen, C.-H. H., & Li, C.-C. C. (2008). A correlation among safety leadership,

safety climate and safety performance. Journal of Loss Prevention in the Process

Industries, 21(3), 307–318. Wu, W., Yang, H., Chew, D. A. S., Yang, S., Gibb, A. G. F., & Li, Q. (2010). Towards an

autonomous real-time tracking system of near-miss accidents on construction sites. Automation in Construction, 19(2), 134–141.

Yang, C., Yeh, T., & Yang, K. (2012). The implementation of technical practices and human factors of the Toyota production system in different industries. Human Factors and

Ergonomics in Manufacturing & Service Industries, 22(6), 541–555. Yeow, P. H. P., & Sen, R. N. (2003). Quality, productivity, occupational health and safety

and cost effectiveness of ergonomic improvements in the test workstations of an electronic factory. International Journal of Industrial Ergonomics, 32(3), 147–163.

Zacharatos, A., Barling, J., & Iverson, R. D. (2005). High-performance work systems and occupational safety. The Journal of Applied Psychology, 90(1), 77–93.

Zohar, D. (2002). The effects of leadership dimensions, safety climate, and assigned priorities on minor injuries in work groups. Journal of Organizational Behavior, 23(1), 75–92.

Zohar, D. (2003). The effects of leadership dimensions, safety climate, and assigned priorities on minor injuries in work groups. Journal of Organizational Behavior, 23(1), 75–92.

Zohar, D., & Luria, G. (2003). The use of supervisory practices as leverage to improve safety behavior: A cross-level intervention model. Journal of Safety Research, 34(5), 567–577.

Zwetsloot, G. I. J. M., Drupsteen, L., & de Vroome, E. M. M. (2014). Safety, reliability and worker satisfaction during organizational change. Journal of Loss Prevention in the

Process Industries, 27, 1–7.

82

Appendices

Appendix A

OHS leading indicators proposed to lean practices

Relationship between value stream mapping (VSM) and OHS leading indicators

In the VSM method, we follow two subjects: material flow and information flow. When the

current mapping of a process needs to be provided, we must take a gemba walk through lines

and review the processes and details such as material inventory, transportation, and the work-

stations. To identify the non value-added activities, the root causes must be determined. As

we suppose the latent failures in processes as waste, they should be addressed beside the other

activities in creating VSM. Some of popular failures that should be addressed in VSM include

ergonomics problems, a loud level of occupational noise, inadequate illumination, and chemi-

cal hazards. More, in regards to the VSM method, creating continuous flow of information

and material is undertaken. Therefore, after mapping the current value stream in the process-

es, modifications are needed to reach an optimum situation for processes in the future. During

this development, changes are made to the processes that likely create some safety and health

problems for employees. Do these changes ignore safety issues? Is it allowed to overlook the

safety issues for the sake of smooth production? Is it allowed to overlook the safety issue for

the sake of increasing work speed, such as, unloading machinery guards?

Two main aspects of VSM are related to the identification of human activity and machinery

operations, which both can contribute to safety performance when changes to them are im-

plemented. On the other hand, using the VSM method shortens cycle time, which this issue

again contributes to safety performance through occupational stress, work pace, and machine

operations.

Do the implemented changes in the workplace through VSM induce occupational stress and

maximize human errors?

The above-mentioned notes shine a light on the possible usage of OHS leading indicators to

measure the impact of VSM on safety and health of employees. Therefore, in this part, sever-

al OHS leading indicators linked to the VSM method are introduced.

OHS leading indicators linked to VSM method

- Train employee in safety principles due to undergone changes within production line

- Measure occupational fatigue of employees

- Conduct risk assessment before and after implementing the final VSM state

- Assess occupational stress after implementing the VSM

- Test safety knowledge of employees due to undertaken changes induced by the VSM

- Periodically inspect tools and equipment

- Survey personal activities due to implemented changes

- Investigate safety procedures

- Assess management commitment

- Determine safety attitude of employees due to VSM implementation

- Assess workload

- Assess job demand

83

- Inspect machinery guards

- Assess accumulation of created risks

- Assess ergonomics of workstations

- Observe suppliers' safety principles due to extended value stream mapping

Relationship between 5S and OHS leading indicators

5S is a method that provides the right quantity of and time for products and services for cus-

tomers. This approach causes quality and productivity improvement and provides a safe

workplace for employees. This method covers tool, equipment, and workplace cleaning in

one hand and applying discipline within the workplace in the other hand. When utilizing 5S,

safety principles cannot be ignored. For instance, unloading machinery guards should not

happen due to production issues, and it should be ensured that safety guards are not regarded

as waste during 5S utilization.

