Running head: 1

Louis Oberdiear, Taylor Herrick, Stephanie Persaud, Maryann Stassen, & Stephen Steiner

Missouri State University

2

Abstract

The popularization of emotional intelligence has sparked controversy over the definition,

dimensionality, and operationalization of the construct. Traditionally, EI is measured through

explicit measures including performance-based tests and self-report questionnaires. The use of

such measures has been highly scrutinized; the susceptibility to faking, need for individual

introspection, and the subjective nature of EI demonstrates threats to validity. The objective of

this study is to develop and validate an Implicit Association Test (IAT) for EI. IAT’s reveal

associations that are introspectively assessed and reported verbally. Through the measurement of

associations between concepts and attributes, an EI-IAT will yield a true evaluation of one’s EI.

The EI-IAT demonstrated convergent and discriminate validity and showed good internal

consistency. The EI-IAT was significantly positively correlated with the TEIQue global trait EI

factor. The EI-IAT was also significantly negatively correlated with the TAS total factor and the

TAS subfactor, difficulty identifying feelings.

3

Table of ContentsAbstract................................................................................................................................2

Table of Contents.................................................................................................................3

Introduction..........................................................................................................................4

Emotional Intelligence.....................................................................................................4

IAT Measures...................................................................................................................7

Method...............................................................................................................................10

Construction of the IAT.................................................................................................10

Participants....................................................................................................................12

Procedure.......................................................................................................................12

Measures........................................................................................................................12

Trait Emotional Intelligence Questionnaire (TEIQue-SF)........................................12

Toronto Alexithymia Scale (TAS -20).......................................................................12

Data Analysis.................................................................................................................13

Results................................................................................................................................15

Discussion..........................................................................................................................20

Summary of Findings....................................................................................................20

Study Implications.........................................................................................................20

Future Research.............................................................................................................20

References..........................................................................................................................21

Appendix A: TAS-20.........................................................................................................23

Appendix B: TEIQue-SF...................................................................................................25

4

Introduction

From Forbes to elite academic journals, emotional intelligence (EI) has received

substantial interest in both the academic and practitioner arena (Lievens & Chan, 2010).

Simultaneously, this construct has stimulated criticism and controversy. Described as an “elusive

term”, researchers have debated its definition, dimensionality, and operationalization (Lievens &

Chan, 2010; Mayer, Salovey, & Caruso, 2004).

In modern society, being emotionally intelligent has been attributed to personal and

professional success (Matthews, Zeidner, & Roberts Need date). Some have argued that

emotional intelligence is a better predictor than general intellectual ability of “academic

achievement, occupational success and satisfaction, and emotional health and adjustment” (Elias,

Zins, Weissberg, Frey et al., 1997; Goleman, 1995). Furthermore, EI has been conceived as

trainable; many workplaces and educational settings are developing EI interventions.

The public exposure and media attention EI has received, has fostered misrepresentations

of the construct (Mayer, Salovey, and Caruso, 2004). Given the popularization of EI, it is

imperative for researchers to promote a consistent conceptualization and valid measure of EI that

is rooted in theory and supported through empirical evidence.

Emotional Intelligence

Ability versus Trait. Generally, EI is conceptualized as either an ability or trait (Lievens

& Chan, 2010). Mayer, Caruso, and Salovey (2000) describe intelligence as an umbrella term

encompassing various types of mental abilities. Ability models of EI conceptualize EI as true

intelligence; thus, EI is referred to “as emotional cognitive ability or information processing

emotional intelligence” (Lievens & Chan, 2010; Mayer, Caruso, & Salovey, 2000).

