Structural equivalence and item bias of a self–report emotional intelligence measure in the mining industry

(1)

mining industry

(2)

mining industry

Francois de Wet (BCom Hons)

Mini-dissertation submitted in partial fulfilment of the requirements for the degree Magister Commercii in Industrial Psychology at the North-West University (Potchefstroom Campus)

Supervisor: Prof. Cara Jonker Co-Supervisor: Dr. A Nel

(3)

Declaration of originality of research

DECLARATION

I, Francois de Wet, hereby declare that Structural equivalence and item bias of a self-report

emotional intelligence measure in the mining industry is my own work, and that the views and

opinions expressed in this study are those of the author and the relevant literature references as shown in the reference list. I also declare that the content of this research will not be submitted for any other qualification at any other tertiary institution.

FRANCOIS DE WET NOVEMBER 2012

(4)

COMMENTS

The reader is reminded of the following:

• The references and the editorial style as prescribed by the Publication Manual (6th edition) of the American Psychological Association (APA) were followed in this dissertation. This practice is in line with the policy of the Programme in Industrial Psychology of the North-West University to use APA style in all scientific documents, as of January 1999.

• This mini-dissertation is submitted in the form of a research article. The editorial style specified by the South African Journal of Industrial Psychology (which agrees largely with the APA style) is used, but the APA guidelines were followed when constructing tables.

(5)

SUMMARY

Topic: Structural equivalence and item bias of a self-report emotional intelligence measure in

the mining industry.

Keywords: Emotional intelligence (EI), psychometric properties, Greek Emotional Intelligence

Scale, item bias, structural equivalence, collectivistic cultures, individualistic cultures, culture and emotion

Emotional intelligence (EI) in organisations has grown immensely over the past two decades. Considerable research regarding this concept and the advantages it poses for the individual as well as the organisation has been conducted; however, one aspect that has not been explored sufficiently is the extent to which EI can be viewed as a culturally relevant concept. The presumption that emotions can be explained in the same way across different culture cannot be made; therefore measuring EI across cultures becomes important and challenging. Language can be viewed as a vehicle of culture, and emotions are shaped by the language spoken in the specific culture.

A quantitative research design was used in this study. The sample consisted of mid-level miners from the Gauteng and North West Province (N = 357). Stratified sampling was used to include the West-Germanic (English and Afrikaans; n = 158) and Sotho group (North Sotho, South Sotho, and Setswana; n = 199). Questionnaires were distributed amongst the participants from the different mines, were completed within a set time, and collected immediately afterwards. The first objective of the study was to determine whether the Greek Emotional Intelligence Scale (GEIS) is a reliable test when measuring West-Germanic (English and Afrikaans) and Sotho (Northern Sotho, Southern Sotho, and Setswana) languages. A four-factor model on the combined sample as well as the two language groups was tested. The four factor model of the West-Germanic group showed poor alphas. (Expression and Recognition of Emotions = 0.66; Caring and Empathy = 0.63; Control of Emotions = 0.80 and Use of Emotions to Facilitate Thinking = 0.62.) Several items from the expression and recognition scale cross-loaded on the

(6)

other three factors, and it was decided to test a three-factor model. The three factor model indicated the best goodness-of-fit indices and showed acceptable alpha coefficients (Use of Emotion to Facilitate Thinking = 0.83; Caring and Empathy = 0.83 and Control of Emotions = 0.77).

The second objective was to determine if the Greek Emotional Intelligence Scale is an equivalent measuring instrument when measuring the West-Germanic (English and Afrikaans) and Sotho (Northern Sotho, Southern Sotho, and Setswana) languages. Goodness-of-fit was tested on the total population as well as the two language family groups. The four factors are Expression and Recognition of Emotions, Caring and Empathy, Control of Emotions and Use of Emotions. The model indices (GFI, CFI and RMSEA) were satisfactory on the total population as well as the Sotho groups, but there were problems noted when testing the goodness-of-fit for the West- Germanic language group. It was therefore decided to test a three factor model (Use of Emotions, Caring and Empathy and Control of Emotions). These problems could possibly be explained by the cultural differences between the two language groups.

The final research objective was to investigate whether the items of the Greek Emotional Intelligence Scale are unbiased when measuring West-Germanic (English and Afrikaans) and Sotho (Northern Sotho, Southern Sotho, and Setswana) languages. Firstly, ANOVAS were produced to determine the mean differences between the groups. There weren’t many differences, indicating none or little biasness between the groups. Then, the uniform and non- uniform biasness was tested by means of Ordinal Logistic Regression to asses Differential Item Functioning. The majority of the items did not have both uniform and non-uniform biasness. The few that did however, (41, 37, 36, 14 and 18) can be explained by the different ways in which cultures interpret emotions as proven in the literature.

(7)

OPSOMMING Onderwerp:

Struktuur gelykheid en item sydigheid van ‘n selfverslaggewende emosionele intelligensie skaal in die myn industrie.

Kernwoorde: Emosionele Intelligensie (EI), psigometriese einskappe, Griekse Emosionele

Intelligensie Skaal, item sydigheid, struktuur gelykheid, kollektivisties-, individualistiese kulture, kultuur en emosie

Emosionel intelligensie (EI) as ‘n organisatoriese konsep het ongelooflik baie gegroei oor die laaste 20 jaar. Baie navorsing is al gedoen rakende die konsep en die voordele wat dit inhou vir die individu sowel as organisasies. Een aspek wat egter nie genoeg aandag verleen is nie, is die mate waarna EI as ‘n kulturele konsep verwys kan word. Die aanname dat emosies dieselfde interpreteer word oor verskillende kulture kan nie gemaak word nie en dis hoekom die meet van EI so belangrik is. Taal kan gesien word as ‘n medium of vervoer middel van kulturre en ons emosies word gevorm deur die taal wat ons praat in ons spesifieke kultuur.

‘n Kwantitatiewe navrosings ontwerp was gebruik in die betrokke studie. Die steekproef was geneem onder middel-vlak myn werkers in die Gauteng en Noord Wes provinsies (N = 357). Gestratifiseerde steekproefneming was gebruik om die twee taalgroepe nl. Wes Germaanse (Engels en Afrikaans; n = 158) asook die Sotho (Noord Sotho, Suid Sotho, en Setswana; n = 199) in te sluit. Vraelyste was uitgedeel aan die deelnemers by die ver skillende myne en in n gegewe tyd ingevul en weer opgeneem direk na afhandeling.

Die eerste doelwit van die studie was om vas te stel of die Griekse Emosionele Intelligensie Skaal betroubaar is met die meet van die Wes Germaanse (Engels en Afrikaans) asook die Sotho (Noord Sotho, Suid Sotho, en Setswana) taalgroepe. ‘n Vier-faktor model is getoets op ‘n gekombineerde populasie sowel as die individuele taal groepe. Die vier-faktor model van die Wes Germaanse groep het lae alpha koefisiënte gelewer (Uitdrukking en Herkenning van Emosie

(8)

= 0.66; Omgee en Empatie = 0.63; Kontrole van Emosie = 0.80 en Gebruik van Emosie om denke te fasiliteer = 0.62). Verskeie items van die uidrukking en herkenning van emosie skaal het gekruis-laai op die ander drie faktore en daar is besluit om ‘n drie-faktor model te toets. Die drie- faktor model het die beste gepas en aanvaarbare alpha koefisiënte getoon (Gebruik van Emosie om denke te fasiliteer = 0.83; Omgee en Empatie = 0.83 en Beheer van Emosies = 0.77).

