1911804

31
The moderating effect of supply chain role on the relationship between supply chain practices and performance An empirical analysis Lori S. Cook and Daniel R. Heiser  DePaul University, Chicago, Illinois, USA, and Kaushik Sengupta  Hofstra University, Hempstead, New York, USA Abstract Purpose  – The purpose of this paper is to examine the relationships between specific supply chain practices and organizational performance and whether this relationship is moderated by the role that a company assumes in its respective supply chain. Design/methodology/approach – This paper uses regression analysis and the relative weights method to analyze a set of survey data from respondents within the non-academic, North American membership of the Institute of Supply Management. Findings – The results show that the supply chain role for a company makes a difference in terms of the specific supply chain practices that lead to better performance. Further, there is a clear indication that the relative importance of a specific practice varies across the supply chain roles thereby indicating that a general link between practice and performance may be erroneous without considering the specific context of the company concerned. Research limitations/implications  – Supply chain practices are complex constructs. While this study shows the effect of broadly-accepted supply chain practices on performance, not all possible practices are covered in the study. Additional practices not considered may have an effect on company performance and future research may improve upon the findings by extending the analysis to include an expanded segmentation of supply chain role. Practical implications  – The results of the study serve as a practical guideline for managers that not all practices would be effective for all companies. Managers must look at the role-specific context of their organization in the supply chain before deciding which practices are likely to be appropriate. Originality/value – This paper expands the current body of research in the supply chain area by examining the supply chain roles of manufacturer, distributor, retailer and service provider. This is a muchbroader constructthan themore commondyadic treatment ofa supply chainconsisting onlyofa customer and supplier, and adds a new contextual dimension to supply chain research. In addition, service provider as a supply chain role has been hardly researched before. Keywords Supply chain management, Organizational performance Paper type Research paper The current issue and full text archive of this journal is available at www.emeraldinsight.com/0960-0035.htm Author last names are in alphabetical order. All authors contributed equally to this research. IJPDLM 41,2 104 Received October 2009 Revised May 2010, August 2010 Accepted August 2010 International Journal of Physical Distribution& LogisticsManagement Vol. 41 No. 2, 2011 pp. 104-134 q Emerald Group Publishing Limited 0960-0035 DOI 10.1108/09600031111118521

Upload: shradha-gawankar

Post on 18-Oct-2015

5 views

Category:

Documents


0 download

DESCRIPTION

article

TRANSCRIPT

  • The moderating effect of supplychain role on the relationshipbetween supply chain practices

    and performanceAn empirical analysis

    Lori S. Cook and Daniel R. HeiserDePaul University, Chicago, Illinois, USA, and

    Kaushik SenguptaHofstra University, Hempstead, New York, USA

    Abstract

    Purpose The purpose of this paper is to examine the relationships between specific supply chainpractices and organizational performance and whether this relationship is moderated by the role that acompany assumes in its respective supply chain.

    Design/methodology/approach This paper uses regression analysis and the relative weightsmethod to analyze a set of survey data from respondents within the non-academic, North Americanmembership of the Institute of Supply Management.

    Findings The results show that the supply chain role for a company makes a difference in terms ofthe specific supply chain practices that lead to better performance. Further, there is a clear indicationthat the relative importance of a specific practice varies across the supply chain roles therebyindicating that a general link between practice and performance may be erroneous without consideringthe specific context of the company concerned.

    Research limitations/implications Supply chain practices are complex constructs. While thisstudy shows the effect of broadly-accepted supply chain practices on performance, not all possiblepractices are covered in the study. Additional practices not considered may have an effect on companyperformance and future research may improve upon the findings by extending the analysis to includean expanded segmentation of supply chain role.

    Practical implications The results of the study serve as a practical guideline for managersthat not all practices would be effective for all companies. Managers must look at the role-specificcontext of their organization in the supply chain before deciding which practices are likely to beappropriate.

    Originality/value This paper expands the current body of research in the supply chain area byexamining the supply chain roles of manufacturer, distributor, retailer and service provider. This is amuch broader construct than the more common dyadic treatment of a supply chain consisting only of acustomer and supplier, and adds a new contextual dimension to supply chain research. In addition,service provider as a supply chain role has been hardly researched before.

    Keywords Supply chain management, Organizational performance

    Paper type Research paper

    The current issue and full text archive of this journal is available at

    www.emeraldinsight.com/0960-0035.htm

    Author last names are in alphabetical order. All authors contributed equally to this research.

    IJPDLM41,2

    104

    Received October 2009Revised May 2010,August 2010Accepted August 2010

    International Journal of PhysicalDistribution & Logistics ManagementVol. 41 No. 2, 2011pp. 104-134q Emerald Group Publishing Limited0960-0035DOI 10.1108/09600031111118521

  • IntroductionOrganizations are faced with an array of challenges as they strive to compete intodays dynamic global markets. To remain competitive, organizations mustrecognize the importance of supply chain practices that improve not only their ownperformance, but also coordinate with their supply chain partners to improve their jointperformance. Yet, despite the significant advances in research and practices, manyorganizations continue to struggle to understand the complex issues associated with thecoordinated planning and supply activities amongst the members of their supplynetworks.

    Supply chain management (SCM) research has evolved to a stage where analyticaland empirical methodologies have allowed researchers to identify and validate basicSCM models and constructs. Numerous studies have also investigated the effects ofvarious SCM related practices affecting organizational performance. As SCM researchcontinues to develop, many researchers are focusing on the cross-industry validity ofprevious findings. One of the aspects of interest is the effect of employing various bestpractices by organizations in different positions of the supply chain. This is a significantissue to address to determine whether commonly advocated practices are equallyrelevant across the length of the supply chain. While a few studies have examined thedifference in effectiveness of SCM practices based on whether these are applied on thesupply side or the distribution side of the supply chain (Frohlich and Westbrook, 2001,2002; Kim, 2006; Li et al., 2005b), most of these studies have treated the supply and thedistribution sides of the supply chain as one overall stage. Therefore, the treatment hasbeen largely based on a dyadic basis. Such an aggregated view of supply chain positionmasks a number of issues, which companies in specific supply chain roles may face.For instance, should distributors and retailers look at supply chain practices the sameway? From the dyadic standpoint, these two types of companies should face the sameissues and supply chain practices adopted for one, should be equally effective for theother. However, this may not be the case since the distributor stage is an intermediatestage in the supply chain while the retailer stage is typically the final stage before thecustomer.

    The strictly dyadic treatment of the supply chain into the supply and distributionsides also prevents the inclusion of other roles, which play a significant role in theeffectiveness of supply chain. For instance, some recent research studies have startedlooking at service supply chains. The aspects of service supply chains emphasize therelative importance of non-physical flows within the supply chain. Prior research onservice aspects in supply chains has been limited, with only a few studies specificallyassessing the importance of services as a separate supply chain related construct (Fieldand Meile, 2008; Sengupta et al., 2006; Ellram et al., 2004).

    Expanding the scope of inquiry to include supply chain position allows us to examinethese additional stages for divergence in the effectiveness of specific SCM practices,which is a significant contribution to the existent body of knowledge. More specifically,the main contributions of this study are twofold:

    (1) it extends the results from previous studies (Frohlich and Westbrook, 2001,2002; Kim, 2006; Li et al., 2005b) which have largely treated supply chains in adyadic manner; and

    (2) it introduces analysis on supply chain roles which have not been examinedextensively in extant research.

    Supply chain role

    105

  • The results of the study not only provide additional insights into effectiveness of SCMpractices from a research standpoint, the results also provide contextual implicationsto practicing managers who are more interested in knowing what specific SCMpractices improve performance for their type of company.

    The remainder of this paper is organized as follows. The next section discusses therelevant literature supporting the proposed research framework, commonly researchedSCM practices and the linkage of practices to performance. Next, the researchhypotheses are proposed and the methodologies used to test them are presented. Finally,we summarize key findings and conclude with a discussion of limitations and futuredirections for research.

    Literature review and research frameworkIn many of the initial SCM research studies, the supply chain was viewed simply as anextension of traditional areas such as operations, purchasing and logistics. This myopicview considered a limited set of organizational stakeholders and issues in the analysis ofSCM practices. However, over the past decade, SCM research has evolved and broadenedto encompass a variety of perspectives such as supplier relationships, supply chainnetwork structure and collaboration (Tan et al., 1999, 2002; Croom et al., 2000; Tan, 2002;Chen and Paulraj, 2004a, b; Chen et al., 2004; Cigolini et al., 2004; Frohlich andWestbrook, 2001, 2002; Ho et al., 2002; Giannakis and Croom, 2004; Lejeune and Yakova,2005; Li et al., 2005b; Kampstra et al., 2006; Sandberg, 2007; Narasimhan et al., 2008).