5

Emotional cognitive ability is defined as mental processes that include “appraising and

expressing emotions in the self and others, regulating emotion in the self and others, and using

emotions in adaptive ways” (Salovey & Mayer, 1990). Utilizing this conceptualization of EI,

abilities and skills of EI can be divided into four branches: (a) perceive emotion, (b) use emotion

to facilitate thought, (c) understand emotions, and (d) manage emotion (Mayer, Salovey, and

Caruso, 2004). The first branch describes one’s aptitude to identify emotion in others’ non-verbal

expressions (Mayer, Salovey, and Caruso, 2004). Branch two includes the capacity of emotions

to assist thinking; hence, the ability to link and utilize emotions to drive future behavior (Mayer,

Salovey, and Caruso, 2004). The third branch includes the ability to reflect, analyze, and

understand the outcomes of emotions (Mayer, Salovey, and Caruso, 2004). Lastly, the fourth

branch reflects the capacity to manage emotions to achieve one’s goals (Mayer, Salovey, and

Caruso, 2004).

The trait EI model views “emotional intelligence as akin to personality” (Lievens &

Chan, 2010). The EQ-i is the most commonly used trait EI measure within the literature (Perez,

Petrides, & Furnham, 2005). The EQ-i is based on the premise that trait EI is “an array of non-

cognitive capabilities, competencies, and skills that influence one’s ability to succeed in coping

with environmental demands and pressures” (Bar-On, 1997). The EQ-i contains 15 subscales

with 5 higher-order factors including: “intrapersonal”, “interpersonal”, “adaptation”, “stress

management”, and “general mood” (Perez, Petrides, & Furnham, 2005).

Measuring Emotional Intelligence. Trait EI and ability EI do not measure the same

constructs (Lievens & Chan, 2010). The key distinction between trait EI and ability EI is the type

of measurement employed. Ability EI utilizes performance-based tests and trait EI measures it

via self-report questionnaires (Lievens & Chan, 2010). Some research studies employ mixed

6

models; mixed models incorporate cognitive abilities and personality traits (Perez, Petrides, &

Furnham, 2005).

The use of performance-based and self-report measures to capture EI has been

scrutinized. Van Rooy, Viswesvaran, & Pluta’s (2005) meta-analytic research demonstrated that

the models (ability and trait measures) only correlated .14 (Lievens & Chan, 2010). Further,

research has demonstrated that the two constructs (ability EI and trait EI) hold different

correlates (Lievens & Chan, 2010). EI models based on ability are more correlated with

cognitive ability than personality (Lievens & Chan, 2010). Mixed models of EI corresponded

with personality scores but not with cognitive ability scores (Lievens & Chan, 2010). Thus, the

use of performance-based and self-report tests to capture EI yields construct validity concerns.

Problems concerning the use of performance-based and self-report measures have also

been conveyed. Self-report measures are cost effective and easily distributed; however, self-

report measures are susceptible to “various validity problems such as faking and inflation of

correlations because of common method variance” (Lievens & Chan, 2010). Trait EI measured

through performance-based tests presents participants with problems or situations thought to

have a correct response. The main critique of EI performance-based tests is (unlike performance-

based tests to measure general mental abilities) emotional intelligence tests “do not have

objectively correct answers” (Lievens & Chan, 2010).

Given the wide controversy with the existing explicit measures of EI, it can be argued

that a new instrument to measure EI is warranted. Lievens and Chan (2010) state “in comparing

the findings from the ability and the trait models, a major methodological problem exists because

of a method-construct confound resulting from the fact that ability model is often measured using

performance-based tests whereas the trait model is often measured using self-reports.”

7

IAT Measures

The susceptibility to faking and need for individual introspection demonstrate a major

threat to the validity of the explicit measures that are currently being used to measure EI. These

threats introduce a great need for a more accurate measure of EI. We can focus on one such

measure, The Implicit Association Test (IAT). Lane, Banaji, Nosek, & Greenwald (2007)

provide an overview of the basic concepts behind the IAT.

When people see a stimulus in the environment an automatic thought is presented

reflecting the individual’s attitudes and beliefs toward the stimulus. For example, when an

individual sees a woman they may be quick to think she is a mother or caretaker; seeing a doctor

may elicit thoughts of intelligence or skill associated with that doctor; a spider may produce a

reaction of immediate fright or disgust; upon hearing the word elder, words such as gray, slow,

and perhaps ill might come to mind. Although you may think these invisible thoughts and have

these implicit attitudes, you may make statements contradicting these thoughts, saying that you

like spiders or that you don’t actually think of women as caretakers. When using explicit

measures, one can see how an individual may mark answers that are not accurately capturing

their thoughts, attitudes, and beliefs.

