THE VALIDITY AND RELIABILTY STUDY OF RACIAL/ETHNIC HARASSMENT AND DISCRIMINATION, AND DIVERSITY ASSESSMENT SCALES IN TURKEY’S POPULATION

Ertugrul Godelek 1 and Necip Yildirim 2 . 1. PhD, University of Mersin, Faculty of Science and Literature, Department of Psychology. 2. University of Mersin, Faculty of Science and Literature, Department of Psychology. ...................................................................................................................... Manuscript Info Abstract ......................... ........................................................................ Manuscript History Received: 2 August 2018 Final Accepted: 4 September 2018  Published: October 2018

When we analyse the social history, it is seen that biological and cultural differences have always brought along a discrimination in the mindscapes that determined the social behaviour. Certain people in the societies have more opportunities and power when compared to others. When social psychologists compared the resources, different people have in the USA, they emphasised the resources that affect life-earning capacities of the individuals, such as education, occupational prestige, and earnings. Determining the class of the people basing upon these is called stratification. The place of individuals in the stratification hierarchy is called the individuals' socioeconomic status. The socioeconomic status of individuals affects their life choices such as opportunities for living better, happiness, health and justice. The resources that differentiate individuals from each other are not only the socioeconomic resources such as education, occupational prestige and earnings. At the same time, it is the function of other personal treats such as sex, race, identity, religion, age, marital status, physical attributions and mental capacities. These features serve a symbolic function for other resources. For instance, being man and being white are sources of power in the American society, because white men keep a hold of most of the positions of power, in which most people related these features with power. It is not necessary to believe that white men have more power in their hands, it is sufficient to believe that only white men have achieved this. This kind of a belief affects the way an individual with a certain socioeconomic status is behaved. Similarly, advanced age might be the symbol of wisdom or despair; and youth might be the symbol of health and liveness or inexperience. Put it differently, our stereotypes ISSN: 2320-5407 Int. J. Adv. Res. 6(10), 144-153 145 about social categories refer to the members of these categories that symbolise the sources of power (Wiggins, et al., 1994).
Characteristics such as race, sex, sexual identity and age are not only individuals' sources of power of about themselves, but at the same time, they could affect the access to important socioeconomic resources. Individuals from certain race, sex, age, religion or physical capacity might be more advantageous with regard to education, occupational prestige and earnings. Race and sex are primary statuses. These statuses affect the individuals' own attitudes and behaviours as well as they affect how individuals see others and how they behave against others; the positions held are insignificant. Perhaps for this reason, race and sex are the categories in which social psychologists conduct research the most. All nations comprising with more than one ethnic groups have encountered ethnic conflicts. It is ironic that, even in the democratic nations, which claim to live in freedom, equality and justice, an ethnic group faces oppression and exploitation by the other. In reality, the dominant group in the society, generally, is not in favour of sharing equally the political and economic power and richness at hand with other ethnic groups.
When the literature on discrimination is reviewed, it is seen that the USA has been involved in discrimination, particularly racial discrimination since its first days. Certain minority groups in the USA have experienced many unequal treatments. The minority group can be defined as a group which is at a lower status in the society and which is exposed to discrimination (Zastrow, 1992). Kanter (1977) identified three perceptual tendencies about the minorities, which are few in number in a greater group: visibility, contrast and assimilation. Kanter (1977) found that these three perceptual tendencies have perceptible impacts in the experiences of the minorities, which are few in number. He argued that each tendency increased the amount and burden of stress for the minorities. Visibility could result in minority behaviours to be more social and more attention grabbing. This increases the pressure on the minorities to observe their own behaviours and their interaction in the workplace. Contrast results in the majority group to see the minorities as an external group. The majority group behaves in a way that asserts the majority perspective or that draws attention to the differences between the majority culture and the minority culture. Assimilation, according to Kanter, tends to force the minority group to play limited and caricaturised roles or results in the majority group to pattern their stereotypes on the minority group and to approach the minority groups in line with these stereotypes.