    Extensive literature reviews have examined a variety of issues pertaining to SCM.Chen and Paulraj (2004a, b) identify the growth of SCM practice from fields such aspurchasing, logistics, operations, organizational theory, information systems andstrategic management. Sachan and Datta (2005) reviewed more than 400 paperspublished in three peer-reviewed journals, Journal of Business Logistics, InternationalJournal of Physical Distribution & Logistics Management, and Supply ChainManagement: An International Journal. While they conclude that survey-basedresearch still ranks as the most widely used methodology in supply chain and logisticsresearch, studies have increasingly adopted more advanced techniques and have beenmoving towards model building and testing. In addition, the authors highlight the needfor looking at all the firms in the value chain as a single entity which partially hints to theeffect of the position of the firms in their respective supply chains and whether this hasoverall effects on strategies and practices. Ho et al. (2002) highlight major conceptualgaps in the current research. They discuss the need to operationalize and model SCMconstructs so they may be viewed from a business process perspective with betterperformance as a desired objective. Their research sought to highlight the need fordefinitive frameworks for managers to put into practice. In a similar manner, Lejeuneand Yakova (2005) propose a typology of supply chain configurations and attempt tolink terms and concepts previously used to separately describe various forms of a supplychain. This approach of developing frameworks is continued in Li et al. (2005b) wherethe authors use an empirical survey to validate six dimensions of SCM practices relatingto strategic supplier partnership, customer relationship, information sharing,information quality, internal lean practices and postponement. Cigolini et al. (2004)propose a normative conceptual framework for SCM practices and a contingency modelto assist managers in matching demand characteristics with supply chain requirements.In the resulting demand-supply matrix, demand was mapped to the stage of the

    IJPDLM41,2

    106

  • product life cycle and the associated supply chain requirements were classified asefficient, lean or quick supply chains. For each normative level of the associated SCMpractices, they proposed management tools and techniques such as just-in-time (JIT),network design, distribution resource planning, transportation optimization, orderingand replenishment policies. Giannakis and Croom (2004) propose a 3S modelemphasizing three dimensions including:

    (1) the synthesis of the business and resources network;

    (2) the synergy between different participants in the network; and

    (3) the synchronization of operational decisions related to the control of theproduction and delivery of goods and services.

    Three levels of analysis, dyadic, chain and network are included in the classificationsystem proposed. Similar to other research studies, common management techniquesare proposed to help improve supply chain operations. Kim (2006) uses results of anempirical survey to develop a framework to assess the SCM integration and SCMpractices and the ensuing linkage to competitive capability. A companys integrationconsiders suppliers, customers and organizational cross-functional integration.A similar perspective regarding knowledge sharing among supply chain partnerswas examined by Wang et al. (2008). They analyze the gap between the theoreticalbenefits of mutual knowledge sharing among supply chain partners and the practicalcomplexity of achieving such a synergy because of various complexities in theinteraction processes between companies. Using a case-based methodology, theydevelop a model to enhance knowledge sharing among supply chain partners. Mentzeret al. (2001) conducted a comprehensive analysis of the various definitions and aspectsunder the umbrella of SCM. They define SCM as a comprehensive model whichencompasses inter-functional and inter-corporate coordination under an envelope ofglobal environment and a multi-tiered supply chain. Clearly, multiple studies haveshown the importance of interaction among the supply chain members as a way toimprove the competency of all supply chain partners.

    While existing studies have extended the theoretical understanding of supply chainrelated issues and have highlighted the need for better interaction among the supplychain members, the question of whether the SCM practices are equally effective acrossall stages of the supply chain has not been explored beyond a few studies. Frohlich andWestbrook (2001, 2002) used an empirical survey to test whether manufacturers shouldhave different integration practices between suppliers and customers. In addition, theydetermined that the extent of integration with the customers and suppliers could bedifferent, with a higher level of integration usually associated with better performance.Although Frohlich and Westbrook (2001, 2002) contributed to understanding thenuances of integration practices, the treatment of suppliers and/or customers did notdifferentiate with respect to the specific position in the supply chain.

    The differences in supply chain and organization performance across various stagesin the supply chain have been implicitly analyzed in a number of other studies.Simulation models were used in Zhao and Xie (2002) and Zhao et al. (2002a) to examinethe impact of forecasting and information sharing in a supply chain. Both studiesconsidered various independent variables such as demand patterns, informationsharing and forecasting related variables. While the dependent variables were totalcosts for the retailers, total cost for the supplier and the total cost for the supply chain.

    Supply chain role

    107

  • In both studies, the results showed that forecasting parameters significantly influencedthe performance of the supply chain and the value of information sharing. In Zhao et al.(2002b), results of another simulation model were presented to show the impact ofinformation sharing and ordering coordination on the performance of a supply chainwith one capacitated supplier and multiple retailers under demand uncertainty. Thismodel simulated inventory replenishment decisions by the retailers and productiondecisions by the supplier under different demand patterns and capacity tightness.Similar to the other studies, the results indicated that information sharing results insignificant benefits to the supplier and the entire supply chain. However, across all threestudies some of the scenarios simulated suggested that the retailers costs actuallyincreased. Therefore, the authors recommend that an incentive should be provided to theretailers to encourage and motivate partners to share information in the supply chain.These studies highlight the potential differences between a supplier and a retailer interms of the benefits of information sharing.

    Singh et al. (2005) in their study of the Australian automotive supply chain highlightimportant issues related to logistics and information sharing. An exploratory study isconducted with vehicle manufacturers, component suppliers, industry bodies andgovernment agencies. Specific issues highlighted include low volume of production,threat of obsolescence, supplier relationship, supplier development, unavailability oflogistics providers and inadequacy of information systems. While several of these issuesrelate to some of the supply chain factors examined in our study, it also highlights theneed for examining the importance of the different factors based on the roles played byspecific companies. For instance, infrastructure issues related to logistics play asignificant role in the suppliers performances while these have a limited effect on themanufacturers. This study illustrates differences in the effectiveness of supply chainpractices based on the specific role of a company in the supply chain. This perspective isalso shown by Lau et al. (2004) where they examine the impact of information sharing oninventory replenishments. In their study, three stages of the supply chain are modelled:manufacturer, distributor and retailer. They conclude that there is a role-based differencein the effectiveness of the replenishment practices. For instance, it is more beneficial tohave information sharing among downstream partners (like retailers and distributors)than upstream partners (like distributors and manufacturers). They also analyzed theseeffects with regard to operating costs and conclude that information sharing initiativesshould be first initiated with downstream partners. Downstream partners hold moreinventory than upstream partners and the cost of backlog is also higher with thedownstream partners. With the upstream partners, the accuracy of demand informationdecreases resulting in any benefits of information sharing with the upstream partners tobe less significant. Therefore, it is more beneficial to initiate information sharing with thedownstream partners. Information sharing with supply chain partners is one of thesupply chain practices examined in our study and Lau et al.s (2004) study shows thatthe effects of information sharing differ by the specific role or position of the company inthe supply chain; the same perspective as the one undertaken in our research.

    In this study, we extend the results from previous research by viewing the supplychain role of a company as a moderating factor in determining the relationship betweenspecific SCM practices and organizational performance. The supply chain role is definedas the position of a company in its supply chain including the roles of: manufacturer,distributor, retailer and service provider. The supply chain role of manufacturer

    IJPDLM41,2

    108

  • includes component suppliers, fabricators and final assemblers of products. The supplychain role of distributor includes wholesalers, distributors and logistics companies. Thesupply chain role of retailer includes organizations which are positioned as the last linkin the supply chain interacting with the final consumer. Finally, the supply chain role ofservice provider includes organizations which provide an array of supporting services tothe other supply chain partners, such as utilities, construction, information services andprofessional services. Analyzing these four supply chain roles as a moderating factorsignificantly expands the results from previous research studies on the effect of SCMpractices on organizational performance. In addition, the inclusion of service providersintroduces a supply chain role that has received sparse coverage in extant research.

    In order to examine the effect of specific SCM practices on organizational performanceand to determine whether the supply chain role moderates this relationship, this studyconsiders commonly advocated SCM practices used in previous research. The intentis not to provide exhaustive coverage of all possible SCM practices, but rather to assessthe relationship between a few commonly advocated practices and the resultingorganizational performance when moderated by the supply chain role. The resultsidentify context-dependent relationships and thereby identify rich pools of inquiry forfuture context-specific analysis.

    The proposed research framework, shown in Figure 1, illustrates the specific SCMpractices examined in this study, their proposed relationship to organization performanceand the proposed moderating effect of the four supply chain roles on this relationship.

    One of the objectives of this study is to exploring the relationship between commonsupply chain practices and organizational performance. We assume that a higher level ofadoption of a supply chain practice has a positive relationship to organizationalperformance. Since we examine six specific supply chain practices, the following sixhypotheses are proposed:

    H1-H6. A higher level of adoption of the six supply chain practices:(a) SHARE, (b) RELATION, (c) PLAN, (d) INTERNET, (e) SOURCE and(f) DISTRIBUTION is positively related to organizational performance.

    The primary objective of this research is to examine whether the supply chain role of anorganization (manufacturer, distributor, retailer or service provider) has a moderating

    Figure 1.Research framework

    Dependent variableOrganizationalperformance

    Moderating factorSupply chain roles

    ManufacturerDistributorRetailerService provider

    H1-H6

    H7-H10Independent variables

    SCM PracticesShare-Information sharingRelation-Long term relationshipsPlan-Advanced planning systemsInternet-Leveraging the internetSource-Supply network structureDistribution-Distribution network structure

    Supply chain role

    109

  • effect on the positive relationship between the SCM practices and organizationalperformance. Accordingly, the following additional hypotheses are proposed:

    H7. The supply chain role of manufacturer in an organization moderates thepositive relationship between the SCM practices: (a) SHARE, (b) RELATION,(c) PLAN, (d) INTERNET, (e) SOURCE and (f) DISTRIBUTION andorganizational performance.

    H8. The supply chain role of distributor in an organization moderates the positiverelationship between the SCM practices: (a) SHARE, (b) RELATION,(c) PLAN, (d) INTERNET, (e) SOURCE and (f) DISTRIBUTION andorganizational performance.

    H9. The supply chain role of retailer in an organization moderates the positiverelationship between the SCM practices: (a) SHARE, (b) RELATION,(c) PLAN, (d) INTERNET, (e) SOURCE and (f) DISTRIBUTION andorganizational performance.

    H10. The supply chain role of service provider in an organization moderates thepositive relationship between the SCM practices: (a) SHARE, (b) RELATION,(c) PLAN, (d) INTERNET, (e) SOURCE and (f) DISTRIBUTION andorganizational performance.