Die tweede doelwit van hierdie studie was om vas te stel of die Griekse Emosionele Intelligensie Skaal ‘n ekwivalente meetinstrument is wanneer die Wes Germaanse (Engels en Afrikaans) asook die Sotho (Noord Sotho, Suid Sotho, en Setswana) taalgroepe gemeet word. The passing van die model was getoets op die totale populasie sowel as die twee taal groepe. Die vier faktore is Uitdrukking en Herkenning van Emosie, Omgee en Empatie, Kontrole van Emosie en Gebruik van Emosies. Die passing (GFI, CFI en RMSEA) van die totale sowel as die Sotho groep was suksevol, maar daar was egter probleme met die passing van die Wes Germaanse groep. Daar is toe besluit om ‘n drie-faktor model te toets. Hierdie problem kan moontlik verduidelik word deur die verskillende maniere hoe die twee kulture emosies interpreteer en uidruk.

Die finale doelwit van die studie was om ondersoek in te stel of die Griekse Emosionele Intelligensie Skaal onsydig is met die meet van die Wes Germaanse (Engels en Afrikaans) asook die Sotho (Noord Sotho, Suid Sotho, en Setswana) taalgroepe. Eerstens was ANOVAS toegepas om die gemiddelde verskille tussen die twee groepe te meet. Daar was nie baie verskille nie, was dui op geen of baie min sydigheid tussen die groepe. Na dit was die uniforme asook nie- uniforme sydigheid gemeet deur middel van ordinale logistiese regressie om Differensiale Item Funktionering te toets. Die meerderheid van die items het nie albei sydigheid gehad nie en die wat wel het soos items (41, 37, 36, 14 en 18) kon verduidelik word deur die verskillende maniere hoe kulture emosies verstaan en uitbeeld.

(9)

ACKNOWLEDGEMENTS

I would genuinely like to thank the following people for making this mini-dissertation possible: • My Father in heaven. Lord, You are the One who gave me the talent, strength, endurance and

persistence to make this task possible. I thank You Lord for giving me the strength, especially during the last couple of months when my life was at its most challenging.

• My supervisors. Prof Cara Jonker, thank you so much for the support and encouragement. Dr Alewyn, thank you for the guidance and patience and friendship. To the both of you, thank you for believing in me.

• My family. To my father, mothers, and brothers, I appreciate you for always believing in me and keeping me motivated.

• My best friends, Marais Bester and NW Smit. Thank you for all the support, listening and encouragement.

• My friends: Louisa, Gerna, Petri, and Wihan. Thank you for believing and helping me. A special thank you to Wihan for all the after hour assistance.

(10)

This mini-dissertation focuses on the measurement of structural equivalence and the item bias of a self-report emotional intelligence measure within the mining environment. In this chapter, the research objectives and specific objective are discussed. The research design and research method are explained, and the chapter summary and the division of chapters then follow.

1.1 PROBLEM STATEMENT

1.1.1 Overview of the problem

The scientific study of Emotional Intelligence (EI) in organisations has gained considerable research attention over the past decade (Ashkanasy, 2003; Joseph & Newman, 2010; Mayer, Salovey, & Caruso, 2004). Goldenberg, Matheson and Mantler (2006) explain t hat EI has been used by many researchers as an umbrella term, encompassing elements such as “people skills,” “soft skills,” and an overall ability to cope with everyday demands. Emotional intelligence has been proven to play a fundamental role in not only an individual’s personal life (Cherniss, 2000), but also in the work environment (Boon, 2007; Human 2005).

Organisations worldwide are becoming increasingly competitive, and therefore a workforce that possesses both the necessary skills and abilities to complete tasks proficiently as well as the emotion competence to cope with the demands faced on a daily basis is required. EI is therefore important for overall employee effectiveness. (Akerjordet, 2009; Jonker, 2002; Sedmar, Robbins, & Ferris, 2006). Seeing that EI has been proven to be of utmost importance to organisations, the importance of measuring this concept has increased over the past two decades (Akerjordet, 2009; Jonker, 2002; Sedmar et al., 2006). Accurate measurements of EI are thus critical in order to improve the efficiency of employees, and ultimately that of organisations (Jonker, 2002).

(14)

Two distinct models have been developed to classify the different theories and measurements of EI: ability models, in which EI is conceptualised as a set of cognitive abilities in emotional functioning (Mayer & Salovey, 1997), and mixed trait-ability models, which is seen as a conglomeration of a wide range of personality attributes or characteristics and other traits (Bar - On, 2000; Goleman, 1995, 1998; Mayer et al., 2004). While theorists have been developing models of EI, there has been a simultaneous development of measurements instruments (Tsaousis, 2008). Measuring EI is done by drawing on the two different conceptual frameworks (i.e. the ability and trait, or mixed models of ability framework), and utilises one of two measurement approaches: Performance-based tests and self-report instruments. The performance-based tests consist of a series of questions which are scored as “right‟ or “wrong‟. An example of such a question is “Which of the following faces corresponds to the emotion of anger?”, and the respondent receives a score of “right‟ when they identify the correct answer (Conte & Dean, 2006). Self-report inventories are tests where the respondent is asked to respond to items such as “I have good control of my anger”, which are evaluated on a Likert -type scale. Even though interest in the field of EI has increased, most of the research on the concept is directed towards leadership and development, work performance, and stress management. As a result, there are many spheres of EI that have been under-researched. For example, Keele and Bell (2008) observe that few of the studies that have been conducted included more than one EI measurement instrument. Jonker and Vosloo (2008) also asserted that although there is a great amount of research on EI and its advantages for the workplace, very little literature has described how it should be measured and the factors that need to be taken into account with EI measurements. According to Pheiffer (2001), the majority of EI assessments are based on self- report instruments, and problems regarding the measurement of EI include deplorable levels of internal consistency and permanence. Conte (2005) state that the most critical concerns regarding the measurement of EI, range from scoring of ability measurements to discriminating validity issues of self-report measurements and item bias of measurements across various cultures.

In addition to a lack of literature on EI measurement, other research issues are highlighted. According to Shipper, Kincaid, Rontondo, and Hoffman (2003), research studies have neglected to explore the extent to which EI is a culturally relevant concept. In other words, if EI can predict

(15)

organisational-related outcomes such as leadership effectiveness, conflict resolution, and engagement as research indicates (Goleman, 1995; Mayer & Salovey, 1995), does this potential exist in different cultures? Together with this statement, the following question arises: Do EI measures yield equivalent results when measuring different groups?