Institutional Racism and Sexism
Discrimination can be institutional or individual. Only one individual can be the source of discrimination. For instance, if one administrator among many administrators responsible for the suggestion of salary rise for lower level employees in a company, routinely proposes higher salaries for white male employees compared to the Afro -American workers with the same qualifications, it means that he or she, as an individual, is discriminating based on race. On the other hand, if most of the administrators of a company are doing the same thing routinely, this company most probably is discriminating based on race. The spreading of discrimination in the company indicates that the company has a racial discrimination policy (Wiggins, et al., 1994).
Institutional racism is an institutional discrimination based on race. Similarly, institutional sexism is an institutional discrimination based on sex. Both include social regulations that hamper the minority group members, and patterns used by a dominant group. According to many sociologists ( A group of researchers (Miller, 1986;Martin et al.1989;Martin, 1992) argue that the organisations not only reflect the racial, ethnic and sexual inequalities of the societies but, at the same time, they create these inequalities (Wiggins et al., 1994). The organisations are practising this by creating ideologies, norms and activities within theirselves that reveal inequality again and again (Wigginset al., 1994).

Prejudice and Discrimination
Gordon Allport(1954) defines prejudice as follows: "Prejudice is to make negative judgements about others without thinking enough." (Zastrow, 1992). This definition includes elements such as an unfounded judgement and belittlement, dislike, fear and a repulsive emotion. With regard to race, prejudiced people attribute their racial 146 prejudice to almost everybody. Racial prejudice originates from the beliefs that people differ with regard to skin colour and other physical characteristics, and at the same time, with regard to behaviours, values, cognitive functions and attitudes.

Stereotypes
Many people relate different characteristics with the race and sex of the individuals. In other words, people, when considering groups, appraise members of certain groups more positively. Stereotypes are belief schemas that attribute a series of characteristics to all members of a social identity (Wiggins et al., 1994). A stereotype can be defined as "a set of beliefs of an individual about the characteristics of a group" (Judd et al., 1993; cited by Ruggiero, 1999;Hartley, 1946). Stereotypes are not supposed to be correct, although they are required. Stereotypes discard some important information and it is certain that many of the members of a social identity do not fit in the stereotypes.

Racial and Ethnic Stereotypes
Racial and ethnic stereotypes involve attributing a static and generally untrue or unpleasant notion to a racial or ethnic group. Stereotypes may involve truth to some degree but they are mostly exaggerated, decontextualized or distorted. Since people try to perceive and understand everything via categories, stereotypes are closely related to the thought patterns of individuals. People need categories to relate similar thing and to group them. People have many stereotypes about numerous categories such as their mothers, their fathers, their teachers, the farmers, the mechanics, Turks, etc. These categories may include useful and true information about a member in those categories. However, each member of a category may have various characteristics that are not suggested by the stereotype, and even may have characteristics contradicting some stereotypes.

Difference between Individual Discrimination and Group Discrimination
Allport (1954), in "The Nature of Prejudice", claimed that the members of the minority group rejected discrimination to protect themselves against discrimination (Ruggiero, 1999). Allport(1954) asserted that although the group members as a whole easily accepted the discrimination against them, they generally rejected their experience of discrimination or they palliated these incidents (Ruggiero, 1999).
The individual/group discrimination difference was proven by Cosby (1982) in a study on working women around Boston. Since this study, perception of discrimination as a group, rather than an individual, has emerged among certain minority groups (Ruggiero, 1999). For instance, studies were conducted on groups such as south Asian and According to the 5-stage model of intergroup relations by Taylor and McKiman (1984), members of the minority group are only interested with common action when they realise that they have been exposed to same amount of discrimination with other members of the group. In addition, when minority group members do not care about their personal experiences about their group being discriminated, they could not take the necessary steps to improve their individual statuses or their group's status (Ruggiero, 1999).
Taylor et al. (1996) introduced a "2-factor explanation" in this respect. Accordingly, two processes may underline the individual/group discrimination difference. When minority members focus on group level discrimination while discarding their personal experiences about discrimination, they could be prejudiced (Ruggiero, 1999). Crosby (1984) who argued that minority group members rejected or at least disregarded their personal experiences about discrimination, asserted that these members could abstained from pointing out the discriminator implicitly and thus tended towards rejecting their personal experiences (Ruggiero, 1999). In addition, other researchers such as Ruggiero and Taylor (1995, 1997) asserted that minority group members disregarded individual discrimination because they wished to believe that they had control on the consequences of this (Ruggiero, 1999).

Method:-Sample
The sample of the study comprises of 500 participants working in five different workplaces, in which at least 30 people work. The demographic characteristics of the individuals such as age, sex, education status were not taken into consideration since this study is a reliability and validity study.