    It is important to note that hypotheses H1 through H6 are treated as the first step in theanalysis and we do not draw conclusive suggestions or managerial implications fromthe associated results. Hence, results from H1 through H6 are used as the basis foranalyzing the moderating effect of the supply chain role indicated in subsequent H7through H10. The same positive relationships between a higher adoption level on thesupply chain practices and organizational performance, as stated in H1 through H6, isalso assumed under H7 through H10. Therefore, when drawing conclusions andmanagerial implications of the study, the discussion of the results combines the two setsof hypotheses.

    Supply chain practicesThis study considers some of the more widely advocated SCM practices examined inprior SCM literature. These include information sharing between supply chain partners(SHARE), building long-term relationships between supply chain partners(RELATION), using advanced planning systems (PLAN), leveraging internet-basedtechnology (INTERNET), design of the supplier network (SOURCE) and design of thedistribution network (DISTRIBUTION). The overall objective is to consider arepresentative set of common SCM practices rather than an exhaustive list of all suchpractices. Table I displays a taxonomy of the literature used to create the specific items tomeasure each of the six SCM practices. Based on prior research studies, three to six suchitems were commonly used to evaluate the level of adoption of the six practices byrespondent organizations.

    Information sharing (SHARE)Information sharing (SHARE) with supply chain partners has been an area of focusin SCM research primarily due to the perceived benefits of enhanced coordination.Prior research has examined the scope and process of information sharing related

    IJPDLM41,2

    110

  • SC

    Mp

    ract

    ices

    /org

    aniz

    atio

    nal

    per

    form

    ance

    Su

    rvey

    item

    sS

    urv

    eyst

    atem

    ents

    Rel

    ated

    stat

    emen

    tsor

    con

    stru

    cts

    from

    pas

    tre

    sear

    chR

    efer

    ence

    s

    Info

    rmat

    ion

    shar

    ing

    (SH

    AR

    E)

    SH

    AR

    E1

    We

    shar

    ein

    form

    atio

    non

    inv

    ento

    ryle

    vel

    sw

    ith

    our

    sup

    ply

    chai

    np

    artn

    ers

    Deg

    ree

    tow

    hic

    hre

    tail

    ers

    shar

    efo

    reca

    stan

    dot

    her

    info

    rmat

    ion

    wit

    hsu

    pp

    lier

    s

    Hsu

    etal.

    (200

    8),

    Zh

    aoetal.

    (200

    2a,b

    )

    SH

    AR

    E2

    We

    shar

    efo

    reca

    sts

    ofcu

    stom

    erd

    eman

    dw

    ith

    our

    sup

    ply

    chai

    np

    artn

    ers

    Ex

    ten

    tto

    wh

    ich

    pro

    pri

    etar

    yan

    dcr

    itic

    alin

    form

    atio

    nis

    com

    mu

    nic

    ated

    toon

    es

    sup

    ply

    chai

    np

    artn

    er

    Lietal.

    (200

    5)

    SH

    AR

    E3

    We

    shar

    ein

    form

    atio

    non

    pri

    cep

    rom

    otio

    ns

    wit

    hou

    rsu

    pp

    lych

    ain

    par

    tner

    s

    Info

    rmat

    ion

    exch

    ang

    ew

    ith

    sup

    pli

    ers

    Hsu

    etal.

    (200

    8),

    Kim

    (200

    6)

    SH

    AR

    E4

    We

    shar

    ein

    form

    atio

    nel

    ectr

    onic

    ally

    wit

    hou

    rsu

    pp

    lych

    ain

    par

    tner

    s

    Sh

    arin

    gd

    iffe

    ren

    tin

    form

    atio

    nw

    ith

    sup

    pli

    ers

    and

    cust

    omer

    s(i

    nte

    gra

    tiv

    eac

    tiv

    itie

    s)

    Fro

    hli

    chan

    dW

    estb

    rook

    (200

    1)

    Com

    mu

    nic

    atio

    nF

    ield

    and

    Mei

    le(2

    008)

    Ch

    enan

    dP

    aulr

    aj(2

    004a

    ,b

    )V

    alu

    eof

    info

    rmat

    ion

    shar

    ing

    Fie

    ldan

    dM

    eile

    (200

    8)L

    ietal.

    (200

    5)(S

    CM

    :IJ)

    Lon

    g-t

    erm

    rela

    tion

    ship

    s(R

    EL

    AT

    ION

    )R

    EL

    AT

    ION

    1W

    ech

    oose

    sup

    pli

    ers

    bas

    edu

    pon

    thei

    rfl

    exib

    ilit

    yan

    dsp

    eed

    ofd

    eliv

    ery

    Str

    ateg

    icsu

    pp

    lier

    par

    tner

    ship

    defi

    ned

    aslo

    ng

    -ter

    mre

    lati

    onsh

    ipb

    etw

    een

    anor

    gan

    izat

    ion

    and

    its

    sup

    pli

    ers

    Lietal.

    (200

    5)

    RE

    LA

    TIO

    N2

    We

    bu

    ild

    lon

    g-t

    erm

    ,m

    utu

    ally

    ben

    efici

    alre

    lati

    onsh

    ips

    wit

    hk

    eysu

    pp

    lier

    s

    Lev

    elof

    stra

    teg

    icp

    artn

    ersh

    ipw

    ith

    sup

    pli

    ers

    Fie

    ldan

    dM

    eile

    (200

    8),

    Kim

    (200

    6)

    RE

    LA

    TIO

    N3

    We

    neg

    otia

    telo

    ng

    -ter

    mco

    ntr

    acts

    wit

    hou

    rsu

    pp

    lier

    sL

    ong

    -ter

    mp

    ersp

    ecti

    ve

    for

    sou

    rcin

    gp

    olic

    yD

    eT

    oni

    and

    Nas

    sim

    ben

    i(1

    999)

    Lon

    g-t

    erm

    rela

    tion

    ship

    and

    rela

    tion

    ship

    clos

    enes

    sC

    hoi

    and

    Har

    tley

    (199

    6)

    Lon

    g-t

    erm

    rela

    tion

    ship

    Ch

    enan

    dP

    aulr

    aj(2

    004a

    ,b

    )A

    dv

    ance

    dp

    lan

    nin

    gsy

    stem

    s(P

    LA

    N)

    PL

    AN

    1M

    anag

    ing

    raw

    mat

    eria

    lan

    dfi

    nis

    hed

    goo

    din

    ven

    tori

    esA

    dv

    ance

    dm

    anag

    emen

    tan

    dm

    anu

    fact

    uri

    ng

    tech

    nol

    ogy

    Kim

    (200

    6)

    (continued

    )

    Table I.Taxonomy of selected

    literature for the six SCMpractices and

    organizationalperformance

    Supply chain role

    111

  • SC

    Mp

    ract

    ices

    /org

    aniz

    atio

    nal

    per

    form

    ance

    Su

    rvey

    item

    sS

    urv

    eyst

    atem

    ents

    Rel

    ated

    stat

    emen

    tsor

    con

    stru

    cts

    from

    pas

    tre

    sear

    chR

    efer

    ence

    s

    PL

    AN

    2M

    anag

    ing

    wor

    k-i

    n-p

    roce

    ssin

    ven

    tori

    esR

    eal-

    tim

    ese

    arch

    ing

    ofin

    ven

    tori

    esK

    im(2

    006)

    PL

    AN

    3U

    sin

    gm

    ater

    ial

    req

    uir

    emen

    tsp

    lan

    nin

    g(M

    RP

    )sy

    stem

    sIn

    teg

    rati

    ve

    inv

    ento

    rym

    anag

    emen

    tK

    im(2

    006)

    PL

    AN

    4U

    sin

    gE

    RP

    syst

    ems

    Deg

    ree

    ofIT

    adop

    tion

    (ten

    item

    sre

    late

    dto

    ED

    I,E

    RP

    ,re

    al-t

    ime

    info

    rmat

    ion

    syst

    ems)

    Ak

    yu

    zan

    dR

    ehan

    (200

    9),

    Jin

    (200

    6)

    PL

    AN

    5U

    sin

    gco

    llab

    orat

    ive

    pla

    nn

    ing

    ,fo

    reca

    stin

    g,

    and

    rep

    len

    ish

    men

    t(C

    PF

    R)

    Info

    rmat

    ion

    tech

    nol

    ogy

    Ch

    enan

    dP

    aulr

    aj(2

    004a

    ,b

    )

    Au

    tom

    ated

    syst

    ems

    (on

    -lin

    ean

    din

    teg

    rate

    d)

    Cig

    olin

    ietal.