Botma (2009) has noted that very little research has focused on validation studies in terms of EI. Arguing that only a few studies have compared EI in terms of different gender, culture, and occupation groups, Matsumoto, Angua-Wong, and Martinez (2008) agreed with this view by stating that numerous studies have shown that differences exist between cultural groups, especially in the way in which the different groups recognise emotions. Visser and Viviers (2010) state that because of a globalised world and a migrating workforce, the multicultural nature of populations and organisations has become increasingly prominent. This has presented a number of challenges for the assessment of psychological constructs such as EI (Van der Vijver & Rothmann, 2004). The conclusion can be drawn that EI measures should therefore equivalently be applicable for different groups, including cultural groups.

One of the most pertinent issues dominating current research in psychology is the cross-cultural study of emotion. Culture plays an important role in the understanding and expression of emotions (Matsumoto, 2002). Regardless of how EI has been measured, the possibility for cultural relevancy has been mostly ignored (Shipper et al., 2003). Hofstede (2001) mentioned that the presumption that emotions can and should be discussed by people is certainly not transferable to all cultures, and because Mayer and Solevey (1997) conceptualise EI as the ability to access and accurately use emotions, it can be concluded that EI can vary amongst different culture groups. It is generally recognized that emotions are shaped and maintained by culture (Markus & Kitayama, 1991), and cultural values are thus likely to affect how people perceive, express, and regulate emotions. EI measurements therefore need to pay attention when measuring across various groups of culture whom perceive emotions differently.

Contrasts between individualistic and collectivistic social life undeniably include many influences that have substantive implications for the processing of emotional information. Previous studies have suggested that Africans might be more collectivistic and Westerners more

(16)

individualistic by nature (Bierbrauer, Meyer, & Wolfradt, 1994; Triandis, 1995). Eaton and Louw (2000) add that in general, modernised Western countries like the USA were found to be highly individualistic, whereas developing Eastern nations like Taiwan more collectivistic. Smit and Cronje (2011) mentioned that culture can be manifested in various different ways, with language almost always playing a prominent role. Gelfand and Brett (2004) stated that emotions are culturally determined, because they are shaped by language and used to extract meaning in social contexts. In this regard, the differences in language groups, which serve as vehicles of culture, needs further investigation. Russel and Barrett (1999) raised the possibility that different languages express different emotions, and Ogarkova, Borgeaud, and Scherer (2009) confirmed this statement, adding that the same language groups (e.g. the West-Germanic (English and Afrikaans) and Sotho (Northern Sotho, Southern Sotho, and Setswana) language groups) tend to shape emotions in a similar manner. Several overarching classifications of cultures such as East vs. egocentric West (Ratner, 2000) and individualism vs. collectivism (Hofstede, 2001) have been instrumental in explaining the differences in the various emotion languages use (Ogarkova et al., 2009).

In a study conducted on an African culture, Ghana, which possesses collectivistic characteristics, the results indicated that their language lacked the indigenous equivalent of the word “loneliness‟, which suggests that individuals living in this culture may not experience situations where they are alone as a result of the collectivistic nature of their culture (Hofstede, 2001). Contrasting to these findings, in a previous investigation, the authors observed no noteworthy differences between collectivistic and individualistic cultures with specific regards to how they experience values (Ghorbani, Bing, Watson, Davidson, & Mack, 2002).

It is prevalent theme in the literature that a variety of EI measures exist (Bar-on, 1997; Mayer & Salovey, 1997; Petrides & Furnham, 2003; Saklofski, Austin, & Minski, 2003; Schutte et al., 1998.). However, it is not yet known whether the structure underlying EI scoring is equivalent for employees in collectivistic and individualistic cultures or from different language groups. Goliath-Yarde and Roodt (2011) add that measuring equivalence in cross-cultural comparisons is the key concern in cross-cultural comparative research. Constructs like EI are considered to be

(17)

equivalent amongst different cultural groups when one obtains the same or comparable scores when using the same or different versions of the items or tests (Goliath-Yarde & Roodt, 2011). However, when constructs aren‟t equivalent, item bias is present. Item bias can be referred to as the presence of certain factors in the instrument that pose problems for the validity of cross- cultural comparisons (Goliath-Yarde & Roodt, 2011).

In order to address the aforementioned research issues, Libbrecht, Lievens, and Schollaert, (2010) state that measuring equivalence of any questionnaire across different rating groups is a prerequisite for making substantive interpretations and recommendations. The purpose of this study is to determine the structural equivalence and item bias of a self-report EI measure. Botma (2009) propose the Greek Emotional Intelligence Scale (GEIS) as a possible EI measure showing promise as cultural equivalent EI measure. The equivalence of two language family groups will be tested. The West-Germanic language family group (Afrikaans and English) and Sotho group (Northern Sotho, Southern Sotho, and Setswana). The choice of these two language family groups will be explained in more detail further on. The structure of the rest of the article follows: Firstly, an overview of emotional intelligence and culture will be provided. Then, item bias and equivalence will be discussed in order to understand how these statistical inferences can be utilised as a method with regards to the GEIS‟s psychometric properties.

1.1.2 Literature review

Emotional intelligence and culture

According to Shipper et al. (2003), EI manifests differently across various cultures, and the diversification in the current workforce facing present times gives rise for the need to investigate the impact of culture and language differences on the measurement of EI. Prior research (Shipper, 2003) indicates that emotional intelligence is influenced by culture. One of the reasons provided for this variance is the cultural dimension of individualism-collectivism explained by Hofstede (2001). Tang, Yin, and Nelson (2010) explained that individualism-collectivism could be used to describe the cross-cultural differences that exist when individuals express emotions. Kang and Shaver (2003) contributed to this statement by mentioning that emotional expression is

(18)

more important than emotional differentiation with regards to interpersonal relationships in individualistic cultures, while the opposite can be observed in collectivistic cultures where emotional differentiation is more important.

In a theoretical study conducted by Ilangovan, Scroggins, and Rozell (2007), EI’s effect and influence on cultures was hypothesised. By referring to the four dimensions conceptualised by Hofstede, the authors proposed the cultural differences impacting emotional intelligence. When discussing cultures and examining cultural research, Hofstede’s findings remain the benchmark. Hofstede (2001) identified four basic dimensions namely power distance, individualism- collectivism, masculinity-femininity, and uncertainty avoidance. Power distance refers to the equality or inequality between people in a country. Individualism-collectivism represents the degree to which individuals or collective relationships are reinforced within a society. In highly individualistic cultures, the rights of the individual are emphasized and relationships with others are viewed more loosely. People in an individual culture tend to be more self-reliant and look out for themselves. In low individualistic cultures, large families and mutual relationships that support mutual responsibility are cherished. These cultures are characterized by the close ties that exist between individuals (Hofstede, 2003). The conclusion can be drawn that EI measures must take cultural (and by extension the inherent language) differences into consideration in the construction of EI scales.