People experience countless invisible thought processes every day. These implicit

cognitive notions lead to many explicit behaviors, which inherently reflect those implicit

attitudes. Lane et al. (2007), as well as Greenwald, McGhee, and Schwarz (2008), describe the

ways in which an IAT captures those implicit attitudes using reaction time, or the time that

elapses between a stimulus presentation and a response production. The idea behind an IAT is

that “the easier a mental task, the quicker the decision point is reached and the fewer the errors

that result” (Lane et al, 2007).

8

An IAT “provides an estimate of the strength of association between concepts and

attributes” (Lane et al., 2007). IAT’s are a computer based test in which two categories of words

(e.g., flower and good, or insect and good) are assigned to each of two response keys. If your

implicit attitudes more closely associates flower with good, it should be easier to respond to them

as a pair and your reaction time will therefore be faster for this pair over insect and good.

Participants are first given directions on which words go into which category and then are given

the chance to practice. When taking the IAT, participants are to use the response keys to sort the

words into the correct categories as quickly as possible.

We gain a measure of the implicit attitude when responding is faster with good/flower

than with good/insect pairing. The faster response time indicates greater positivity of the implicit

attitude associated with flower. The use of IAT’s provides an indirect measurement of individual

differences in implicit social cognition.

For example, in Stage 1 of the IAT, participants quickly organize words into the

categories flower (by pressing the left computer key) and insect (by pressing the right computer

key). In Stage 2, the task is repeated for the categories good and bad. In Stage 3, the categories

are combined and participants press either the left computer key when items referring to flower

or good appear on the screen, and press the right computer key when items in the insect or bad

category appear on the screen. A further set of trials are conducted in Stage 4. In Stages 5-7, the

earlier combined pairings are reversed, now pairing flower with bad and insect with good.

Flower and bad now share the left computer key, and insect and good share the right response

key. The structure of this IAT can be seen in Table 1. Typically attitudes toward flowers are

more positive than attitudes toward insects and so it can be predicted that participants will

9

respond more rapidly, on average, when the categories flower and good are combined, and insect

and bad are combined, as opposed to flower and bad, and insect and good. (Lane et al., 2007)

Table 1. Outline Structure of the IAT.

Block Number of Trials Left key response Right key response

1* 20 Flower Insect

2* 20 Good Bad

3* 20 Flower + Good Insect + Bad

4 40 Flower + Good Insect + Bad

5* 40 Insect Flower

6* 20 Insect + Good Flower + Bad

7 40 Insect + Good Flower + Bad

 *Practice blocks

In addition to effectively discriminating between groups of people who perform a

behavior and those who do not (such as smoking, or avoidance of spiders), the IAT successfully

predicts behavior (Greenwald, McGhee, & Schwarz, 2008). Poehlman, Uhlmann, Greenwald,

and Banaji (2005) produced a meta-analysis using 86 independent samples. Results for that

meta-analysis found that the IAT predicted multiple criterion variables, including social

judgments, physiological responses, and social action. Most research using IATs focus attention

on stereotyping and prejudice, and health-related behaviors, such as food choices, alcohol use,

and smoking. A review of the IAT literature suggests there is substantial evidence that the IAT

predicts behaviors and judgments in these domains (Spruyt, Lemaigre, Salhi, Van Gucht,

Tibboel, Van Bockstaele, & Nackaerts, 2015; Bongers, Jansen, Houben, & Roefs, 2013; Houben,

& Wiers, 2006).

By measuring associations between concepts and attributes, the IAT can reveal

associations that often differ from those that are introspectively accessed and reported verbally

10

(Lane et al., 2007). The IAT may prevent against self-presentational forces that may hide

personally or socially undesirable evaluative associations, such as the ethnic and racial attitudes

an individual may have. Further, the IAT is an adaptable tool that provides the opportunity for

assessment of a wide variety of associations (Greenwald, McGhee, & Schwarz, 2008).