    (200

    4)

    PL

    AN

    6U

    sin

    gac

    tiv

    ity

    -bas

    edco

    stin

    g(A

    BC

    )ac

    cou

    nti

    ng

    met

    hod

    sO

    per

    atio

    ns-

    orie

    nte

    dap

    pli

    cati

    ons

    (in

    clu

    din

    gE

    RP

    ,C

    PF

    R)

    San

    der

    san

    dP

    rem

    us

    (200

    2)

    Lev

    erag

    ing

    the

    inte

    rnet

    (IN

    TE

    RN

    ET

    )IN

    TE

    RN

    ET

    1S

    har

    ing

    info

    rmat

    ion

    over

    the

    inte

    rnet

    wit

    hsu

    pp

    lych

    ain

    par

    tner

    s

    Su

    pp

    ly-s

    ide

    inte

    gra

    tion

    pra

    ctic

    esu

    sin

    gw

    ebte

    chn

    olog

    y(i

    ncl

    ud

    esin

    form

    atio

    nsh

    arin

    g,

    dem

    and

    and

    inv

    ento

    rytr

    ack

    ing

    ,or

    der

    man

    agem

    ent

    via

    the

    inte

    rnet

    )

    Gim

    enez

    and

    Lou

    ren

    co(2

    008)

    Fro

    hli

    chan

    dW

    estb

    rook

    (200

    2)G

    imen

    ezan

    dL

    oure

    nco

    (200

    9)

    INT

    ER

    NE

    T2

    Pu

    rch

    asin

    gm

    ater

    ial

    and

    serv

    ices

    via

    the

    inte

    rnet

    Cu

    stom

    er-s

    ide

    inte

    gra

    tion

    pra

    ctic

    esu

    sin

    gw

    ebte

    chn

    olog

    y(i

    ncl

    ud

    esor

    der

    man

    agem

    ent,

    sell

    ing

    and

    dem

    and

    fore

    cast

    ing

    via

    the

    inte

    rnet

    )

    Fro

    hli

    chan

    dW

    estb

    rook

    (200

    2)G

    imen

    ezan

    dL

    oure

    nco

    (200

    9)

    INT

    ER

    NE

    T3

    Sel

    lin

    gp

    rod

    uct

    san

    dse

    rvic

    esv

    iath

    ein

    tern

    etW

    eb-b

    ased

    mar

    ket

    ing

    orie

    nte

    dap

    pli

    cati

    ons

    San

    der

    san

    dP

    rem

    us

    (200

    2)

    (continued

    )

    Table I.

    IJPDLM41,2

    112

  • SC

    Mp

    ract

    ices

    /org

    aniz

    atio

    nal

    per

    form

    ance

    Su

    rvey

    item

    sS

    urv

    eyst

    atem

    ents

    Rel

    ated

    stat

    emen

    tsor

    con

    stru

    cts

    from

    pas

    tre

    sear

    chR

    efer

    ence

    s

    Su

    pp

    lyn

    etw

    ork

    stru

    ctu

    re(S

    OU

    RC

    E)

    SO

    UR

    CE

    1D

    ecid

    ing

    wh

    eth

    er,

    and

    how

    mu

    ch,

    toou

    tsou

    rce

    Log

    isti

    csin

    itia

    tiv

    ean

    dlo

    cati

    onp

    olic

    ies

    Kim

    (200

    6)

    SO

    UR

    CE

    2S

    elec

    tin

    gan

    dce

    rtif

    yin

    gsu

    pp

    lier

    sS

    up

    pli

    erse

    lect

    ion

    ,su

    pp

    lyb

    ase

    red

    uct

    ion

    ,si

    ng

    leso

    urc

    ing

    De

    Ton

    ian

    dN

    assi

    mb

    eni

    (199

    9)

    SO

    UR

    CE

    3R

    atio

    nal

    izin

    gth

    esu

    pp

    lyb

    ase

    (e.g

    .st

    rate

    gic

    par

    tner

    ing

    ,v

    erti

    cal

    inte

    gra

    tion

    and

    sin

    gle

    sou

    rce

    sup

    ply

    )

    Ex

    ten

    ded

    sup

    ply

    net

    wor

    kC

    hoi

    and

    Wu

    (200

    9)

    Ver

    tica

    lin

    teg

    rati

    onC

    hoi

    and

    Kim

    (200

    8)In

    vol

    vem

    ent

    insu

    pp

    lier

    dec

    isio

    ns

    Sez

    en(2

    008)

    Su

    pp

    lyn

    etw

    ork

    stru

    ctu

    re;

    sup

    ply

    bas

    ere

    du

    ctio

    nC

    hen

    and

    Pau

    lraj

    (200

    4a,

    b)

    Dis

    trib

    uti

    onn

    etw

    ork

    stru

    ctu

    re(D

    IST

    RIB

    UT

    ION

    )D

    IST

    RIB

    UT

    ION

    1D

    ecid

    ing

    wh

    ere

    tolo

    cate

    faci

    liti

    esL

    ogis

    tics

    init

    iati

    ve

    and

    loca

    tion

    pol

    icie

    sK

    im(2

    006)

    DIS

    TR

    IBU

    TIO

    N2

    Dec

    idin

    gw

    her

    eto

    hol

    din

    ven

    tory

    ina

    dis

    trib

    uti

    onn

    etw

    ork

    Insi

    de-

    out

    cap

    abil

    itie

    s(i

    ncl

    ud

    esou

    tbou

    nd

    tran

    spor

    tati

    on,

    war

    ehou

    sin

    gan

    din

    ven

    tory

    pol

    icie

    s)

    Tra

    ceyetal.

    (200

    5)

    DIS

    TR

    IBU

    TIO

    N3

    Ch

    oosi

    ng

    bet

    wee

    nd

    iffe

    ren

    ttr

    ansp

    orta

    tion

    and

    dis

    trib

    uti

    onm

    odes

    Inte

    gra

    tiv

    eac

    tiv

    itie

    sw

    ith

    cust

    omer

    sF

    roh

    lich

    and

    Wes

    tbro

    ok(2

    001)

    War

    ehou

    sen

    etw

    ork

    red

    esig

    n,

    tran

    spor

    tati

    onop

    tim

    izat

    ion

    Cig

    olin

    ietal.

    (200

    4)

    (continued

    )

    Table I.

    Supply chain role

    113

  • SC

    Mp

    ract

    ices

    /org

    aniz

    atio

    nal

    per

    form

    ance

    Su

    rvey

    item

    sS

    urv

    eyst

    atem

    ents

    Rel

    ated

    stat

    emen

    tsor

    con

    stru

    cts

    from

    pas

    tre

    sear

    chR

    efer

    ence

    s

    Org

    aniz

    atio

    nal

    per

    form

    ance

    OR

    GP

    ER

    F1

    We

    del

    iver

    our

    pro

    du

    cts

    and

    serv

    ices

    fast

    erth

    anou

    rco

    mp

    etit

    ors

    Tim

    eto

    mar

    ket

    mea

    sure

    Lietal.

    (200

    5b)

    OR

    GP

    ER

    F2

    Ou

    ron

    -tim

    ed

    eliv

    ery

    per

    form

    ance

    isb

    ette

    rth

    anou

    rco

    mp

    etit

    ors

    Del

    iver

    yd

    epen

    dab

    ilit

    yD

    eliv

    ery

    fact

    orL

    ead

    -tim

    ep

    erfo

    rman

    ce

    Lietal.

    (200

    5b),

    Ver

    eeck

    ean

    dM

    uy

    lle

    (200

    6)Ji

    n(2

    006)

    OR

    GP

    ER

    F3

    Ou

    rp

    rod

    uct

    and

    serv

    ice

    qu

    alit

    yis

    bet

    ter

    than

    our

    com

    pet

    itor

    sQ

    ual

    ity

    fact

    orQ

    ual

    ity

    dim

    ensi

    onV

    eree

    cke

    and

    Mu

    yll

    e(2

    006)

    Tan

    etal.

    (200

    2)O

    RG

    PE

    RF

    4O

    ur

    oper

    atin

    gco

    sts

    are

    low

    erth

    anou

    rco

    mp

    etit

    ors

    Cos

    tfa

    ctor

    Ver

    eeck

    ean

    dM

    uy

    lle

    (200

    6)

    OR

    GP

    ER

    F5

    Ou

    rop

    erat

    ing

    pro

    fits

    are

    hig

    her

    than

    our

    com

    pet

    itor

    sP

    rofi

    tab

    ilit

    yR

    etu

    rnon

    sale

    sN

    aras

    imh

    anetal.

    (200

    2),

    Nar

    asim

    han

    and

    Kim

    (200

    2)Ji

    n(2

    006)

    Ret

    urn

    onsa

    les

    Table I.

    IJPDLM41,2

    114

  • to operational issues such as inventory, forecasting, orders and production plan issues(Chen and Paulraj, 2004a; Kim, 2006; Lee and Whang, 1999; Li et al., 2005a, b; Zhao andXie, 2002; Zhao et al., 2002a,b; Narasimhan et al., 2008; Hsu et al., 2008; Sezen, 2008). Mostof the research in this area is based on a simplified model of a supply chain consisting oftwo stages, single product or a limited number of supply chain partners. Some of themore recent studies have examined other issues with information sharing including thepower of the relationship between the supply chain partners (Williams and Moore, 2007)and the extent of value realized by different partners within a supply chain (Lumsdenand Mirzabeiki, 2008). The latter study uses published papers and textbooks, andadditional interviews with practitioners in different segments of the supply chain todetermine that different type of information are valued differently by the supply chainpartners. In this research, we explore a similar need that the specific supply chain role ofa company may lead to different perspectives on information sharing and we use a morerobust definition of information sharing by including factors related to inventory levels,demand forecasts and pricing information. Specific items examined under this factorinclude information sharing on inventory levels, forecasts and price promotions; and theextent to which information is shared electronically with the supply chain partners.These modifications were appropriate to make the survey items more relevant to all ofthe manufacturing and service-oriented supply chain roles considered in this researchand are similar to those studied in previous research dealing with information sharingamong supply chain partners in both manufacturing and service areas (Field and Meile,2008; Sezen, 2008; Hsu et al., 2008; Frohlich and Westbrook, 2001).

    Long-term relationships (RELATION)Building long-term relationships (RELATION) within a supply chain allows forfamiliarity and the opportunity for mutual benefit that correspond with a greater level ofcoordination in business decisions (Hahn et al., 1983; De Toni and Nassimbeni, 1999;Choi and Hartley, 1996). Maintaining extended relationships with partner companies is apillar of JIT principles. Prior studies have assessed long-term relationships from severalperspectives including strategic supplier partnership, long-term sourcing policy andgeneral structures of long-term relationships (Chen and Paulraj, 2004a, b; De Toni andNassimbeni, 1999; Choi and Hartley, 1996; Kim, 2006; Lee and Whang, 1999; Li et al.,2005b; Zhao et al., 2002b).