Hofstede (2003) found a strong negative correlation between a cultures’s score on the power distance index and its scores on the individualism-collectivism index. High power cultures tend to be more collectivistic, as opposed to low power distance cultures being more individualistic. People from individualistic cultures like the West-Germanic language groups (English and Afrikaans), do not perceive large psychological distance between in-group and out-group members like collectivistic cultures (such as the Sotho language group and other African cultures) do. According to the South African (SA) Language Bill (2000) as provided in the constitution of SA (1996), the country has a linguistic diversity consisting of 11 official languages. Most of these languages are grouped into three different language families (Zeller, 2006), namely the Nguni group (consisting of Zulu, Xhosa, Swati and Southern Transvaal Ndebele), the West-Germanic group (English and Afrikaans), and the Sotho-Tswana group

(19)

(Southern Sotho, Northern Sotho, and Setswana). Some authors in the literature refer to the latter group as the Sotho-Tswana group (Badenhorst, 2010; Zeller, 2006), while others refer to it as the Sotho group. For the purposes of this study, reference to the Sotho group will be utilised. The subsections in the Languages Bill categorise the languages as follow; the Nguni group, the Sotho group, the West-Germanic group, and Tshivenda/Xitsonga (2000). Van Dyk and De Kock (2004) found that Afrikaans- and English-speaking South Africans had more Eurocentric, Western and individualistic values, as opposed to the African groups who produced a more Afrocentric value system of a collectivistic nature. In another study, Eaton and Louw (2000) also found evidence that members of African language families are collectivistic by nature, while members of the West-Germanic language group tend to be more individualistic.

According to Ilangovan et al., (2007), individualistic cultures value self-expression, and use verbal communication strategies to resolve conflict and deal with personal problems. In contrast, people from collectivistic cultures like the Sotho language family, value paternalism, family relationships, and tend to focus more on loyalty within an organisational environment and place value on loyalty to their company. For them, acting in accordance to these values is more important than personal achievement. Given the individualistic-collectivistic differences between these two distinct culture groups, Ilangovan et al. (2007) postulate that people from collectivistic groups will have higher levels of the emotional intelligence dimension of empathy than individualistic groups. However, before such differences can be determined, EI measures should first be found to be equivalent for different language groups (Visser & Viviers, 2010).

One explanation for the hypothesis presented above can be attributed to the influence that universalism and relativism has on different cultures. In the first few decades of the 20th century, cultural relativism was conceptualised by anthropologists as an individual’s beliefs and behaviours are understood by others in terms of that individual’s own culture (Boas, 1989). Opposite to relativism stood the concept of universalism which introduced the belief that some emotions, beliefs and behaviours are universal and are experienced in similar events across all cultures. A study conducted between Chinese and English culture groups found that the culture- relativistic stance proclaims that emotions are determined by the presumptions, judgements and desires of the cultural community people find themselves in (Ogarkova et al., 2009). Therefore,

(20)

one can acknowledge that the way in which emotions are perceived, used, and judged (i.e. EI) can either vary or be the same between cultural groups. Gignac and Ekermans (2010) make a very important statement by saying that should ethnic groups differ at the mean level, it does not necessarily mean that the instrument used is biased to one of the particular groups, because of universalism and relativism emotion in cultures. They further argue that determining the structural equivalence and item bias of measurements before it can be applied in different cultures and language groups therefore becomes important.

According to Ogarkova et al. (2009), language is seen as a vehicle of culture. De Klerk and Mostert (2010) state that language is important as a way of expressing ethnic and cultural identity amongst societies. The knowledge, beliefs, and practices of a certain society are reflected in the language spoken by the individuals from that society. Urban (2011) cont ributes to this statement, stating that it is important to explore boundaries when examining cultural encounters, distinguishing one group from another, understanding language use, and determining inclusion and exclusion. To date, a few empirical studies have been conducted on language differences in EI. Van Rooy and Viswesvaran (2005) conducted a study on American university students by utilising the Schutte EI scale. The results showed that Black students scored higher than White students on the total EI scale. Another study conducted in Canada Parker et al., (2005) found that aboriginal youth scored lower than non-aboriginal youth. However, the equivalence of these measures in different groups has not been determined. It is therefore obvious that in the existing literature there is an inconsistency. In any cross-cultural study, one very important question is posed: can test scores obtained in different cultural groups be interpreted in the same way. According to Van de Vijver and Tanzer (2004), bias and equivalence can be utilised as a method to investigate if a measurement measures the same construct in different cultural/language groups.

Bias and Equivalence

Cole and Moss (1989, p. 205) defined bias as being “when a test score has meanings or implications for a relevant, defining subgroup of test takers that are different from the meanings or implications for the remainder of the test takers”. This definition implies that test scores

(21)

gathered from different subgroups in a society cannot be interpreted in the same way for different culture groups (Visser & Viviers, 2010). Gregory (2007) defined bias as the objective statistical indices that examine the patterning of tests scores for relevant subpopulations. Three kinds of bias can be distinguished in the literature, namely construct bias, method bias, and item bias. Construct bias can be noted when the construct measured is not similar across the various subgroups being compared (Gregory, 2007, as cited by Visser & Viviers, 2010). In this instance, item bias can thus be explained as the inability of the items in the GEIS to measure the same emotional constructs across the two cultural groups.

As with bias, there also exist three types of equivalence, namely construct equivalence, measurement unit equivalence, and full scale equivalence. According to Visser and Viviers (2010), construct equivalence exists when the same construct is measured in the various groups; as opposed to construct in-equivalence occurs when a test measures different constructs in the population. Construct equivalence can be assessed by using structural equation modeling. According to Visser and Viviers (2010), equivalence and bias are the key concepts when cross- cultural comparisons are made, because inferences made on biased or non-equivalent scores are not valid. It is therefore important to ensure that assessment measures, such as the GEIS, used in South Africa across various culture groups are tested for bias and equivalence. Van de Vijver and Tanzer (1997) state that it is important to remember that these concepts (bias and equivalence) do not refer to properties inherent in any instrument, but are rather specific characteristics of an instrument in a specific comparison amongst groups, like the ones in the present study.

There exist a large amount of possibilities as to why groups differ in test scores. Amongst the reasons are ethnicity, socio-economic circumstances, education opportunities, language, and culture (Meiring, Van de Vijver, Rothmann, & Barrick, 2005). Most of the testing done in South Africa is primarily conducted in either English or Afrikaans. According to Visser and Viviers (2010), apart from other possible reasons for these differences between test scores in different groups, results support the suspicion that language can affect the performance of both ability and cognitive tests. The possibility therefore also exists (as the literature indicates) that language use can influence the scores obtained in EI measures. The most extensive South African bias study conducted by Meiring et al. (2005) involved 12 different cultural groups in the police service.

(22)

Results from that study indicated low alphas, particularly for the African languages. It was further noted that a significant number of items were found to be biased. It is clear that in South Africa, with its multicultural population; far too little research has been published that investigate the presence of bias and equivalence when measuring psychological constructs such as EI.