Method

Construction of the IAT

When constructing the IAT, guidelines from Lane et al (2007) were used. The structure of

the IAT was kept the same as outlined in Table 1. Using 40 trials in blocks 4 and 7 has shown to

reduce order effects when the categories are switched. The design chosen was the balanced-

valance IAT meaning category names and attribute stimuli have similar valances. Valance can be

described as the emotion that is evoked and can be positive or negative to different degrees.

When choosing the comparison construct, the researchers conducted a literature review in

order to find a construct that was negatively correlated with trait emotional intelligence. The

intent here was to exaggerate the IAT effect as most people who are high in trait EI will be low in

the comparison construct and vice-versa. The researchers found narcissism to be negatively

correlated with empathy (Watson, Grisham, Trotter, & Biderman, 1984). In Thomas and

Walker’s (2014) article on the development of an environmental IAT, they used Egoistic Values

as a comparison construct which is in close relation to narcissism. The researchers used words

such as ambition, money, and prestige. It was in the opinion of the researchers that this

comparison construct would work well in contrast to trait EI. The researchers were polled to rate

the valence of “Emotional Intelligence” and “Egoistic Values.” The researchers concluded that

“Emotional Intelligence” had a slight positive valence and “Egoistic Values” had a slight

11

negative valence. Since a balanced-valance IAT design was being used, the name “Egoistic

Values” was changed to “Individualism.”

The researchers then developed stimuli for each category construct. Nosek, Greenwald,

and Banaji (2005) advise on using a minimum of four items per category and this guideline was

followed. EI stimuli were chosen to represent each facet of Daniel Goleman’s model of EI

(1995); self-awareness, social awareness, self-management, and relationship management. One

word for each facet was chosen and can be seen in Table 2. Three words were borrowed from

Thomas and Walker’s article: prestige, reputation, and money. The researchers added legacy to

more accurately represent narcissism.

Table 2. Word Stimuli for categories (Emotional Intelligence and Comparison construct)Emotional Intelligence Individualism

Relationships Legacy Empathy Prestige

Poise ReputationAdaptability Money

Attribute categories were chosen to be “Good/Bad” (Table 3) on the evidence produced

from the Nosek and Hansen (2008) article suggesting using other attribute categories can elicit

more conscious processing thus reducing the effectiveness of the IAT.

Table 3. Word Stimuli for AttributesGood Bad

Marvelous Tragic Superb Horrible

Glorious TerribleWonderful Awful

12

Participants

A total of 181 psychology students were recruited from Missouri State University and

received course credit for participating in the study. No demographic information was collected.

Procedure

Subjects were recruited using the university’s SONA systems, which allowed students to

select and sign-up for studies online. Participants were then sent a link to the study using their

school email. Upon clicking the link, participants viewed an informed consent statement and

agreed by clicking continue. After clicking continue, the three different measures would be

launched on the participants computer screen.

Measures

Trait Emotional Intelligence Questionnaire (TEIQue-SF). Petrides and Furnham

(2006) created a shorter EI questionnaire aimed to capture the global trait of EI. The TEIQue-SF

includes 30 items on a 7-point Likert scale (1 = “Completely Disagree” to 7 = “Completely

Agree”). There are 2 items from each of the 15 subscales of the original TEIQue included in this

shorter form (e.g. adaptability, self-esteem, and social awareness). Items were selected based on

their high correlations with total subscale scores. These items ensure adequate internal

consistency and a broad coverage of the EI domain. Internal consistencies were satisfactory for

both males and females (αmale = .84, αfemale = .89) (Petrides & Furnham, 2006).