    Advanced planning systems (PLAN)The high adoption patterns of sophisticated advanced planning systems (PLAN)suggest these systems are commonly advocated tools for improving supplychain-related efficiencies. Companies such as SAP, Oracle and others have beendeveloping state-of-the-art systems over the last decade. Prior to the year 2000 (Y2K),advanced planning and inventory control systems were perceived by many to be a coretool of SCM. More recently, the role of enterprise resources planning (ERP) systems inintegrating business processes was found to be relevant and important to remaincompetitive in a web-based environment (Akyuz and Rehan, 2009; Narasimhan et al.,2008; Auramo et al., 2005). However, in the post-Y2K environment and with thedownturn in technology-related equities, the central role of such advanced planningsystems has evolved to be less certain in non-internet dependent environments. In anycase, it is without dispute that the implementation of advanced planning systems

    Supply chain role

    115

  • represents a strategic decision requiring a large investment of capital and managerialfocus. In addition, with greater coordination between supply chain partners comes theneed for applying the advanced planning systems (which were largely developed andinstituted internally by companies) to an inter-firm environment ( Jonsson et al., 2007).Simultaneously, companies need to determine the purpose of using these systems froman inter-firm perspective and they need to identify the key drivers for such systems(Karkkainen et al., 2007). As a result, the costs and benefits of these investments are nowbeing weighed more carefully by managers. This factor has been considered in otherSCM frameworks in a variety of forms including use of advanced management andmanufacturing technology, integrative inventory management systems, degree of ITadoption, scope of automated systems and operations-oriented applications (Chen andPaulraj, 2004a, b; Cigolini et al., 2004; Jin, 2006; Kim, 2006; Sanders and Premus, 2002),and our survey instrument items related to this practice includes a combination of thespecific aspects of the practices examined in previous studies.

    Leveraging the internet ( INTERNET)Even beyond its role in advanced planning systems, the evolution and leveraging of theinternet (INTERNET) has served as one of the main technological developmentssupporting increased collaboration and coordination among supply chain partners.In recent years, companies have continued to adopt internet-based collaboration to makeeffective decisions concerning forecasts, inventory and orders. Multiple studies over thelast few years have examined the impact of the internet on the different processes thatconstitute SCM. Gimenez and Lourenco (2008) conducted a literature review based studyfor the period 1995-2005. They found that the effect of internet on SCM has beenrecognized as an important topic of research with information flows, e-procurement ande-fulfilment being the main areas of research. Therefore, it is important to analyze thisfactors unique role as a supply chain practice in affecting organizational performance.While Frohlich and Westbrook (2002) also examined this factor, they only differentiatedbetween manufacturing and service-oriented organizations. The broader definition ofsupply chain roles in our study allows a deeper understanding of this factor and itsinfluence on organizational performance. The growth in internet usage and this factorsimportance in supply chain research has been reflected in other studies throughsupply-side integration practices using web technology, customer-side integrationpractices using web technology and web-based marketing-oriented applications(Gimenez and Lourenco, 2008; Sanders and Premus, 2002).

    Supply network structure (SOURCE)The design of the supply network structure (SOURCE) refers to practices related to theupstream supply chain of a company. Typically, network structure implies the numberof suppliers and the number of stages in the supply chain (Frohlich and Westbrook,2001; Li et al., 2005b). While Frohlich and Westbrook (2001) consider the extent ofintegration with suppliers and customers, they do not consider some of the factors thatmay affect the nature and extent of integration. Tracey et al. (2005) consider thesupply-side structure in their SC framework through the outside-in capabilities. Otherstudies have also considered logistics initiatives, location policies, supplier selection,supply base reduction, sourcing issues and the network structure (Chen and Paulraj,2004a, b; De Toni and Nassimbeni, 1999; Kim, 2006; Sezen, 2008). Likewise, recent studies

    IJPDLM41,2

    116

  • have also looked at the aspect of the supplier network from an extended networkstandpoint (Choi and Wu, 2009; Choi and Kim, 2008; Narasimhan et al., 2008). Thesestudies have attempted to extend the dyadic buyer-supplier relationship to a moreextended network on a triadic mode where the suppliers suppliers are also included inthe analysis. Continuing this theme, this research assumes the supply network structurefocuses on the upstream portion of the supply chain and includes specific items related tosupplier relations and outsourcing. The items could be important factors when oneconsiders both the dyadic network and also the extended supplier network.

    Distribution network structure (DISTRIBUTION)The design of the distribution network structure (DISTRIBUTION) has been examinedin a variety of frameworks and directly relates to the downstream supply chain for acompany (Frohlich and Westbrook, 2001; Kim, 2006; Li et al., 2005b; Tracey et al., 2005).Some frameworks have focused on customer-related practices, while others haveemphasized distribution in the downstream portion of the supply chain. Kim (2006)proposes a construct which considers the organizations integration with its customers.Tracey et al. (2006) incorporate distributors through the inside-out capability construct.Other studies have also considered logistics initiatives, location policies, integrativeactivities with customers, warehouse network redesign and transportation optimization(Cigolini et al., 2004; Frohlich and Westbrook, 2001; Kim, 2006; Tracey et al., 2005).Shang et al. (2009) apply a mixed integer model to not only reduce overall costs but alsoto improve service levels for Glaxo Smithkline and Beecham. Their study shows thataspects related to facilities location and choosing the most effective distribution networkis still relevant to companies. Based on these studies, our research uses a generalizeddefinition of the distribution network construct (DISTRIBUTION) and encompassesfactors such as facility location, inventory positioning and the choice of alternativedistribution channels.

    Organizational performanceAlthough prior research suggests there is a direct link between the level of adoption ofSCM practices and organizational performance, there have been various definitions oforganizational performance, with some studies emphasizing operational measures,while others stressing financial measures. For example, Li et al. (2005b) use deliverydependability and time to market as performance measures, while firm performancedefined by sales growth, market share growth and profitability are used in other studies(Narasimhan et al., 2008; Narasimhan and Kim, 2002). Many studies have selected acombination of pertinent operational and financial measures to reflect overallorganizational performance. For example, Vereecke and Muylle (2006) use factoranalysis to extract five components of performance related to delivery, cost, flexibility,procurement and quality. Tracey et al. (2005) measure performance through fourseparate dimensions including perceived value, customer loyalty, market performanceand financial performance. Similarly, Tan et al. (2002) use six items for performanceincluding product quality, customer service, competitive position, market share,average selling price and return on assets. Chen and Paulraj (2004a, b) use supplierperformance and buyer performance to assess the financial performance of the buyingfirm. Vickery et al. (2003) use customer service performance followed by financialperformance as the performance constructs. Finally, Jin (2006) assesses operational

    Supply chain role

    117

  • performance via three levels of performance criteria: strategic, operational and financial.Strategic performance is measured by market share and sales growth, operationalperformance is measured by lead-time performance and financial performance isassessed through return on investments and return on sales.

    Based on other SCM studies, Table I displays how this research defines organizationalperformance as a combination of operational and financial results as measured bythe respondents perceived performance relative to their competitors. The operationalmeasures selected are commonly used to assess operational excellence and measures acompanys relative performance with its main competitors on three competitive priorities:speed, delivery and quality. Financial performance is measured by the companys cost-and profit-related performance as compared to their direct competitors. These metrics arecommonly used to assess an organizations financial performance. The performancemeasures in this study were chosen for their applicability across a broad spectrum ofindustries. Given the diversity of the respondents in this study, the dependent variablewas designed to capture evidence of an organizations perceived performance relative totheir direct competitors to avoid confounding results with disparate inter-industrystandards of performance. Similar methods have been used by several other studies(Tan et al., 1999; Gunasekaran et al., 2001; Sanders and Premus, 2002; Tan, 2002; Lockamyand McCormack, 2004). For instance, Tan (2002) proposes that due to a lack of consensusregarding a valid cross-industry measure of corporate performance, performance can beoperationalized by managements perceptions of a firms performance in comparison tothat of major competitors.

    Research methodologyThis study is part of a larger research project exploring supply chain-related practices,their relevance to managers and their impact on firm performance. This arm of the studycompares common SCM practices among working managers across supply chain roles.The respondents were drawn from the non-academic, North American membership ofthe ISM. Data collection for this study included several key items of interest for this studyincluding the primary industrial sector, supply chain role, company demographics, SCMpractices and organizational performance. All items were measured on a seven-pointLikert scale with higher scores indicating a higher level of respondent agreement that theitem accurately describes their organization.

    The content validity of each construct was ensured through pre-testing of thequestionnaire and structured interviews with academic experts and managers in the field.A two-step process was used to develop and refine the survey instrument. In the first step,a panel of SCM academic experts examined the questionnaire items to check for relevancyor possible ambiguity in the wording of specific items. There were six academic expertswho participated in this step. In the second step, a panel of supply chain professionalscompleted the survey. Subsequent interviews with these individuals assessed whetherthey found any non-relevant or ambiguous items. The supply chain professionals weredrawn from two separate samples from two different graduate MBA classes in an urbanuniversity in the USA. The sample was chosen such that all participants were experienced,working professionals in various functions associated with their respective companiespurchasing, operations or distribution functions. Feedback from this two-step processresulted in minor changes to the survey instrument.