Therefore because the idea that bias and equivalence may yield important information about cross-cultural differences has grown tremendously in the literature, further investigation is needed to confirm this proposition. As Van de Vijver and Tanzer (2007, p. 132) so appropriately state, “cross-cultural research has grown beyond the stage of mere demonstration and into an era where the interpretations of results are of utmost importance.” The purpose of this study will therefore be to determine structural equivalence and item bias of the GEIS in two language family groups of South Africa: one of an individualistic cultural group and the other of a collectivistic cultural group. The following research questions are therefore formulated:

• How can EI be conceptualised with regards to emotion and culture in a literature review?

• Is the Greek Emotional Intelligence Scale reliable when measuring West-Germanic (English and Afrikaans) and Sotho (Northern Sotho, Southern Sotho, and Setswana) languages?

• Is the Greek Emotional Intelligence Scale an equivalent measuring instrument when measuring West-Germanic (English and Afrikaans) and Sotho (Northern Sotho, Southern Sotho, and Setswana) languages?

• Are the items of the Greek Emotional Intelligence Scale unbiased when measuring West- Germanic (English and Afrikaans) and Sotho (Northern Sotho, Southern Sotho, and Setswana) languages?

• What future recommendations can be formulated with regards to EI, emotion, and culture studies?

(23)

1.2 RESEARCH OBJECTIVES 1.2.1 General objective

The general objective of this study is to determine whether the Greek Emotional Intelligence Scale is an equivalent measure, which is free of item bias, when measuring a West-Germanic (English and Afrikaans) and Sotho (Northern Sotho, Southern Sotho, and Setswana) language groups.

1.2.2 Specific objective

The specific objectives of this research are:

• To determine whether EI can be conceptualised with regards to emotion and culture in a literature review?

• To establish if the Greek Emotional Intelligence Scale is reliable when measuring West- Germanic (English and Afrikaans) and Sotho (Northern Sotho, Southern Sotho, and Setswana) languages.

• To establish if the Greek Emotional Intelligence Scale is an equivalent measuring instrument when measuring West-Germanic (English and Afrikaans) and Sotho (Northern Sotho, Southern Sotho, and Setswana) languages.

• To determine if the items of the Greek Emotional Intelligence Scale are unbiased when measuring West-Germanic (English and Afrikaans) and Sotho (Northern Sotho, Southern Sotho, and Setswana) languages.

• To make future recommendations with regards to EI, emotion, and culture studies.

1.3 RESEARCH METHOD

The research method consists of a literature review and an empirical study. The results gathered are presented in the form of a research article.

(24)

1.3.1 Phase 1: Literature review

A complete review regarding EI, culture, bias, and equivalence is conducted. Relevant articles to the study are obtained by utilising computer searches on the internet. The following sources are consulted: Academic Search Premier; Business Source Premier; PsycArticles; PsycInfo; EbscoHost; Emerald; ProQuest; SACat; SaePublications; Science Direct; and Nexus. Various journals and books will also be utilised during the research of information relevant to the research topic:

• Journal of Cross-Cultural Psychology, • Journal of Managerial Psychology, • Journal of Positive Psychology, • South African Journal of Psychology, • Review of General Psychology, • Journal of Applied Psychology, • Journal of Organizational Behaviour, • Social Indicators Research,

• Management Dynamics,

• South African Journal of Industrial Psychology, • Administrative Science Quarterly,

• American Psychologist,

• Personality and Individual Differences, • Google scholar,

• Books

1.3.2 Phase 2: Empirical study

The empirical study consists of the research design, participants, measuring battery, and the statistical analysis. These are outlined below.

(25)

1.3.2.1 Research Design

This specific study is quantitative of nature. According to Creswell (2003), quantitative research is when the researcher primarily employs strategies such as experiments or surveys, and then collects data by means of predetermined instruments. Quantitative research is definite research, consisting of large representative samples (Struwig & Stead, 2001). With regards to the data collection, the procedure followed is structured in nature and will not be as informal as when sampling for a qualitative study. The research is cross-sectional, where numerous groups of people at a single point in time are examined (Salkind, 2009). Advantages regarding this approach are the time and economic significance.

1.3.2.2 Participants

For the purposes of this study, a combined random and stratified probability sample (N = 357) is extracted from the mining sector in South Africa. The participants are not blue collar workers. The focus of the research is on the white collar workers, in the organisation due to the literacy level of the participants. The inclusion criteria are mid-level managers and higher positions. The sample is conducted in the Gauteng and North-West provinces. The sample is stratified to include the West-Germanic language group (Afrikaans and English; n = 158) and the Sotho group (Northern Sotho, Southern Sotho, and Setswana; n = 199). It is essential that participants have a good command of the English language in order to complete the questionnaire in a successful manner.

Contact is established with the HR coordinator of different mining cites in the chroming, platinum, and gold mining industries in the North West and Gauteng provinces. After explaining the research to the HR co-ordinator, permission is requested to conduct the research. The coordinator has a very good relationship with the employees, and this relationship is utilised in order to group the employees in the training centre in accordance to their language. The reason for the research is then explained to the research participants. There the GEIS is administered, which the participants complete. It is then anonymously dropped in a box. Participation to the study is voluntary in nature; and the confidentiality and anonymity of participants is emphasised.

(26)

The different organisations’ that participate in the study are given comprehensive feedback, individually, regarding the results that are obtained.

1.3.2.3 Measuring Battery

The Greek Emotional Intelligence Scale and a biographical questionnaire are utilised in this study.

The Greek Emotional Intelligence Scale (GElS; Tsaousis, 2008) measures four basic emotional skills, namely: 1) expression and recognition of emotion, which corresponds to the capability of the individuals to express and recognise accurately their own emotional reactions ("I am unable

to explain my emotional state to others"; "I find it difficult to express my emotions to others"); 2)

control of emotions, which corresponds to the capability of the individuals to control and regulate emotions in themselves and others ("When I am under pressure I snap"; "I often get

angry and afterwards I find my anger inexcusable”); 3) use of emotions to facilitate thinking,

which corresponds to the capability of the individuals to harness their own emotions in order to solve problems through optimism and self-assurance, two emotional states that facilitate inductive reasoning and creativity ("I deal with my problems in a positive way by trusting

myself'; "I think of the positive side of things"); and 4) caring and empathy, which corresponds to

the willingness of the individual to help other people and his or her capability to comprehend another's feelings, and to re-experience them ("I am always willing to help someone who is

confronted with personal problems"; "1 like to talk with others about their problems"). The fifty-

two-item instrument demonstrated acceptable psychometric properties, which justified its use as a reliable and valid measure of emotional intelligence (Tsaousis, 2008). More specifically, the factor analytic data suggested a four-factor solution, which had a close similitude to Mayer and Salovey's (1997) theoretical framework. The Cronbach alpha coefficients for the four factors ranged between 0.80 and 0.92. All of the scales confirmed high internal consistency, indicating that they were homogeneous in their measurements. Furthermore, test-re-test data covering a four-week period indicated the temporal reliability of the GElS in that correlation coefficients ranged between 0.79 and 0.91 (Tsaousis, 2008). Also, according to Tsaousis (2008), data from

(27)

five different studies provided support for good convergent and discriminant validity of the GElS scales, suggesting that the test presented a fairly broad range of related emotional constructs, such as positive correlation with empathy, social skills, and well-being, as well as negative correlation with locus of control, negative affect, low physical and psychological well-being, and work stress. These findings justified the concurrent validation of the newly developed instrument, and the GElS will therefore be utilised in this study.