Data Collection Tools Racial/Ethnic Harassment and Discrimination Scale
Racial/Ethnic Harassment and Discrimination Scale (REHD) was developed and first used by Schneider, Hitlan and Radhakrishnan (2000), and Ormerod, Bergman, Palmieri, Drasgow and Juraska (2001). This scale, comprising of 40 items, questions the racial/ethnic harassment experiences of individuals in the last 12 months. The first 14 items of this scale are evaluated in the form of 4 points Likert scale. They have four categories as "Never", "Once or Twice", "Sometimes", and "Often", and these response categories are assessed between 0 point and 3 points. The remaining 26 items question whether individuals experienced racial/ethnic discrimination in the last 12 months. The response categories of these last 26 items are as "No, nothing like this has ever occurred", "Yes, but not because of my race", "Yes, it was because of my race". These 26 items are assessed as zeroes and ones, "Yes = 1" and "No = 0". The score that can be obtained from the first 14 items of the scale varies between 0 and 42. The score that can be obtained from the remaining 26 items of the scale varies between 0 and 26.

Diversity Assessment Scale
Diversity Assessment Scale (DAS) was developed by McArthur (1999). In the first section of the scale, the definitions of difference, discrimination and minority group are provided to inform the respondents of the scale. The original survey form of the scale comprises of three sub-scales as Section A -Stages of Change, Section B -Corporate Environment, and Section C -Personal Behaviour and Beliefs. In the first section of the scale, respondents are provided with six workplaces examples scored between 1 point and 6 points, and they are requested to give points to their own workplace between 1 point and 6 points. In the second section, there is a scale comprising of 15 items aiming at learning the experiences and beliefs about corporate environment. In the third section, there is another scale comprising of 15 items aiming at learning the personal behaviours and beliefs of the respondent. The scales in the last two sections are 5-point Likert scales with "Strongly Disagree", "Disagree", "Neutral", "Agree" and "Strongly Agree" as options.
The score to be obtained from the first section of the Diversity Assessment Scale varies between 1 point and 6 points. The score that can be obtained from the second and third sections each varies between 15 and 75 points.

Procedure
These two scales, after being translated from English into Turkish, were evaluated by 10 people who had full command of English and Turkish, and the statements with the greatest consensus were selected to comprise the Turkish translations of the scales. Later, these scales were distributed 10 workers from varying demographic characteristics, to see whether the statements could be understood, and the scales were given their final form with the feedback collected.

Analysis of the Data
The Statistical Software Package (SPSS) was used to analyse the data obtained from the sample group. The reliability analyses of the first 14 items of the Racial/Ethnic Harassment and Discrimination Scale and the second and third sections of the Diversity Assessment Scale were conducted in SPSS. However, the analysis of the last 26 items of the Racial/Ethnic Harassment and Discrimination Scale was conducted by the researchers themselves since these items were scored as 0 or 1. The Cronbach' Alpha was calculated for the 14 first 14 items of the Racia l/Ethnic Harassment and Discrimination Scale and the second and third sections of the Diversity Assessment Scale, and the Kuder-Richardson internal consistency coefficient was calculated for the last 26 items of the Racial/Ethnic Harassment and Discrimination Scale. At the same time, the item-total test correlations were assessed using the Pearson Product Moments method and the significance of the items was tested. For the validity study, all scales were tested with t-test using the traditional split half method.

148
The Limitations of the Study The procedure required for a research has been followed in this study; however, the number of participants sufficient for a validity and reliability study could not be achieved. The reason for this was that some workers in some workplaces refused to fill out some scales due to certain concerns.