    IJPDLM41,2

    118

  • The initial mailing consisted of 666 instruments. Each mailing of the surveyinstrument included a cover letter stating the purpose of the overall study. Several stepswere taken to maximize the response rate, including the inclusion of a postage-paidbusiness reply envelope, a financial incentive to complete the survey and the use of afollow-up letter to non-respondents. All of these steps are considered effective ways toincrease response rates in other operations management research studies (Frohlich,2002). Ten survey packets were returned by the postal service as undeliverable.Approximately, 15 percent of the targeted recipients replied within the first four weeks.After the second mailing, a total of 161 useable surveys were received, which representsa response rate of 24.2 percent. For this research, another 16 surveys were eliminated dueto incomplete responses, leaving 145 responses (21.8 percent) as the basis for thisanalysis. The possibility of non-response bias was investigated through a series oft-tests comparing the responses from the first and second mailing. The t-tests yielded nostatistically significant differences between the early and late responders suggestingthat non-response bias was not an area of concern in this study (Tan, 2002; Lambert andHarrington, 1990; Armstrong and Overton, 1977). Factor analysis was then conductedon the data using the extraction method of principal component analysis followed by avarimax rotation (Tan, 2002).

    The data were analyzed using hierarchical regression and the relative weights (RWs)technique developed by Johnson (2000). The primary distinction between the methods isthat stepwise regression introduces independent variables into a regression equationbased upon their additional contribution to explained variance. In contrast, the RWstechnique examines the multi-linear relationships with all of the proffered independentvariables in the equation. The RWs technique then parses the unique contribution ofeach independent variable to the observed variance. To facilitate comparison betweenthe methods, we also report the explained variance (R 2) for a standard multipleregression with all variables introduced along with the results of the RWs technique.

    The application of the RWs technique is appropriate when simultaneously examiningthe effect of multiple SCM initiatives on organizational performance. This is because it iscommon for firms to pursue multiple SCM-related practices at the same time. As a result,there is a possibility that the overlapping effects and interactions of the relationships ofinterest will affect the results of the stepwise regression. The RW technique is robustagainst the presence of multi-collinearity between the independent variables. Under suchconditions of potential multi-collinearity, the RW method is suggested for scenarioswhen there is no inherent ordering of the predictors and the researcher is interested in therelative contribution each variable makes to the prediction of a dependent variable( Johnson, 2000). It should also be noted that when using multiple regression analysis, thestandardized regression coefficients are often assumed to suggest the relative importanceof the individual independent variables. However, these coefficients are only useful in thisregard when there are no significant correlations between the predictors. In the presenceof multi-collinearity, using standardized regression coefficients to explain the relativeimportance of the predictors may lead to erroneous interpretation (Johnson, 2000).

    Johnson (2000) and Johnson and LeBreton (2004) provide detailed explanations of theRW indices and the associated calculations. They have also made the SPSS syntax file,RWEIGHT, publicly available to compute RWs (Johnson, 2001a). Subsequently,the RWs technique has been used in a variety organizational settings and researchstudies (Cochran, 1999; Johnson, 2000, 2001b; Lievens et al., 2003).

    Supply chain role

    119

  • Results and discussionTable II displays the profile of the respondents by supply chain role, number of full-timeemployees and annual sales volume. Similar organizational dimensions have beenreported and used in other SCM studies (Li et al., 2005b; Tan et al., 2002). The externalvalidity of the collected data was considered by testing the data for normality to assesswhether a bias towards a specific company size or annual sales revenue was present.The results show there is considerable variation and lack of bias in the data thusallowing the results to be generalizable.

    Although the survey items are oriented towards measuring the key supply chainpractices examined in the study, we needed to first analyze whether these statementsactually reflect a unitary construct. In other words, multiple statements related to aspecific strategic factor should generate relatively consistent responses from thecompanies. For example, in the case of the information sharing (SHARE) factor, it isexpected that a company that shares information electronically with its supply chainpartners, will also share information regarding inventory levels, demand and pricepromotions. Although some variation may be expected across the four items, the moreconsistent the responses are, the more likely the factor actually measures a cohesiveconcept. In order to achieve this, a factor analysis was conducted on the survey items withfactors extracted using the principal component analysis followed by a varimax rotation.

    The main criteria used to decide which factors would be used for further analysiswere that total variance explained had to be greater than 60 percent for all factorscombined; there should be a minimum of three variables per factor; factor loadings(eigenvalues) for each of the variables for a factor should be at least 0.45 (which is higherthan the normally accepted level of factor loads of 0.3 (Hair et al., 1998)); and theCronbachs alpha for each factor had to be at least 0.5 (Hair et al., 1998).

    Table III displays the results of the factor analysis including the survey items thatload on each factor, the factor loadings (eigenvalues) and the value of the reliabilitycoefficient or the Cronbachs alpha. The specific SCM practices identified have a totalvariance explained greater than 60 percent and the factor loadings (eigenvalues) for eachof the items is at least 0.525. This threshold value corresponds to a 0.05 level of

    Dimensions Levels Retailer DistributorServiceprovider Manufacturer Total

    Number of full-timeemployees

    1-500 employees 3 16 17 29 65 (44.8%)500-5,000employees

    4 8 12 24 48 (33.1%)

    Greater than5,000 employees

    11 4 10 7 32 (22.1%)

    Annual sales volume Less than $10million

    1 1 7 6 15 (10.3%)

    $10-$50 million 1 10 8 13 32 (22.1%)$50-$250 million 1 7 7 23 38 (26.2%)Greater than$250 million

    15 10 17 18 60 (41.4%)

    Total 18 28 39 60 145(12.4%) (19.3%) (26.9%) (41.4%) (100.0%)

    Note: n 145Table II.Respondents profileby supply chain role

    IJPDLM41,2

    120

  • significance, with an 80 percent power level for a sample size of 150 which is close to thesample size of 145 in this study (Hair et al., 1998). As shown in Table III, the values forCronbachs alpha for the SCM practices range from 0.608 to 0.878. The factor fororganizational performance was also extracted using a similar method. The fivestatements related to performance loaded onto a single factor. The factor loadings andreliability coefficients show a high level of reliability with a Cronbachs alpha value of0.812. The results show that there is a fairly high level of congruence among the itemsmeasuring a particular factor. The relatively high values of the Cronbachs alpha whichmeasures the reliability of the factors and the eigenvalues provides excellent validity tousing the factor scores for all subsequent analysis. For instance, the SOURCE factorrelates to the supply network structure and has a reliability coefficient of 0.776. The threeitems that load on the factor relate to: deciding whether, and how much, to outsource;selecting and certifying suppliers; and, rationalizing the supply base (e.g. strategicpartnering, vertical integration and single source supply). All three items relate to thesupplier-related practices followed by companies and hence, the loading is logical. Theloadings of the items on the other factors can also be explained in a similar manner.Based on these results, each respondents individual survey items under each extractedfactor are normalized by taking the average scores over the items and the resultingfactor scores are used as the independent variables.

    SCM practices Items Factor loadings Cronbachs alpha

    Information sharing (SHARE) SHARE1 0.678 0.795SHARE2 0.777SHARE3 0.756SHARE4 0.615

    Long-term relationships (RELATION) RELATION1 0.525 0.608RELATION2 0.715RELATION3 0.746

    Advanced planning systems (PLAN) PLAN1 0.708 0.878PLAN2 0.813PLAN3 0.805PLAN4 0.716PLAN5 0.615PLAN6 0.605

    Leveraging the internet (INTERNET) INTERNET1 0.729 0.792INTERNET2 0.819INTERNET3 0.667

    Supply network structure (SOURCE) SOURCE1 0.605 0.776SOURCE2 0.806SOURCE3 0.713

    Distribution network structure(DISTRIBUTION)

    DISTRIBUTION1 0.656 0.692DISTRIBUTION2 0.678DISTRIBUTION3 0.642

    Organizational performance (ORGPERF) ORGPERF1 0.823 0.812ORGPERF2 0.860ORGPERF3 0.757ORGPERF4 0.649ORGPERF5 0.661

    Table III.Factor analysis forSCM practices and

    organizationalperformance

    Supply chain role

    121

  • Table IV displays the means and standard deviations for the entire sample and thefour supply chain roles. These overall results suggest the most commonly pursued SCMpractices include design of the supply network (SOURCE) and building long-termrelationships with supply chain partners (RELATION). The least commonly pursuedSCM practices were information sharing with supply chain partners (SHARE) and thedesign of the distribution network (DISTRIBUTION). It is also interesting to note that areview of the standard deviation suggests that there is the most agreement of the value ofbuilding long-term relationships (RELATION) and the least agreement on the emphasisplaced on information sharing (SHARE).

    There are also differences in the means and standard deviations among the foursupply chain roles. Some of the differences are intuitive. For example, the mean scores ofservice providers are lower than those of the other supply chain roles for the design of thedistribution network (DISTRIBUTION) and the use of advanced planning systems(PLAN), which reflects the absence of physical goods in the value propositions.However, other findings were unexpected, such as the low emphasis placed on sharinginformation with supply chain partners (SHARE) by the service providers. This mayindicate that service providers remain on the periphery of their respective supply chains.It is also interesting to note that the observed standard deviations of the retailers arehigher than those for the other supply chain roles on four of the six constructs (SHARE,RELATION, INTERNET and DISTRIBUTION). This may indicate that compared withthe other roles examined, there is less agreement on what the appropriate SCM practicesare within the retail sector. The results reported in Table IV suggest that the level ofadoption on the commonly advocated SCM practices vary according to the supply chainrole of respondents.