It is essential to address the safety and lean principles simultaneously when employing the 5S

approach. It is not allowed to overlook one of these principles due to maintaining the other

one. By having addressed the safety and lean principle simultaneously, final results will be

significant for both safety and lean approaches. Safety principles of workplace are maintained

and advantages of lean efforts are provided, including quality and productivity improvements.

It is recommended to conduct a risk assessment before employing 5S to identify hazard

points. By doing so, it is assessable whether the points have been removed from the work-

place or not. It is helpful to suggest to 5S team regarding the safety principle during 5S utili-

zation to ensure no safety rules are ignored due to production issues. Even, after 5S comple-

tion, risk assessment can be repeatedly undertaken to guarantee the status of safety in the

workplace.

OHS leading indicators linked to 5S method

-Inspect tools and equipment after 5S implementation

- Inspect safety procedures

- Review workload in regards to removing waste from the workplace

-Inspect machinery guards after 5S implementation

- Inspect work processes from a safety perspective

-Assess ergonomics risks due to an increase to documentation processes within the 5S ap-

proach

- Survey working environment

- Assess management commitment

- Survey safety culture

- Assess organization's policies on safety issues

Relationship between single-minute exchange of dies (SMED) and OHS leading indicators

The SMED has been developed to improve the machine setup time and attempts to convert

the internal setup to an external setup, which significantly reduces the changeover time of

machinery. In other words, it tries to change the tooling in the machine while is operating.

This procedure leads to reduced downtime of the machine, which finally increases the

productivity. The SMED has various benefits such as, fewer physical adjustments, lesser set-

up time than takt time, less expense of excess inventory, less material, reduced variation be-

84

tween each setup, reduced defects, more cost-effective products, simplified setup, and effi-

cient use of spaces.

OHS leading indicators linked to SMED method

- Assess occupational stress due to time reduction of machinery set up

- Assess ergonomics risks due to extensive and frequent force and frequent lifting of tools and

equipment within the SMED technique

- Assess workload after the SMED implementation

- Identify physical hazards such as noise, vibration, and incidents due to tools and equipment

loading

- Inspect using personal protective equipment (PPE)

- Monitor working environment including illumination, noise, and vibration

- Assess workstations in regards to available space for activities

- Survey changeover instructions in order to prevent human error

- Assess employee training, especially changeover training

- Survey visual markers for changeover and its inspection

- Survey job conflict in regards to parallel operation technique.

- Inspect safety signals for changeover

- Conduct training on safe procedures to perform the changeover

Relationship between total productive maintenance (TPM) and OHS leading indicators

The TPM was developed in the 1970's as a method of involving machine operators in preven-

tive maintenance of the machineries. This method attempts to eliminate equipment-related

defects. By doing so, downtime of machinery will be reduced. Further, the barriers between

departments in this approach will be eliminated. In order to reach the goals of the TPM, three

objectives have been introduced by Cudney et al. (2014):

"- Total employee involvement

- Hands-on approach

- Improve the organization's competitiveness" (p.103).

Two techniques are known for this approach: preventive maintenance (PM) and predictive

maintenance.

In summary, TPM methodology attempts to eliminate the downtime of machinery.

OHS leading indicators linked to TPM methodology

- Periodically inspect of tools and equipment regarding safety issues

- Inspect safety signals to ensure their existence and operation

- Monitor employee training in regards to safety principles during protection and maintenance

processes

- Survey ergonomics factors and workstations

- Assess workload of employee

- Survey working environment factors associated with inspecting tools and equipment to

identify defects

- Inspect using PPE during machinery protection and maintenance

-Survey TPM instruction in regards to safety principles

- Survey Tag out/Lock out procedures during machinery maintenance

- Survey job conflicts in order to prevent human error

85

- Assess occupational stress

- Assess ergonomics factors including awkward positions, extensive and frequent force, and

frequent lifting

- Identify physical hazards such as noise, vibration, and incidents due to tools and equipment

loading

Relationship between cellular manufacturing and OHS leading indicators

The cellular manufacturing methodology uses multiple cells in production lines. Each cell

includes various types of machines to accomplish a certain task. The products move from one

cell to another, so some part of product is completed in each station. The formed cell is gen-

erally arranged in a "U" shape. The biggest advantage of this methodology is its flexibility.

That is, simple changes are made very rapidly because of the existence of automatic machine

in production processes. By utilizing this methodology, products are manufactured as quickly

as possible, and various types of similar products are produced, which results in little waste in

the production processes.

OHS leading indicators linked to cellular manufacturing methodology

- Assess accumulation of created risks linked to joined workstations

- Survey occupational stress and work pressure

- Assess ergonomics risks

- Assess the degree of job conflict

- Assess time pressure

- Survey working environment such as physical, chemical, and biological hazards

- Assess workload and pressure

Relationship between one-piece-flow and OHS leading indicators

The one-piece-flow methodology attempts to reduce inventory between work cells, which

leads to improvement and work balance. Therefore, little work in process inventory exists be-

tween work cells resulting streamline flow in the most processes. The implementations of

one-piece-flow create some safety problems that should be addressed.