A biographical questionnaire is included to measure the characteristics of the participants. Internal dimensions (i.e. age, gender, and language), as well as external dimensions (i.e. qualification level).

1.3.2.4 Statistical analyses

The SPSS program and AMOS program is used to carry out the analysis of the data that is collected. Descriptive statistics is used in this study and it involves the testing of assumptions (Pallant, 2005) and it provides the researcher with a summary of the data he / she collects. The purpose of descriptive statistics is to provide the researcher with an overall, logical, and simple picture of the data that is collected (Struwig & Stead, 2007). Descriptive statistics used include the mean, standard deviation, skewness, and kurtosis and the alpha coefficient (Pallant, 2005). The mean is the sum of the observations that is made, divided by the number of observations that will constitute the group (Marascuilo & Serlin, 1988); in other words, it can be seen as the average. The standard deviation “measures the deviation of each score from the mean and the averages the deviations” (Struwig & Stead, 2007, p. 158). Skewness and kurtosis refers to the distribution of the scores (Struwig & Stead, 2007). “Skewness refers to the degree of deviation from symmetry, while kurtosis refers to how flat or peaked the distribution is (Struwig & Stead, 2007, p. 159). The degree of reliability is articulated by the Cronbach Alpha Coefficient; it ranges from 0.00 to 1.00 and the closer the alpha coefficient is to 1.00, the closer it will be to the true score (Struwig & Stead, 2007). The Cronbach Alpha Coefficient is acceptable when α > 0.70. SPSS is utilised to determine item bias by means of ordinal logistic regression. According to Kim (2001), there are various ways to detect Differential Item Functioning (DIF). DIF can be

(28)

defined as the interference of some demographic characteristic or grouping of the tight relationship between trait level and item responses (Crane, Gibbons, Jolley, & Van Belle, 2006). One of these methods is Zumbo‟s (2009) ordinal logistic regression approach. A variation of ordinal logistic regression to detect DIF between cultural groups is carried out by using several comparisons for different response categories on each item. When applying the regression procedure, outliers beyond the 95% confidence interval is used as DIF items. Crane, Gibbons, Jolley and Van Belle (2006) note that the ordinal logistic regression approach for testing DIF is not a complicated statistical analysis to accomplish. They further state that because of the ordinal LR framework, many demographic characteristics can be evaluated to determine whether items display DIF.

In order to determine whether the items of the GEIS are unbiased, a pre-test is conducted by means of various analyses of variance (ANOVA) on the four factors of the GEIS. An ANOVA shows the expression of the tests of interests in terms of variance estimates (Muller & Fetterman, 2002). In order to determine if there is a difference with regards to equivalence between the groups four, ANOVAS are conducted on each factor.

In the AMOS program (Arbuckle, 1997), confirmatory factor analysis and structural equation modelling methods is used to construct and test a four-factor model of emotional intelligence across language groups. Hypothesised relationships are tested empirically for goodness-of-fit

2

with the sample data. The X and several other goodness-of-fit indices summarise the degree of 2 correspondence between the implied and observed covariance matrices. However, the X test is widely recognised to be problematic (Joreskog, 1969); it is sensitive to sample size, and could also be invalid when distributional assumptions are violated, leading to the rejection of good

2

models or the retention of bad ones. Due to the drawbacks of X test, many alternative fit statistics have been developed. The Goodness-of-Fit Index (GFI) is commonly used to indicate the relative amount of variance and co-variance in the sample predicted by the estimates of the population. It usually varies between 0 and I, and a result of 0.90 or above indicates a good model.

(29)

The Comparative Fit Index (CFI) also compares the hypothesised and independent models, but considers sample size when doing so. The Tucker-Lewis Index (TLI) is a relative measure of covariation explained by the hypothesised model, which has been specifically designed for the assessment of factor models (Tucker & Lewis, 1973). In order to acquire good model fit, it is recommended for the NFI, CFI, and TLI to be acceptable above the 0.90 level (Bentler, 1992), although Hu and Bentler (1999) recommended a cut-off value of 0,95. The Root Mean Square Error of Approximation (RMSEA) estimates the overall amount of error; it is a function of the fitting function value relative to the degrees off freedom (Brown & Cudeck, 1993). Hu and Bentler (1999) suggest a value of 0,06 to indicate acceptable whereas MacCullum, Browne, and Sugawara (1996) suggested that values between 0,08 and 1,00 indicate mediocre fit and values above 1.00 poor fit.

In accordance with Hu and Bentler (1999), a combined approach is used to evaluate model fit. Following Hu and Bentler (1999), several fit indices are used to evaluate the fit of each CFA model; specifically an absolute close-fit index (RMSEA) and two incremental close-fit indices were chosen (TLI and CFI), as it has been argued that they provide more stable and accurate estimates than several other indices (Hu & Bentler, 1999; Maruyama, 1998). Further, these three indices have been used in other confirmatory factor analysis studies of emotional intelligence (Gignac et al., 2005). Other fit indices are included as support to the TLI, GFI, and RMSEA, as they were used in other studies for evaluation of psychological tests (Parker, Taylor, & Bagby, 2003) and provided easy comparisons to the other data sets.

1.3.2.5 Ethical considerations

The gathering of research data is an ethical practice. A code of moral guidelines on how to conduct research is provided so that the researcher can conduct the research in an ethical manner. In order for this research to be successful, a conscious awareness of fair and ethical practices must be utilised. Aspects such as informed consent, voluntary participation, protection from harm, confidentiality, and the protection of privacy are considered at all time (Salkind, 2009). The research proposal is submitted to the North-West University‟s ethical committee for

(30)

evaluation. The following guidelines are applicable at all times to retain an ethical environment (Struwig & Stead, 2001):

• At all times the researcher will endeavour to be honest, fair, and respectful towards the participants, and not attempt to mislead or deceive the research participants.

• The researcher is also sensitive to individual differences among people, such as age, ethnicity, religion, language, and socioeconomic status.

• The rights and dignity of the participants will at all times be respected and protected. This includes respecting their rights of privacy, confidentiality, and autonomy.

• The researcher will never discriminate against people on the basis of such afo rementioned factors.

• The welfare of others is extremely important. Therefore, no harm must ever come from the interaction between the researcher and the participants.

1.4 CHAPTER DIVISION

The chapters in this mini-dissertation are presented as follows: Chapter 1: Introduction and problem statement.

Chapter 2: Research article.

Chapter 3: Conclusions, limitations, and recommendations.