    The inter-correlations of each of the six SCM practices are shown in Table V. Thecorrelations between the independent variables range from 0.083 to 0.398. For example,the significant correlation between RELATION and SHARE of 0.353 shows thatcompanies which have long-term relationships with their supply chain partners wouldbe more likely to share information with these partners. Likewise, the correlationbetween INTERNET and PLAN is significant at 0.366. This points to the strong synergybetween using web-based business practices and using advanced planning systems.Given that many of the correlations shown in Table V are significant, it also points to the

    Supply chain rolesAll supplychain

    members Retailers DistributorsService

    providersManu

    facturersSCM practices M SD M SD M SD M SD M SD

    SHARE 3.87 1.54 3.90 1.75 3.87 1.55 3.50 1.61 4.13 1.42RELATION 5.24 1.09 4.78 1.63 5.20 0.85 5.44 1.09 5.28 0.96PLAN 4.29 1.48 4.38 1.49 4.34 1.34 3.21 1.54 4.94 1.07INTERNET 4.17 1.52 3.99 1.61 4.42 1.34 4.40 1.57 3.95 1.55SOURCE 5.32 1.19 5.06 1.09 5.07 1.24 5.17 1.44 5.61 0.97DISTRIBUTION 3.88 1.41 4.20 1.57 4.26 1.37 3.10 1.43 4.11 1.18

    Notes: Mean M; standard deviation SD; scores on a seven-point Likert scale; higher scoresindicate higher preference for the specific SCM practices

    Table IV.Descriptive statisticsfor the SCM practice

    IJPDLM41,2

    122

  • notion that many of these practices are undertaken simultaneously and some aspects ofa particular practice overlap with aspects of a different practice.

    From a statistical perspective, presence of correlation between independent factors (likethe supply chain practices examined here) results in the possibility of multi-collinearitywhere the independent factors are not really independent of one another. However, in ourstudy, although many of the observed correlations are statistically significant, bivariatecorrelations below 0.80 are not considered indicative of the presence of multi-collinearity inthe dataset (Mason and Perreault, 1991). This means that the application of stepwiseregression remains appropriate under the evidence presented. Johnson (2000) hasrecommended using a RW analysis to supplement the interpretation of the regression resultsin such cases where the independent factors have significant inter-correlations. The RWsmethod provides additional explanatory power regarding the relative importance of specificrelationships between the factors and the dependent variable.

    It is important to test for the discriminant validity of the factors since there aresignificant correlations among some of the independent factors and the scales of thefactors were compiled based on items derived from a number of previous researchstudies. Campbell and Fiske (1959) first introduced the concept of discriminant validityby stating that multiple factors should exhibit discriminant validity when a uniquefactor should not measure similar concepts as the other factors. In other words,measurement error should not exist between multiple factors. This was furtherconfirmed in a subsequent study (Campbell, 1960) where specific recommendations weremade to measure the correlation index between two factors, x and y, as equal to:

    rxy

    rxx:ryyp ;

    where rxy is the correlation between factors x and y; and rxx, ryy are the reliabilitycoefficients of factors x and y, respectively. If the calculated value of the index is less than0.85, the two factors under consideration are distinct from each other and there isdiscriminant validity between the two factors. Table VI displays the results of the indexfor the test of discriminant validity on the SCM factors. The results show that the valuesrange from 0.114 (between PLAN and RELATION) to 0.508 (between SOURCE andINTERNET), which are considerably lower than the accepted threshold of 0.85.In addition, as suggested by Straub (1989), constructs are deemed to be different from oneanother if, during the factor analysis, the respective items load most heavily on differentfactors. This is true based on our results for the factor analysis reported in Table III.

    SHARE RELATION PLAN INTERNET SOURCE DISTRIBUTION

    SHARE RELATION 0.353 * * PLAN 0.349 * * 0.083 INTERNET 0.235 * * 0.187 * 0.366 * * SOURCE 0.253 * * 0.299 * * 0.237 * * 0.398 * * DISTRIBUTION 0.217 * * 0.179 * 0.379 * * 0.228 * * 0.338 * *

    Notes: n 145; significance at: *p , 0.05, * *p , 0.01 levelsTable V.

    Correlations forthe SCM practices

    Supply chain role

    123

  • In conclusion, the methods of Campbell and Fiske (1959) and Straub (1989) indicate thatthe SCM factors have discriminant validity and, in fact, measure separate constructs.

    Table VII displays the results of five stepwise regression models designed toexplore the relationship between the six supply chain practices and organizationalperformance by supply chain role. The first regression model consists of all respondents.The remaining four regression models segregate supply chain member by theirrespective role: retailer, distributor, service provider and manufacturer. As shown inTable VII, four of the five regression models were found to be statistically significantwith R 2-values ranging from 0.174 to 0.624, which are similar to those reported in otherSCM research studies (Tan, 2002; Lockamy and McCormack, 2004).

    These findings indicate thatH1 andH6 are accepted while H2-H5 are rejected. In theoverall model, the SCM practices of sharing information with supply chain partners(SHARE) and design of the distribution network (DISTRIBUTION) were found tosignificantly differentiate an organizations performance from its competitors. Withinthe sub-models, additional support was found that indicates that the supply chain

    SHARE RELATION PLAN INTERNET SOURCE DISTRIBUTION

    SHARE RELATION 0.507 PLAN 0.417 0.114 INTERNET 0.296 0.269 0.439 SOURCE 0.322 0.435 0.287 0.508 DISTRIBUTION 0.292 0.276 0.486 0.308 0.461

    Note: n 145Source: aBased on Campbell and Fiske (1959)

    Table VI.Values for test ofdiscriminant validitya

    among the SCM practices

    Supply chainrole Factors Coefficient

    Significancelevela R 2 F

    SignificanceLevelb

    All supplychain members

    Constant 3.036 0.000 * * * 0.174 14.974 * * * 0.000SHARE 0.205 0.001 * * *

    DISTRIBUTION 0.209 0.001 * * *

    Retailer Regression wasnot significant

    Distributor Constant 3.460 0.002 * * * 0.624 13.274 * * * 0.000DISTRIBUTION 0.474 0.000 * * *

    PLAN 0.385 0.001 * * *

    RELATION 20.608 0.009 * * *

    Serviceprovider

    ConstantSHARE

    2.9240.436

    0.000 * * *

    0.001 * * *0.282 14.527 * * * 0.001

    Manufacturer Constant 1.700 0.035 * * 0.240 5.901 * * * 0.001DISTRIBUTION 0.247 0.028 * *

    SHARE 0.180 0.052 *

    RELATION 0.318 0.087 *

    Notes: Significance at: *p , 0.10, * *p , 0.05 and * * *p , 0.01 levels; alevel of significance forindividual factors; blevel of significance for overall model

    Table VII.Stepwise regressionresults by supplychain role

    IJPDLM41,2

    124

  • practices of building long-term relationships with supply chain partners (RELATION)and utilization of advanced planning systems (PLAN) also significantly differentiate anorganizations performance relative to its direct competitors. However, evidence was notfound to support a positive relationship between leveraging the internet (INTERNET) ordesign of the supplier network (SOURCE) and differentiated performance.

    When moderated by the supply chain role, the results of the stepwise regressionindicate strong support for H8-H10. Specifically, H8 was accepted for distributors; H9was accepted for service providers; and H10 was accepted for manufacturers. Each ofthe regression equations, when moderated by the supply chain role are unique, withdifferent variables introduced and with different levels of explanatory power. Indeed,with the exception of H7 for retailers, the explanatory power of the regression equationsincreases with the consideration of supply chain role. This is because the values for theR 2 increase from the model encompassing all the supply chain roles to the ones specificfor each supply chain role, except for the retailers. The R 2-value for the overall model is0.174, while those with the supply chain role as a moderating variable result inR 2-valuesof 0.624, 0.282 and 0.240 for the distributors, service providers and manufacturers,respectively. Therefore, when moderated by the supply chain role, the regression modelswith organization performance as the dependent variable and the SCM practices as theindependent variables have a better fit.

    It may be noted that (Table VII), the RELATION variable has a negative coefficient inthe regression model for the distributors. However, this model is the strongest in ourstudy further confirming the evidence that despite this negative relationship betweenRELATION and performance, the inclusion of the variable in the model increases the fitof the model and explains the variance in the dependent variable better than any of theother models examined.

    The inability to identify significant SCM-related relationships within the retail sectormay indicate that marketing-related practices may have a stronger positive relationshipwith organization performance than back office SCM-related practices. An examinationof the respondents profile in Table I shows that for the retailers, the sample was heavilybiased towards large retail companies. As many as 15 out of the 18 retailers reportedannual sales of greater than $250 million and 11 out of 18 reported having more than5,000 employees. A further analysis of the population of US-based retailers that arepublicly traded companies show that more than 91 percent of all such companies havesales greater than $250 million (US Retail Industry Publicly Traded Company List andStock Exchange Symbols, 2010). In other words, the sample bias towards larger retailersis not atypical if one considers the entire population. This also maps to the fact thatrespondents to the survey are members of the ISM and mostly professionals working inlarge corporations are expected to be members of such an organization. Therefore, whilewe cannot make conclusive recommendations about the retailers for the effect of thespecific supply chain practices examined in this study because of the small sample andthe lack of significant independent variables in the regression model, the results do pointto the strong possibility of the existence of additional factors that were not includedin this study. It is possible that specific SCM practices may not have as much an effect onthe retailers as in the other supply chain roles. This is all the more pertinent because theretail industry as classified on the publicly traded companies (US Retail Industry PubliclyTraded Company List and Stock Exchange Symbols, 2010) includes different types ofretailers including online retailers, discount retailers wholesalers, and specialty retailers

    Supply chain role

    125

  • in various industry categories such as dress/apparel, electronics, etc. Therefore, while thelimitations in the current study with regard to the small sample does not allow to examinethe effects of other factors, the results do point to a strong need for further studies todetermine what other (if any) supply chain practices affect retailer performance.