Toronto Alexithymia Scale (TAS -20). Parker, Bagby, and Taylor’s (1992) Toronto

Alexithymia Scale (TAS-20) was employed. The construct, alexithymia, has been explained as a

“deficit in the cognitive capacity to process and regulate emotions (Parker, Bagby, Taylor,

Endler, & Schmitz, 1993). The TAS-20 was developed to encompass the following dimensions:

(a) difficulty identifying and describing feelings; (b) difficulty distinguishing between feelings

13

and the bodily sensations of emotional arousal; (c) constricted imaginal processes, as evidenced

by a paucity of fantasies; and (d) a cognitive style that is concrete and externally orientated

(Parker, Bagby, Taylor, Endler, & Schmitz, 1993). The TAS-20 has strong validity with a

Cronbach’s alpha of .91 and Test-Retest reliability of .77. The 20 items were rated on a 5-point

Likert scale (1 = “Strongly Disagree” to 1 = “Strongly Agree”).

Data Analysis

The goal of the data analysis is to show construct validity through the Multi-Trait Multi-

Method as presented in the seminal paper by Campbell and Fiske (1959). Construct validity will

be shown through both convergent validity and discriminant validity. Convergent validity is the

degree in which concepts that should theoretically be related are shown to be related. Evidence

for convergent validity is shown when different methods measure the same trait and show

relatively similar results. Discriminant validity is the degree in which concepts that should

theoretically not be related are shown to be unrelated. Evidence for discriminant validity is

shown when independent but same methods measure different traits and show different results.

Two techniques will be utilized in showing convergent and discriminant validity of the

new measure. The first technique is the more traditional approach using correlations as evidence

(MTMM correlation matrix). The second technique is more complicated and a newer method

that is a little more precise then using correlations. The second technique is using the Multi-Trait

Multi-Method (MTMM) through Structural Equation Modeling (SEM). The reason for MTMM

through SEM being used is the criticism that the original MTMM design has received and

limitations being shown in the basic analytic strategy. MTMM through SEM has also been the

most widely supported alternative method (Byrne, 2010).

14

Convergent and discriminant validity is shown through comparison of model fit statistics

through nested models as first presented by Widaman (1985). The technique and guidelines

suggest using a nested model and comparing the first model to a series of more restricted models.

Figure 1 shows this study’s hypothesized model (will be referred as Model 1 henceforth)

that will be compared against three other models. Model 1 has correlated traits and correlated

methods. Figure 2 shows the second model (Model 2). Model 2, like stated before, is more

restrictive with the traits being removed (Figure 2). Model 3 has perfectly correlated traits and

freely correlated methods (Figure 3). Model 4 has freely correlated traits and uncorrelated

methods (Figure 4).

Model 1-Model 2 comparison should show Model 2 having poorer model fit statistics

providing evidence for convergent validity due to the lack of specified traits. Model 1-Model 3

comparison should show discriminant validity by comparing a model in which traits are freely

correlated (Model 1) to a model in which traits are perfectly correlated (Model 3). The bigger

difference in model fit statistics, the stronger the support for discriminant validity. However, it

should be noted in this study we would not expect a large discrepancy because emotional

intelligence and alexithymia are expected to be highly correlated. We would still expect a

significant difference in model fit statistics, though. Model 1-Model 4 comparison uses the same

logic as used in the Model 1-Model 3 comparison but in reverse. Model 4 removes the

correlation between the methods. Discriminant validity is shown in this comparison through

Model 1 and Model 4 not being significantly different because we would not expect methods to

be highly correlated thus showing no method bias.

15

Results

Descriptive statistics and alpha coefficients are shown in Table 4. The EI-IAT was shown

to have good internal consistency with an alpha coefficient of .73. Most of the explicit measures

also showed good internal consistency with alphas .7 or greater. Zero-order correlations of all

study variables are shown in Table 5. The EI-IAT was significantly positively correlated with the

TEIQue global trait EI factor. The EI-IAT was also significantly negatively correlated with the

TAS total factor and the difficulty identifying feelings sub-factor. Both of these correlations are

evidence for convergent validity.