1.5 CHAPTER SUMMARY

Chapter 1 discusses the theory regarding measurement instruments, arguing that they should be equivalent and unbiased when used in cross-cultural studies. Emotional intelligence (EI) as an industrial psychology concept has grown tremendously over the past decade and the advantages that it poses has been recognised. There is, however, one issue not mentioned and studies enough and this is the extent to which EI is a culturally relevant concept. The presumption that emotions can be spoken of across various cultures cannot be made, and EI measurements therefore need to pay attention when measuring across various groups of culture whom perceive emotions differently. Language can be viewed as a vehicle of culture, and emotions are shaped within

(31)

specified cultural groups speaking languages that are of the same origin of that culture. The Greek Emotional Intelligence Scale (GEIS) has only been used in one other study previously in SA, but this study was not on equivalence. Two language families will be tested in this study namely the West-Germanic (Afrikaans and English) and the Sotho (Northern Sotho, Southern Sotho, and Tswana) group.

(32)

REFERENCES

Akerjordet, K. (2009). Emotional intelligence: Part 1-Development of scales and psychometric testing. Nursing and Health Sciences, 11(1), 58-63.

Arbuckle, J. (1997). AMOS user's guide version 3.6. Chicago, IL: Smallwaters.

Ashkanasy, N. M. (2003). Emotions organizations: A multi-level perspective. Research in

Multilevel Issues, 2, 9-54.

Badenhorst, S. (2010). Descent of Iron Age farmers in Southern Africa during the last 2000 years. African Archaeological Review. 27, 87-106.

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

Bar-On, R (2000). Emotional and social intelligence: Insights from the Emotional Quotient Inventory. In R. Bar-On, & J. D. A. Parker (Eds.), The handbook of emotional intelligence:

Theory, development, assessment and application at home, school and in the workplace (pp.

363-388). San Francisco, CA: Jossey-Bass.

Bentler, P. M. (1992). On fit of models to covariances and methodology to the Bulletin.

Psychological Bulletin, 112, 400-404.

Bierbrauer, G., Meyer, H., & Wolfradt, U. (1994). Measurement of normative and evaluative aspects in individualistic and collectivistic orientations: The Cultural Orientation Scale (COS). In U. Kim, H. Triandis, C. Kagitcibasi, S. C. Choi, & G. Yoon (Eds.), Individualism

and collectivism: Theory, method, and applications (pp. 189-194). Thousand Oaks, CA: Sage.

Boon, M. (2007). The African way: The power of interactive leadership. Cape Town, South Africa: Zebra Press.

Botma, N. (2009). Emotion experience, emotional intelligence and well-being in South Africa (Unpublished doctoral thesis). North West University, Potchefstroom.

Byrne, B. M. (2001). Structural equation modelling with AMOS: Basic concepts, applications

and programming. Mahwah, NJ: Erlbaum.

Brown, M. W., & Cudeck, R. (1993). Alternative ways of assessing model fit. In K. A. Bollen, & J. S. Long (Eds.), Testing structural equation models (pp. 445-455). Newbury Park, CA: Sage Publications.

(33)

Chemiss, C. (2000, April). Emotional intelligence: What it is and why it matters. Paper presented at the Annual Meeting for the Society for Industrial and Organizational Psychology, New Orleans, Louisiana.

Cole, N. S., & Moss, A. (1989). Bias in test use. In R. L. Linn (Ed.), Educational measurement (3rd ed), (pp. 201-219). Washington, DC: American Council on Education & National Council on Measurement in Education.

Conte, J. M. (2005). A review and critique of emotional intelligence measures. Journal of

Organizational Behaviour, 26, 433-440.

Conte, J. M., & Dean, M. A. (2006). Can emotional intelligence be measured? In K. R. Murphy (Ed.), A critique of emotional intelligence (pp. 59-80). Mahwah, NJ: Erlbaum.

Crane, P. K., Gibbons, L. E., Jolley, L., & Van Belle, G. (2006). Differential item functioning analysis with ordinal logistic regression techniques: DIF detect and dif with par. Medical

Care,

44, S115–S123.

Creswell, J. W. (2003). Research design: Qualitative, quantitative, and mixed approaches. Thousand Oaks, CA: Sage.

De Klerk, M., & Mostert, K. (2010). Work-home interference: Examining socio-demographic predictors in the South African context. SA Journal of Human Resource Management / SA

Tydskrif vir Menslikehulpbronbestuur, 8(1), 7- 10.

Gelfand, M. J., & Brett, J. M. (2004). The Handbook of negotiation: Theoretical advances and

cultural perspectives (pp. 3-6). Palo Alto, CA: Stanford University Press.

Ghorbani, N., Bing, M., Watson, P., Davison, H., & Mack, D. (2002). Self-reported emotional intelligence: Construct similarity and functional dissimilarity of higher-order processing in Iran and the United States. International Journal of Psychology, 37, 297-308.

Gignac, G. E., Palmer, B. R., Manocha, R., & Stough, C. (2005). An examination of the factor structure of the Schutte Self-Report Emotional Intelligence (SSREI) scale via confirmatory factor analysis. Personality and Individual Differences, 39, 1029-1042.

Gignac, G. E., & Ekermans, G. (2010). Group differences in EI within a sample of black and white South Africans. Personality and Individual Differences, 49(6), 639-644.

Goldenberg, I., Matheson, K., & Mantler, J. (2006). The assessment of emotional intelligence: A comparison of performance-based and self-report methodologies. Journal of Personality

(34)

Goleman, D. (1995). Emotional intelligence. New York, NY: Bantam.

Goleman, D. (2005). Emotional intelligence: Why it matters more than IQ. New York, NY: Bantam Dell.

Goliath-Yarde, L., & Roodt, G. (2011). Differential item functioning of the UWES-17 in South Africa. SA Journal of Industrial Psychology/SA Tydskrif vir Bedryfsielkunde, 37(1), 1-11. Gregory, R. J. (2007). Psychological testing (5th ed.). Boston, MA: Pearson Education Group. Hofstede, G. H. (2001). Culture’s consequences: Comparing values, behaviours, institutions and

organizations across nations (2nd ed.). Thousand Oaks, CA: Sage Publications .

Hofstede, G. (2003). Geert hofstede cultural dimensions. Retrieved from http://www.geert- hofstede.com/

Hu, L. T., & Bentler, P. M. (1999). Cut-off criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modelling: A

Multidisciplinary Journal, 6(1), 1-55.

Human, L. (2005, February). Current issues and practices for global diversity. Paper presented at the American Society for Training and Development‟s global conference and exhibition, Cape Town, South Africa.

Ilangovan, A., Scroggins, W. A., & Rozell, E. J. (2007). Managerial perspectives on emotional intelligence differences between India and the United States: The development of research propositions. International Journal of Management, 24, 541-548.

Jonker, C. S. (2002). The compilation and evaluation of a development program aimed at

emotional intelligence (Unpublished doctoral thesis). Potchefstroom University for Christian

Higher Education, Potchefstroom, South Africa.