OHS leading indicators linked to one-piece-flow methodology

- Assess ergonomics risks including awkward postures, frequent lifting, and extensive force

- Survey physical hazards such as noise, illumination, and vibration in work cells

- Assess accumulation of created risks linked to work cells

- Assess the degree of job conflict in work cells


- Survey working environment such as noise, illumination, and dust


- Assess machinery and tool safety

- Assess employee training on safety principle linked to work cells


Relationship between Kanban and OHS leading indicators

The Kanban methodology is based on customer demand that utilizes signals to replenish the

material. Indeed, signals control the production flow. In contrast to a push system, which pro-

duces a high amount of products to be demanded by customers in future, a pull system oper-

86

ates on customer request. Therefore, a pull system is customer- based. The occupational stress

and work procedures are two main concerns related to safety issue.

OHS leading indicators linked to Kanban methodology

- Assess occupational stress related to Kanban methodology

- Assess ergonomics risks including awkward postures, frequent lifting and repetitive motions

due to documentation process in Kanban methodology

- Assess work procedures



- Assess compute-based activities and related ergonomics risks

- Monitor employee training

- Assess workstations and their required space

- Survey defined/ clear job functions

- Survey workplace disciplines

Relationship between poka-yoke and OHS leading indicators

The poka-yoke (mistake proofing) methodology attempts to prevent the mistakes and defects

through ingenious devices. This methodology has a close relationship with safety principle.

However, tools and equipment should be inspected periodically to ensure that safety devices

operate correctly. Therefore, the following OHS leading indicators are proposed for this

method.

OHS leading indicators linked to poka-yoke methodology

- Periodically inspect safety devices and poka-yoke devices


- Survey workstations regarding ergonomics risks

- Assess work procedures

- Monitor employees' training in regards to tools and equipment utilization


- Assess workload

Relationship between standard work and OHS leading indicators

The standard work attempts to ensure that activities are within takt time range through calcu-

lating takt time and timing the activities. Further, this methodology strives to ensure that all

employees perform their tasks similarly and, therefore, variation in the work method is re-

duced. Work teams try to specify the exact approach of task performing and then follow it

consistently.

OHS leading indicators linked to standard work methodology

- Assess workload and pressure - Assess occupational stress - Assess work paces - Assess ergonomics risks including repetitive motions due to documentation process in standard work methodology - Assess workstations in regards to safety principles - Monitor employee training -Periodically inspect tools and equipment

87

Appendix B

The Questionnaire

A study on the relationship between lean and occupational health and safety performance

This research studies the relationship between lean implementation and its effects on occupational

health and safety management and performance. The results of the survey will be published in scien-

tific journals and conference proceedings. By completing this questionnaire, you give your consent

to participate. Participation is voluntary. Refusal to participate will involve no penalty or loss of bene-

fits to which you are otherwise entitled, and you may discontinue participation at any time without

penalty or loss of benefits to which you are otherwise entitled. All responses are anonymous, as no

personal information is collected. The survey should take approximately 10 minutes. All respondents

to the survey must be 18 years of age or older. Thank you for participating in this online question-

naire. Should you have any questions about this research project, please feel free to contact Dr.

Beth Cudney at [email protected]. For additional information regarding human participation in re-

search, please feel free to contact the Missouri S&T Campus IRB Chair, Dr. Kathryn Northcut, at

(573)341-6498. By clicking the 'next' button below, you are indicating you have read the information

above and agree to participate.

NEXT

Q1 Please select the category that include your age

Under 18

18 - 24

25 - 34

35 - 44

45 - 54

55 and above

Condition: Under 18 Is Selected. Skip To: End of Survey.

Q2 Please indicate you gender

Male

Female

Prefer not to answer

Q3 What is the highest education that you have achieved?

High school

BS

MS

PhD

Other, please specify ____________________

88

Q4 In which industry do you work?

Accommodation and Food Services

Administrative and Support and Waste Management and Remediation Services

Agriculture, Forestry & Fishing

Arts, Entertainment, and Recreation

Construction

Educational Services

Finance and Insurance

Health Care and Social Assistance

Information

Management of Companies and Enterprises

Manufacturing

Mining

Professional, Scientific, and Technical Services

Public Administration

Real Estate Rental and Leasing

Retail Trade

Transportation and Warehousing

Utilities

Wholesale Trade

Other Services (except Public Administration)

Q5 In which country /region do you work?

All countries were included in the final version.

Q6 How many total employees in your company (all branches)?

Under 49

50 to 499

500 to 4999

5000 or more

89

Q7 What is the annual revenue for your company/organization?