Findings Validity
To understand the degree to which Racial/Ethnic Harassment and Discrimination Scale and the Diversity Assessment Scale measure what they claim, the construct validity was conducted on all the data collected. To provide information to the construct validity, the "rotation of principal components analysis" was used. The compliance of the data for the factor analysis was tested with the Kaiser-Meyer-Olkin (MKO) coefficient and Bartlett's test for Sphericity (Hair et al., 2006;Howell, 1982;Roscoe, 1975;Walker et al., 1969.). KMO is a statistical method used to determine whether data and sample size are appropriate and sufficient for the selected analysis. If the KMO converges to 1, it means the data are appropriate for the analysis, and if it is 1, it means a perfect fit is present.
According to the analysis, the KMO value was found 0.90 for the Racial/Ethnic Harassment and Discrimination Scale, 0.91 for the Diversity Assessment Scale Section-B, and 0.97 for the Diversity Assessment Scale Section-C. In order to use the parametric method, the measured feature should exhibit normal distribution in the sample. Bartlett Sphericity test is a statistical technique that can be used to control whether the multivariate data come from a normal distribution or not.
The Chi-square results being statistically significant, at the end of this test, is an indicator that the data are from a multivariate normal distribution. At the end of the analysis in the study, the Bartlett test was found statistically significant for all three scales(for Racial/Ethnic Harassment and Discrimination Scale 2 = 3617,52; p<0,01; for Diversity Asssessment Scale Section-B 2 = 4004,11; p<0,0; for Diversity Asssessment Scale Section-C 2 = 4871,44; p<0,01).
To attain a statistically significant construct on the participants' tendencies factor analysis was conducted to reveal the structure or structures, which are called the factors or components, measured by the scale items ( At the end of the factor analysis, the Racial/Ethnic Harassment and Discrimination Scale comprising of 40 items formed a single factor construct with the eigenvalue over 1. In this 40-item single factor construct, the eigenvalue belonging to the factor providing information about the significance level and weight was found 6.61. This factor, alone, explains the 73.81% of the total variance in the Racial/Ethnic Harassment and Discrimination tendency. The factor loads for the items comprising the scale vary between 0.50 and 0.98 (Table-1). All these findings are used as a proof that the scale has satisfactory construct validity. The item test correlations were calculated with regard to the item validity and homogeneity of the scale. As a result, the item test correlations of the scale vary between 0.47 and 0.91. All these findings are considered as a proof that the scale items are valid and they all measure the same construct (Hair et al., 2006;Howell, 1982;Knoke et al., 1994;Roscoe, 1975;Walker et al., 1969.)  At the end of the factor analysis, the scale in the Diversity Asssessment Scale Section-B comprising of 15 items formed a single factor construct with the eigenvalue over 1. In this 15-item single factor construct, the eigenvalue belonging to the factor providing information about the significance level and weight was found 7.72. This factor, alone, explains the 77.14% of the total variance in the Diversity Asssessment Scale Section-B tendency. The factor loads for the items comprising the scale vary between 0.66 and 0.93 (Table-2  At the end of the factor analysis, the scale in the Diversity Asssessment Scale Section-C comprising of 15 items formed a single factor construct with the eigenvalue over 1. In this 15-item single factor construct, the eigenvalue belonging to the factor providing information about the significance level and weight was found 8.67. This factor, alone, explains the 79.63% of the total variance in the Diversity Asssessment Scale Section-C tendency. The factor loads for the items comprising the scale vary between 0.66 and 0.93 (Table-3). All these findings are used as a proof that the scale has satisfactory construct validity. The item test correlations were calculated with regard to the 150 item validity and homogeneity of the scale. As a result, the item test correlations of the scale vary between 0.44 and 0.84.
All these findings are considered as a proof that the scale items are valid and they all measure the same construct (Hair et al., 2006;Howell, 1982;Knoke et al., 1994;Roscoe, 1975;Walker et al., 1969.)