Jonker, C. S., & Vosloo, C. (2008). The psychometric properties of the Schutte Emotional Intelligence Scale. South African Journal for Industrial Psychology, 34, 21-30.

Joreskog, K. G. (1969). A general approach to confirmatory maximum likelihood factor analysis.

Psychometrika, 34, 183-202.

Joseph, D. L., & Newman, D. A. (2010). Emotional intelligence: An integrative meta-analysis and cascading model. Journal of Applied Psychology, 95, 54-78. doi 10.1037/a0017286 Kagitcibasi, C. (1994). A critical appraisal of individualism and collectivism: Toward a new

(35)

Individualism and collectivism: Theory, method, and applications (pp. 52-65). Thousand Oaks, CA: Sage.

Kang, S., & Shaver, P. R. (2003). Individual differences in emotional complexity: Their psychological implications. Journal of Personality, 54, 757-765.

Keele, S. M., & Bell, R. C. (2008). The factorial validity of emotional intelligence: An unresolved issue. Personality and Individual Differences, 44, 487-500.

Kim, S. H. (2001). An evaluation of a Markov chain Monte Carlo method for the Rasch model.

Applied Psychological Measurement, 25, 163-176.

Libbrecht, N., Lievens, F., & Schollaert, E. (2010). Scale across self and other ratings measurement equivalence of the Wong and Law emotional intelligence. Educational and

Psychological Measurement, 70, 1007.

MacCullum, R. C., Browne, M. W., & Sugawara, H. M. (1996). Power analysis and determination of sample size for covariance structure modelling. Psychological Methods, 1, 130-149.

Marascuilo, L. A., & Serlin, R. C. (1988). Statistical methods for the social and behavioural

sciences. New York, NY: Freeman.

Markus, H., & Kitayama, S. (1991). Culture and the self: Implication for cognition, emotion, and motivation. Psychological Review, 98, 224-253.

Matsumoto, D. (2002). Methodological requirements to test a possible in-group advantage in judging emotions across cultures: Comments on Elfenbein and Ambady and evidence.

Psychological Bulletin, 128, 236-242.

Matsumoto, D., Angua-Wong., & Martinez, E. (2008). Priming effects of language on emotion judgments in Spanish-English bilinguals. Journal of Cross-Cultural Psychology, 39, 335-342. Mayer, J. D., Caruso, D. R., & Salovey, P. (1999). Emotional intelligence meets traditional

standards for intelligence. Intelligence, 27, 267-298.

Mayer, J. D., & Caruso, D. R. (1999). Emotional intelligence meets traditional standards for an intelligence. Intelligence, 27(4), 267-298.

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

(36)

Mayer, J. D., & Salovey, P. (1997). What is emotional intelligence? In P. Salovey, & D. Sluyter (Eds.), Emotional development and emotional intelligence: Implications for educators (p. 3- 31). New York, NY: Basic Books.

Mayer, J. D., Salovey, P., & Caruso, D. R. (2000). Models of emotional intelligence. In R. J. Sternberg (Ed.), Handbook of human intelligence (pp. 396-420). New York, NY: Cambridge University Press.

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. (2004). A further consideration of the issues of emotional intelligence. Psychological Inquiry, 15(3), 249-255.

Meiring, D., Van de Vijver, A. J. R., Rothmann, S., & Barrick, M. R. (2005). Construct, item, and method bias of cognitive and personality tests in South Africa. South African Journal of

Industrial Psychology, 31(1), 1−8.

Mulaik, S. A., James, L. R, Van Altine, J., Bennet, N., Lindi, S., & Stilwell, C. D. (1989). Evaluation of goodness-of-fit statistics for structural equation models. Psychological Bulletin,

105, 430-445.

Ogarkova, A., Borgeaud, P., & Scherer, K. (2009). Language and culture in emotion research: a multidisciplinary perspective. Social science information, 48, 339-357.

Pallant, J. (2005). SPSS Survival Manual (2nd ed.). Buckingham, England: Open University Press.

Parker, J. D. A., Taylor, G. J., & Bagby, R. M. (2003). The 20-item Toronto Alexithymia Scale III: Reliability and factorial validity in a community population. Journal of Psychosomatic

Research, 55, 269-275.

Parker, J. D. A., Duffy, J. M., Wood, M., Bond, B. J., & Hogan, M. J. (2005). Academic achievement and emotional intelligence: Predicting the successful transition from high school to university. Journal of First-Year Experience, 17, 1-12.

Parker, J. D. A., Keefer, K. V., & Wood, L. M. (2011). Toward a brief multidimensional assessment of emotional intelligence: Psychometric properties of the emotional quotient inventory-short form. Psychological Assessment, 23(3), 762-777.

(37)

Petrides, K. V., & Furnham, A. (2003). Trait emotional intelligence: Behavioural validation in two studies of emotion recognition and reactivity to mood induction. European Journal of

Personality, 17, 39-57.

Pheifer, S. I. (2001). Emotional intelligence: Popular but elusive construct. Roeper Review,

23(3), 138–143.

Ratner, C. (2000). A cultural-psychological analysis of emotions. Culture and Psychology, 6, 5- 39.

Republic of South Africa (1998). Employment Equity Act No. 55 of 1998. Government Gazette,

400, 19370. Pretoria, South Africa: Government printer.

Russell, J. A., & Barrett, L. F. (1999). Core affect, prototypical emotional episodes, and other things called emotion: Dissecting the elephant. Journal of Personality and Social Psychology,

76, 805- 819.

Saklofske, D. H., Austin, J., & Minski, P. S. (2003). Factor structure and validity of a trait emotional intelligence measure. Personality and Individual Differences, 34, 707-721.

Salkind, N. J. (2004). Statistics for people who (think they) hate statistics (2nd ed). Thousand Oaks, CA: Sage Publications.

Schutte, N. S., Malouff, J. M., Hall, L. E., Haggerty, D. J., Cooper, J. T., Golden, C. J., & Domheim, (1998). Development and validation of a measure if emotional intelligence.

Personality and Individual Differences, 25, 167-177.

Sedmar, A., Robins, G., & Ferris, G. R. (2006). Comparing the validity of multiple social effectiveness constructs in the prediction of managerial job performance. Journal of

Organizational Behavior, 27, 443-461.

Shipper, F., Kincaid, J., Rotondo, D. M., & Hoffman, R. C. (2003). A cross-cultural exploratory study of the linkage between emotional intelligence and managerial effectiveness. The

International Journal of Organizational Analysis, 11, 171-19l.

Smit, P. J., & Cronje, G. J. (2011). Management principles: A contemporary edition for Africa. Cape Town, South Africa: Juta and Company Ltd.

Struwig, F. W., & Stead, G. B. (2001). Planning, designing and reporting research. Cape Town, South Africa: Pearson Education.

Structural equivalence and item bias of a self–report emotional intelligence measure in the mining industry

Declaration of originality of research

COMMENTS

SUMMARY

ACKNOWLEDGEMENTS

CONTENTS