    As previously discussed, when predictor variables are inter-correlated, stepwiseregression does not provide a definitive answer to the relative importance of thepredictor variables. A RW analysis is used to analyze the relative contribution of eachpredictor to the prediction of the dependent variable by considering both the unique andcollective contributions of each of the six SCM practices ( Johnson, 2000). The results ofthe RWs analysis are displayed in Table VIII.

    Similar to the stepwise regression analysis, the RWs have been determined for allrespondents and by supply chain role: retailer, distributor, service provider andmanufacturer. Table VIII shows the percentage of R 2 explained (the RW) by each of theSCM practices. The last row of the table displays the variance explained for each of thefive models with all six variables present in the equation. Therefore, the R 2 shown inTable VIII are higher than the corresponding R 2 shown in Table VII.

    The results from the RW analysis are consistent with the findings from the stepwiseregression models, but also provide additional insight to the relationships underscrutiny. For example, in the model reflecting all respondents, only two factors, SHAREand DISTRIBUTION, entered the regression model. In the RWs analysis, the same twofactors, SHARE and DISTRIBUTION, have the highest RWs (0.070 and 0.063,respectively) and contribute the highest percentages (36.8 and 33.0 percent, respectively)in predicting variation in the dependent variable. However, advanced planning systems(PLAN) also provides a unique contribution of 22.3 percent to the variation explained bythe model. In addition, the SCM practice of leveraging the Internet (INTERNET) was notfound to be significantly related to differentiated performance in the stepwise regressionanalysis. However, using the RWs method, INTERNET is found to contribute12.8 percent to the variance explained in distributor performance and 11.7 percent of thedifferentiated performance of service providers. Although these relationships did notprove to be statistically significant, their unique contribution displayed in the RWsanalysis may indicate they are managerially significant.

    Finally, in the stepwise regression, the design of the supplier network did not appear tosignificantly differentiate organizational performance under any of the contextualparameters considered. However, a review of the RWs output indicates that SOURCEhighly corresponds with the observed variance in differentiated organizationalperformance within the retail sector almost an order of magnitude higher than anyother supply chain role. Combined with the comparatively large relative values of thedesign of the distribution network (DISTRIBUTION) and the use of advanced planningsystems (PLAN), the authors argue that further analysis of SCM practices within the retailsector are warranted. In particular, the results for the retailer role may have been influencedby the fact that franchise agreements and other contractual obligations tie many retailersto specific supply chain partners, and because of this reason, many retailers may havelimited opportunity to control their SCM practices. The potential of this factor to impact theefficacy of SCM practices within the retail sector should be the focus of future research.

    For the distributors, the results in Table VIII are consistent with the stepwiseregression analysis. The RWs indicate a clear hierarchy in the unique contribution of eachpredictor variable found to be significant in the regression analysis DISTRIBUTION

    IJPDLM41,2

    126

  • Su

    pp

    lych

    ain

    role

    sA

    llsu

    pp

    lych

    ain

    mem

    ber

    sR

    etai

    lers

    Dis

    trib

    uto

    rsS

    erv

    ice

    pro

    vid

    ers

    Man

    ufa

    ctu

    rers

    SC

    Mp

    ract

    ices

    RW

    Per

    cen

    tag

    eof

    R2

    RW

    Per

    cen

    tag

    eof

    R2

    RW

    Per

    cen

    tag

    eof

    R2

    RW

    Per

    cen

    tag

    eof

    R2

    RW

    Per

    cen

    tag

    eof

    R2

    SH

    AR

    E0.

    070

    36.8

    00.

    007

    3.40

    0.02

    13.

    200.

    194

    61.8

    00.

    064

    24.7

    0R

    EL

    AT

    ION

    0.00

    52.

    400.

    005

    2.60

    0.09

    514

    .40

    0.14

    4.50

    0.06

    524

    .90

    PL

    AN

    0.04

    222

    .30

    0.03

    417

    .20

    0.17

    125

    .90

    0.04

    514

    .40

    0.03

    011

    .50

    INT

    ER

    NE

    T0.

    006

    3.40

    0.01

    47.

    200.

    084

    12.8

    00.

    037

    11.7

    00.

    011

    4.30

    SO

    UR

    CE

    0.00

    42.

    100.

    061

    30.7

    00.

    023

    3.50

    0.01

    85.

    700.

    009

    3.40

    DIS

    TR

    IBU

    TIO

    N0.

    063

    33.0

    00.

    077

    38.9

    00.

    267

    40.4

    00.

    006

    1.90

    0.08

    131

    .30

    aR

    20.

    190

    **

    0.19

    80.

    661

    **

    0.31

    4*

    0.26

    0*

    Notes:

    Sig

    nifi

    can

    ceat

    :* p,

    0.05

    and

    ** p,

    0.01

    lev

    els;

    aR

    2-v

    alu

    eco

    rres

    pon

    ds

    toen

    ter

    met

    hod

    ofre

    gre

    ssio

    nw

    ith

    all

    fact

    ors

    incl

    ud

    ed;

    rela

    tiv

    eim

    por

    tan

    ceca

    lcu

    late

    du

    sin

    gR

    Ws

    (Joh

    nso

    n,

    2000

    );p

    erce

    nta

    ge

    ofR

    2-v

    alu

    esar

    ep

    erce

    nta

    ges

    ofto

    tal

    var

    ian

    ceac

    cou

    nte

    dfo

    rin

    the

    dep

    end

    ant

    var

    iab

    le(e

    .g.

    org

    aniz

    atio

    nal

    per

    form

    ance

    )

    Table VIII.Relative contributions

    of SCM practices toperformance metricvariance by supply

    chain role

    Supply chain role

    127

  • (40.40 percent), PLAN (25.90 percent), and RELATION (14.40 percent). Moreover,RELATION only marginally exceeds the explanatory power of INTERNET(12.80 percent). Note that RELATION was the third significant factor in the stepwiseregression model and had a negative coefficient. The counter-intuitive direction of thecoefficient is partially related to the small difference between the RWs of the two variables,RELATION and INTERNET and the significant correlation between the two variables.Although the results in Table IV show that respondents within the distributors group hada high score on the RELATION variable, a high score does not necessarily indicate thenature of the relationship of the variable with the performance metric. In fact, a furtherinvestigation of the correlation between RELATION and performance reveals that, fordistributors, the correlation is 20.264 which explains why the regression coefficient isnegative for this variable. Therefore, while the negative coefficient cannot be intuitivelyexplained using just the stepwise regression analysis, the addition of the RWs analysisand the additional insights from the correlations explain a possible cause. The resultsshow that while it has been traditionally accepted that long-term relationships with onessupply chain partners helps a company, the evidence here suggests that this may not betrue for distributors.

    For service providers, information sharing (SHARE) was the only significant factor inthe regression equation. The RWs analysis shows that SHARE contributes 61.8 percentto the observed variance in organizational performance. This confirms the strategicimportance of information sharing for service companies to improving theirperformance the domination of these practices over all other practices.

    For manufacturers, DISTRIBUTION, SHARE and RELATION were significant in theregression equation. The RWs display a similar order with DISTRIBUTION explaining31.3 percent of the variation and the other two factors, SHARE and RELATION,displaying a similar contribution of 24.7 and 24.9 percent, respectively. By comparison,the use of advanced planning systems (PLAN) only contributed 11.5 percent to observedvariance in organizational performance. The use of advanced planning systems was firstinitiated by manufacturers and subsequently adopted by other companies. Combinedwith the insignificant result in the regression model, the comparatively smallcontribution assigned to PLAN suggests that the return on above average investment inadvanced planning systems may be questionable for manufacturers at present. However,the high level of attention paid to advanced planning systems (a mean of 4.94 as shown inTable IV), indicates many manufacturers emphasize usage of advanced planningsystems and our findings may only indicate that the effect of these investments may nolonger be a significant differentiator of organizational performance.

    The preceding comment highlights one limitation of this study. This research wasdesigned to identify SCM practices that positively relate to organizational performancelevels not to establish baseline levels for the SCM practices under review. For example,two of the top three most common practices (SOURCE and INTERNET) were not foundto be significant in any of the stepwise regression models. However, their absence shouldnot be interpreted as implying that investment in outsourcing and supplier selection orutilization of the internet is not important. A similar inference can be drawn for two of theother variables, SHARE and DISTRIBUTION, which had a positive relationshipwith organizational performance for a majority of the supply chain roles examined.While the mean scores on these two variables are low (from Table IV), respondents in thestudy also had a high variance in the adoption levels of the SCM practices. Since we were

    IJPDLM41,2

    128

  • not establishing baseline levels for the practices in the study but were linking them toperformance, it could well be that the higher variances in the scores for these twovariables explain why these two have a positive relationship with organizationalperformance and hence appear in most of the regression models reported in the study. Infact, a further analysis of the correlations between SHARE and DISTRIBUTION withorganizational performance shows that these two variables have the two highestpositive correlations with performance (correlations of 0.335 and 0.321, respectively,with both being significant at the 0.01 level). Hence, what is more significant here is thepositive relationships of the SCM practices to performance rather than the overall meanor variances of the scores across the respondents.

    Conclusions and suggestions for future researchThe results from the study highlight several implications for managers. First, the resultsshow that the importance and effectiveness of specific SCM practices are not the same forall positions in the supply chain. Rather, the mix of SCM practices pursued by anorganization should consider the specific role that a company occupies within a supplychain. This finding is in contrast to the common argument that supply chain partnersshould strive to align their SCM-related strategic initiatives, regardless of their specificroles in the supply chain. Owing to the differences in the results across the supply chainroles, specific SCM practices are more likely to differentiate an organizations performancebased on the organizations role in the supply chain. Managers in companies who areengaged primarily in distribution-related activiti