Reliability
A reliability study was conducted to understand the degree to which Racial/Ethnic Harassment and Discrimination Scale and the Diversity Assessment Scale measure what they claim. In this respect, the Cronbach's Alpha reliability test, item-test correlation analysis and the comparison of the 27% lower-upper group were conducted with regard to the reliability and homogeneity of the scale. All these findings are presented in Tables 4, 5  ).Scores equal to or higher than 0.30 in the interpretation of the item total correlation, prove that they differentiate the individuals better with regard to the measured feature. The item total correlations of the Racial/Ethnic Harassment and Discrimination Scale are between 0.47 and 0.91. This indicates that the items are distinctive with regard to the Racial/Ethnic Harassment and Discrimination tendency. The 27% lower-upper group analysis was found statistically significant for all items. All these findings are used as a proof that the Racial/Ethnic Harassment and Discrimination Scale has satisfactory reliability. The Cronbach's Alpha reliability of the Diversity Assessment Scale Section B is 0.89. The Kuder -Richardson internal consistency coefficient of the Diversity Assessment Scale Section B was found .97 (KR-20=.97).Scores equal to or higher than 0.30 in the interpretation of the item total correlation, prove that they differentiate the individuals better with regard to the measured feature. The item total correlations of the Diversity Assessment Scale Section B are between 0.38 and 0.93. This indicates that the items are distinctive with regard to the Diversity Assessment Scale Section B tendency. The 27% lower-upper group analysis was found statistically significant for all items. All these findings are used as a proof that the Diversity Assessment Scale Section B has satisfactory reliability. .39 59.12*** 5.
.40 60.12*** *p<.05 *p<.01 ***p<.001 The Cronbach's Alpha reliability of the Diversity Assessment Scale Section C is 0.96. The Kuder -Richardson internal consistency coefficient of the Diversity Assessment Scale Section C was found .90 (KR-20=.90).Scores equal to or higher than 0.30 in the interpretation of the item total correlation, prove that they differentiate the individuals better with regard to the measured feature. The item total correlations of the Diversity Assessment Scale Section C are between 0.40 and 0.93. This indicates that the items are distinctive with regard to the Diversity Assessment Scale Section C tendency. The 27% lower-upper group analysis was found statistically significant for all items. All these findings are used as a proof that the Diversity Assessment Scale Section C has satisfactory reliability (Hair et al., 2006;Howell, 1982;Knoke et al., 1994;Roscoe, 1975;Walker et al., 1969.).

Discussion and Conclusion:-
In this study, the validity and reliability tests of two scales about a topic that has not been yet studied in Turkey that should not be disregarded, i.e. workplace discrimination was conducted. At the end of the analyses conducted, the reliability coefficients of all subscales of these two scales are found high. Item 31 of the Racial/Ethnic Harassment and Discrimination Scale has a statistically significant correlation but it is problematic. This item is expressed as "I or my family could not get appropriate medical care." Similarly, Item 26 of the Racial/Ethnic Harassment and Discrimination Scale has a high level of significance, but it has been problematic during the implementation. This item is expressed as "I did not have any professional relationship with somebody would provide guidance to me to improve my career or to become specialized." Items 31 and 26 have made an impression that these items are not understood fully and accurately by the participants or the questions are not stated appropriately, although they are statistically significant. Another reason for Item 31 to be considered problematic is that the participants feel themselves under health coverage since the scale was implemented in workplaces.
The original form of the Racial/Ethnic Harassment and Discrimination Scale was developed to be used in the US army. Therefore, some items were changed to be adapted into Turkish society. For instance, Item 24 in the original form of the scale is expressed as "My current assignment is not good for my career if I continue in the military." This expression was generalised more and expressed as "I will not be able to advance in my career, if I continue this job." A similar situation is valid for Item 32. Item 32 in the original form of the scale is expressed as "I or my family got poorer military support service (for example, at commissaries, exchanges, clubs, and rec centres) than others did." This item also was generalised and expressed as "I or my family received less social support when compared to others." Section A -Stages of Change, in the original form of the Diversity Assessment Scale was not analysed in this study; because this section was not filled out by most of the participants. The researchers think that either the participants missed out this section or they left empty after getting bored from reading. Therefore, the evaluations were made deeming the Diversity Assessment Scale as comprising of two sections.
The data on the Section B -Corporate Environment subscale of and Section C -Personal Behaviours and Beliefs subscale of the Diversity Assessment Scale have given us the idea that both subtests function sufficiently. It is certain that some of the expressions in the items are problematic. However, as it is mentioned in the Findings section, very satisfactory results were obtained in general.
.73 99.12*** *p<.05 **p<.01 ***p<.001 153 Consequently, both scales were adapted successfully for the Turkish sample. However, these two scales need to be implemented with different samples to excel.

Suggestions
It is thought that this study could be further improved by conducting the analyses with more participants. In addition, it is possible to obtain challenging results by considering variables such as sex, age, education status and socioeconomic status.
This study is the first step taken to determine the pattern of discrimination in Turkey. The issue of discrimination, which is deemed as a significant source of stress for both the society in general and for the employees in the workplace in particular, could only be solved after identifying the problem. The discrimination experiences in Turkey seem to occur less overtly and intensely than the US, however, this is not a proof that discrimination does not exist and it will not occur. Therefore, developing policies to prevent discrimination would only be possible by the scientific study of the issue.