Table 4. Descriptive Statistics for Study VariablesVariables N Min Max Mean SD AlphaImplicit Measurea                 EI-IAT 181 -1.47 1.20 -0.01 0.47 .73Explicit Measures                 TEIQue     Global Trait EI 181 3.07 6.53 4.99 .65 .88     Well Being 181 2.00 7.00 5.62 .97 .86     Self-Control 181 2.33 6.33 4.51 .83 .62     Emotionality 181 1.88 6.75 4.85 .81 .66     Sociability 181 2.67 6.67 5.00 .83 .68     TAS TAS Total 181 25.00 77.00 49.97 10.35 .83     Diff. ID. Feel 181 7.00 30.00 15.83 5.41 .85 Diff. Desc. Feel 181 5.00 24.00 13.58 4.06 .76 Ext. Orient. Think 181 9.00 33.00 20.56 4.37 .65

a The IAT measure is a D score (standardized mean differences measured in milliseconds); larger (positive) values reflect stronger associations with the global trait EI.

Individual model goodness-of-fit indicies are listed in Table 6. The hypothesized model

(Model 1) showed excellent fit with a RMSEA below .10 with a confidence interval only

containing values below .10 and a CFI above .90. The differential of the goodness-of-fit indicies

of each model comparison are shown in Table 7. Evidence of convergent validity was shown in

the comparison of Model 1 and Model 2 with a significant change in χ2 and CFI. Discriminant

16

validity evidence was shown in the comparison of Model 1 and Model 3 with a significant

change in χ2 and CFI. More evidence for discriminant validity was shown in the comparison of

Model 1 and Model 4 when there was no significant difference or practical difference between

the models.

Running head: 17

Table 5. Zero-order Correlations for Study Variables

Variables 1 2 3 4 5 6 7 8 9 10

Implicit Measure

1. EI-IAT -

Explicit Measures

2. TEIQue Total 0.15* -

3. Well-Being 0.14 .83** -

4. Self-Control 0.13 .72** .48** -

5. Emotionality 0.09 .71** .43** .36** -6. Sociability 0.11 .70** .51** .38** .32** -

7. TAS Total -0.18* -.70** -.52** -.42** -.70** -.40** -

8. Diff. Ident. Feel -0.17* -.64** -.50** -.53** -.47** -.39** .79** -

9. Diff. Desc. Feel -0.13 -.63** -.44** -.34** -.69** -.37** .86** .61** -

10. Ext. Orient. Thghts -0.09 -.28** -.20** -.02 -.44** -.13 .59** .06 .36** -

* p < .05; ** p < .01

Running head: 18

Table 6. Summary of Goodness of Fit Indicies for MTMM Models

Model χ2 df CFI RMSEA 90% C.I.

1. Freely correlated traits; freely correlated methods

20.387 16 0.991 0.039 .000, .084

2. No traits; freely correlated methods

105.653 26 0.845 0.130 .105, .157

3. Perfectly correlated traits; freely correlated methods 38.911 18 0.951 0.080 .045, .115

4. Freely correlated traits; uncorrelated methods

22.514 18 0.991 0.037 .000, .080

Table 7. Differential Goodness-of-Fit Indicies for MTMM Nested Model Comparisons

Difference in

Model Comparisons χ2 df CFI

Test of Convergent Validity

Model 1 vs. Model 2 85.27* 10 0.15*

Test of Discriminant Validity

Model 1 vs. Model 3 18.52* 2 0.04*

Model 1 vs. Model 4 2.13 2 0.00

19

Discussion

Summary of Findings

The EI-IAT demonstrated convergent and discriminate validity and showed good internal

consistency. The EI-IAT was significantly positively correlated with the TEIQue global trait EI

Figure 1. Hypothesized Model (Model 1)Figure 2. Model 2

Figure 3. Model 3 Figure 4. Model 4

20

factor. The EI-IAT was also significantly negatively correlated with the TAS total factor and the

TAS subfactor, difficulty identifying feelings.

Study Implications

The findings in this study have unique implications for organizations, practitioners, and

the field of EI research. As noted, EI is measured through performance-based and self-report

tests. However, these current explicit measures are susceptible to faking and require introspective

ability and therefore, yield validity concerns. The use of a valid implicit measure of EI provides

organizations and I/O practitioners with a tool to capture an individual’s “true” EI. Because EI is

predictive of job performance, it is imperative that it is accurately evaluated. Further, an accurate

assessment of EI can lead to better organizational decisions including promotional and selection

decisions.