Under $10,000

$10,000 to $49,999

$50,000 to $99,999

$100,000 to $499,999

$500,000 to $999,999

$1,000,000 to $9,999,999

$10,000,000 to $49,999,999

$50,000,000 to $99,999,999

$100,000,000 to $ 1 Billion

Over $ 1 Billion

Don’t know

Q8 Which of the following most accurately describes your primary job function?

Account Management

Administrative

Health Services

Business Development

Clerical, Processing

Creative, Design

Consulting

Customer Service, Support

Distribution

Education

Engineering

Executive Management

Finance

Human resources

Heath, Safety, Environment

Information Systems, Information

Operations/Production

Purchasing

R&D/Scientific

Sales

Other, Please specify ____________________

90

Q9 In your opinion, how thorough have lean practices been implemented in your organization. (Fi-

delity)

How Thorough

To a Great

Extent Somewhat Very Little Not at All Do not know

Quality management pro-

grams (e.g. ISO, QS, EFQM)

Formal continuous im-

provement programs (e.g.

Kaizen)

Visual tools/ management

Standard Work

Level Loading (Heijunka)

5S

Cellular layout

Kanbans (internal)

Bottleneck identification

and removal

Cycle time reduction

Re-engineering processes

Quick changeover tech-

niques/ SMED

Preventive /predictive

maintenance techniques

Job rotation

Problem-solving groups

Flexible/cross-functional

workforce

91

Q10 In your opinion, how wide have lean practices been implemented in your organization. (Exten-

siveness)

How Wide

No Depart-

ments

Some De-

partments

All Depart-

ments Do not know

Quality management programs (e.g.

ISO, QS, EFQM)

Formal continuous improvement pro-

grams (e.g. Kaizen)

Visual tools / management

Standard Work


5S

Cellular layout

Kanbans (internal)

Bottleneck identification and removal



Quick changeover techniques/ SMED

Preventive /predictive maintenance

techniques

Job rotation


Flexible/ cross-functional workforce

92

Q11 How long have the following lean practices been implemented in your organization?

(Experience)

How Long

Never Less than 2

years

Between 2

to 5 years

More than 5

years Do not know

Quality management programs (e.g.

ISO, QS, EFQM)

Formal continuous improvement

programs (e.g. Kaizen)

Visual tools / management

Standard Work


5S

Cellular layout

Kanbans (internal)

Bottleneck identification and re-

moval



Quick changeover techniques/ SMED

Preventive /predictive maintenance

techniques

Job rotation


Flexible/ cross-functional workforce

Q12 Are the lean facilitators in your organization certified?

Yes (1)

No (2)

Q12.1 If yes, which certifications do they hold?

SME Lean certification (1)

ASME Black Belt (2)

ASME Green Belt (3)

ASQ CSSBB (4)

ASQ CSSGB (5)

Other, Please specify (6) ____________________

93

Q13 By implementing lean practices, what effects has your organization experienced regarding the

following issues?

Effects

Worse Same Better Do not know

Exposure to workplace noise

Exposure to vibration

Status of workplace illumination

Exposure to poisoning chemicals

Exposure to flammable and/or explosive chem-icals

Exposure to dust and/or smoke at the work-place

Exposure to biological hazards (e.g. bacteria and viruses)

Awkward/strained positions

Frequent lifting

Repetitive motion

Extensive and frequent force

Motivation for safe working

Employee involvement in creating a safe envi-ronment

Knowledge about safety issues

Risk-taking behavior

Define/clarify job functions

Skills utilization

Employee involvement

Workload and pressure

Work pace

Breaks

Work intensity (responsibility, cognitive de-mands)

Job autonomy

Machinery and tool safety

Job stress

Job safety

Job satisfaction

Time pressure (e.g. deadlines)

Safety culture

Management commitment to safety issues in the workplace

Organization's policy regarding safety issues at workplace

Training on safety and health principles

Teamwork and communication

Employee involvement in improving work methods

Labor-management relations

Reward systems for safety

Workplace health promotion programs

Safety systems (e.g. light curtains, lock out-tag out)

94

Q14 What trend of recordable injuries has your organization experienced directly related to lean im-

plementation?

Increasing

Stable

Decreasing

Do not know

Q15 What trend of worker's compensation costs has your organization experienced directly related

to lean implementation?

Increasing

Stable

Decreasing

Do not know

Q16 What trend of accident records has your organization experienced directly related to lean im-

plementation?

Increasing

Stable

Decreasing

Do not know

Q17 What trend of lost working days has your organization experienced directly related to lean im-

plementation?

Increasing

Stable

Decreasing

Do not know

Q18 What does your company emphasize?

Lean over safety

Safety over lean

Both equally important

Neither emphasized

Other, Please specify ____________________

measuring the impact of lean implementation on ...€¦ · thus, lean philosophy is perceived as...

Documents