Future Research

The development of an implicit measure of EI presents valuable information and

opportunities for the field of EI research. Given the inconsistencies and contradictions in the

measurement of EI, future research of implicit measures is warranted.

We advise that the EI-IAT be utilized in future studies to further assess reliability and

validity. We also recommend the EI-IAT be used with diverse samples to ensure generalizability.

Moreover, future studies should compare explicit measures to EI-IAT to demonstrate incremental

predictive validity over the self-report and performance-based measures.

21

References

Bar-On, R. (1997). The Emotional Intelligence Inventory (EQ-i): Technical manual. Toronto: Multi-Health Systems.

Bongers, P., Jansen, A., Houben, K., & Roefs, A. (2013). Happy eating: The Single Target Implicit Association Test predicts overeating after positive emotions. Eating Behaviors, 14(3), 348-355. doi:10.1016/j.eatbeh.2013.06.007

Elias, M. J., Zins, J. E., Weissberg, R. P., Frey, K. S., Greenberg, M. T., Haynes, N. M. et al. (1997). Promoting social and emotional learning: Guidelines for educators. Alexandria, VA: Association for Supervision and Curriculum Development.

Goleman, D. P. (1995). Emotional Intelligence: Why It Can Matter More Than IQ for Character, Health and Lifelong Achievement. Bantam Books, New York.

Greenwald, A. G., McGhee, D. E., & Schwartz, J. L. (1998). Measuring individual differences in implicit cognition: The implicit association test. Journal of Personality and Social Psychology, 74(6), 1464-1480.

Greenwald, A. G., Poehlman, T. A., Uhlmann, E. L., & Banaji, M. R. (2009). Understanding and using the Implicit Association Test: III. Meta-analysis of predictive validity. Journal Of Personality And Social Psychology, 97(1), 17-41. doi:10.1037/a0015575

Houben, K., & Wiers, R. W. (2006). Assessing implicit alcohol associations with the Implicit Association Test: Fact or artifact?. Addictive Behaviors, 31(8), 1346-1362. doi:10.1016/j.addbeh.2005.10.009

Lane, K. A., Banaji, M. R., Nosek, B. A., & Greenwald, A. G. (2007). Understanding and Using the Implicit Association Test: IV: What We Know (So Far) about the Method. In B. Wittenbrink, N. Schwarz, B. Wittenbrink, N. Schwarz (Eds.), Implicit measures of attitudes (pp. 59-102). New York, NY, US: Guilford Press.

Lievens, F., & Chan, D. (2010). Practical intelligence, emotional intelligence, and social intelligence. In Farr and Tippins (Ed.), Handbook of employee selection (pp. 231-337). NY: Routledge.

Nosek, B., Greenwald, A., & Banaji, M. (2005). Understanding and using the Implicit Association Test: II. Method variables and construct validity. Personality and Social Psychology Bulletin, Vol. 31 No. 2, February 2005 166-180.

Nosek, B. A., & Hansen, J. J. (2008). Personalizing the Implicit Association Test increases explicit evaluation of target concepts. European Journal of Psychological Assessment, 24, 226-236

22

Matthews, G., Zeidner, M., & Roberts, R. D. (2002). Introduction. In Emotional intelligence, science and myth (pp. 1-28). MA: Cambridge.

Mayer, J. D., & Salovey, P. (1995). Emotional intelligence and the construction and regulation of feelings. Journal of Applied & Preventative Psychology, 4, 197-208.

Mayer, J. D., Salovey, P., & Caruso, D. R. (2004). Emotional intelligence: Theory, findings, and implications. Psychological Inquiry, 15(3), 197-215.

Mayer, J. D., Salovey, P., & Caruso, D. R. (2000). Emotional intelligence meets traditional standards for an intelligence. Intelligence, 27(4), 267-298.

Parker, J. D., Bagby, R. M., Taylor, G. J., Endler, N. S., & Schmitz, P. (1993). Factorial validity of the 20-item Toronto alexithymia scale. European Journal of Personality, 7, 221-232.

Pérez, J. C., Petrides, K. V., & Furnham, A. (2005). Measuring trait emotional intelligence. In R. Schulze and R. D. Roberts (Eds.), International Handbook of Emotional Intelligence. Cambridge, MA: Hogrefe & Huber. [Reprinted in German].

Petrides, K. V. & Furnham, A. (2006). The role of trait emotional intelligence in a gender-specific model of organizational variables. Journal of Applied Social Psychology, 36, 552-569.

Salovey, P. & Mayer, J. D. (1990). Emotional intelligence. Imagination, Cognition, and Personality, 9, 185- 211.

Spruyt, A., Lemaigre, V., Salhi, B., Van Gucht, D., Tibboel, H., Van Bockstaele, B., & ... Nackaerts, K. (2015). Implicit attitudes towards smoking predict long-term relapse in abstinent smokers. Psychopharmacology, doi:10.1007/s00213-015-3893-2

Tett, R. P., Fox, K. E., & Wang, A. (2005). Development and validation of a self-report measure of emotional intelligence as a multidimensional trait domain. Personality and Social Psychology Bulletin, 31, 859-888.

Van Rooy, D. L., & Viswesvaran, C. (2004). Emotional intelligence: A meta-analytic investigation of predictive validity and nomological net. Journal of Vocational Behavior, 65, 71-95.

Watson, P., Grisham, S. O., Trotter, M. V., & Biderman, M. D. (1984). Narcissism and Empathy: Validity Evidence for the Narcissistic Personality Inventory. Journal Of Personality Assessment, 48(3), 301.)

23

Appendix A: TAS-20

24

25

Appendix B: TEIQue-SF

Instructions: Please answer each statement below by putting a circle around the number that best reflects your degree of agreement or disagreement with that statement. Do not think too long about the exact meaning of the statements. Work quickly and try to answer as accurately as possible. There are no right or wrong answers. There are seven possible responses to each

statement ranging from ‘Completely Disagree’ (number 1) to ‘Completely Agree’ (number 7).

1 . . . . . . . . . 2 . . . . . . . . . . 3 . . . . . . . . . . 4 . . . . . . . . . . 5 . . . . . . . . . . 6 . . . . . . . . . . 7Completely Completely

Disagree Agree

1. Expressing my emotions with words is not a problem for me. 2. I often find it difficult to see things from another person’s viewpoint. (R) 3. On the whole, I’m a highly motivated person. 4. I usually find it difficult to regulate my emotions. (R) 5. I generally don’t find life enjoyable. (R) 6. I can deal effectively with people. 7. I tend to change my mind frequently. (R) 8. Many times, I can’t figure out what emotion I'm feeling. (R) 9. I feel that I have a number of good qualities. 10. I often find it difficult to stand up for my rights. (R) 11. I’m usually able to influence the way other people feel. 12. On the whole, I have a gloomy perspective on most things. (R) 13. Those close to me often complain that I don’t treat them right. (R) 14. I often find it difficult to adjust my life according to the circumstances. (R) 15. On the whole, I’m able to deal with stress. 16. I often find it difficult to show my affection to those close to me. (R) 17. I’m normally able to “get into someone’s shoes” and experience their emotions.18. I normally find it difficult to keep myself motivated. (R) 19. I’m usually able to find ways to control my emotions when I want to. 20. On the whole, I’m pleased with my life. 21. I would describe myself as a good negotiator. 22. I tend to get involved in things I later wish I could get out of. (R) 23. I often pause and think about my feelings. 24. I believe I’m full of personal strengths. 25. I tend to “back down” even if I know I’m right. (R) 26. I don’t seem to have any power at all over other people’s feelings. (R) 27. I generally believe that things will work out fine in my life. 28. I find it difficult to bond well even with those close to me. (R) 29. Generally, I’m able to adapt to new environments. 30. Others admire me for being relaxed.

Scoring key: Reverse-score (R) items and then sum up all responses