The Development Of A Scale To Measure Whether University Of Hertfordshire Student’s Attitudes Towards Cross Cultural Relationships.docx

Lab Report Feedback Sheet Research Methods for Psychologists 2

Lab Title: Student ID: Marker:

Date:

Overall Report Grade

How to Improve: Within Specific Sections Title, Abstract and Introduction

Method

Results

Discussion

References and Appendix

Summary of the main strengths of your report:

General Themes in Report Writing Writing: Academic format and a clear writing style Content: Relevant material supported by references in text Comprehension: Accuracy and conveys complex understanding

The development of a scale to measure whether University of Hertfordshire student’s attitudes towards cross cultural relationships, both romantic and friendly, differ according to gender.

Abstract:

The development of a scale to measure whether University of Hertfordshire student’s attitudes towards cross cultural relationships, either romantic or friendly, differ according to gender.

Attitudes can be simplistically explained as a stance or position towards a certain view, event or experience and as suggested by Hogg and Vaughn (2011) it is an evaluation of different issues or aspects that come along they way in ones’ life. Moreover, as Alport (1935) suggests attitudes are a form of acknowledged ideas and experiences which are shaped and predict an individual’s reaction and stance opposite a certain issue or event. It is indeed a gradual evaluation of ideas and thus being the aftermath of ‘subjective experiences’ (Eiser, 1986). As a notion many psychologists (Thurstone, 1931; Alport, 1935; Rosenberg & Hovland, 1960; Himmelfarb &Eagly, 1973; Eagly & Chaiken, 1993), tried to place attitudes into a certain model to try and explain how they are formed and in general give clarity upon the matter. Alport (1935) came about with the twocomponent attitude model, which suggested that disliking or liking a certain object has to be enhanced through the mental stability and readiness of doing so. The three-component attitude model was introduced by Himmelfarb & Eagly (1973), which suggested that an attitude is formed through a generic stimulus examination, observation and acknowledgement, which then passes through three major components; affective response, cognition, behaviour. This means suggest that an individual has to undergo emotional and affective responses towards the object or issue, acknowledge his beliefs and opinions on the matter and moreover how it will affect his intentions of action (Eiser, 1986). Moreover, the three-component attitude model can lead to a inferred measurement of an attitude as the reaction on stimulus can be observed and all three components likewise, which thus can bring out a certain generic assumption of ones’ attitude upon a certain matter.

According to Katz (1960), attitudes provide knowledge and defence to one’s ego, which agrees with the notion that attitudes are permanent, and not just a ‘phase’ about a certain issue (by Himmelfarb & Eagly, 1973). This notion leads to the theory of reasoned action and the theory of planned behaviour, which was developed by Fishbein and Ajzen (1974). They decided to mend the two theories in one to emphasise how a positive or negative attitude towards an issue may affect an individual to later act on a certain semantic matter. The theory stresses how the intent of a behavioural action is a cumulative result of the subjective norms of the individual, the attitude towards the behavioural action and the perception of the action’s control (Hogg & Vaughn, 2011). All of these characteristics of acknowledging the attitudes, and perceived intentions, may lead to the behavioural intention at the right time and with the appropriate cross of the action’s control line, thus the action will occur. Regarding the choice of the most appropriate design for an attitude scale, the type of measurement that will be used is another significant issue. Methods such as, observational methods, interviewing and indirect measures such as Rochbach’s Type Ink Blot, can be considered as options. However, these methods may take a lot of time to be completed and many not give the appropriate result the experimenter is looking for due to the lack of reliability and validity as they open to interpretation. On the other hand, self-competition measurements may be the most efficient method to use to measure attitudes. The Semantic Differential Scale (Osgood, Suci & Tannenbaum, 1957) that allow the subject to complete a statement given by the experimenter, may lead to issues of a limited amount of adjectives available to select in order have a sufficient survey. Moreover, the Thurstone survey (1928), is a type of survey that contains a list of statements which allow the subject to select a ‘yes’ or ‘no’ in terms of preference in the context of the statement. This type of survey may be advantageous in the sense of receiving the appropriate interval data that is needed; however it may be overtly time consuming. The main concern of the following survey was the development of an efficient scale. In order for this to be achieved, precautions were taken. By selecting a between subjects design this enabled the survey of cross-cultural relationships to compare the attitudes of males and females in the University of Hertfordshire.

The hypothesis of the survey is that there will be a difference between the attitudes towards cross-cultural relationships between males and females in the University of Hertfordshire. In order to do so, a Likert scale (1932) was developed. The Likert Scale was appropriate as it is a self-completion scale, which allows a 5-point scale of favour or not a statement that was presented. To ensure that the survey’s results will be reliable and efficient, validity was taken into account. Face and content validity as well as construct validity are important criteria upon which they indicate whether the survey is measuring what it is suppose to be measuring (Oppenheim, 1997). In order for the survey to measure the attitudes towards cross-cultural relationships, the sets of items set must have a certain continuum and measure what they are suppose to measure, signifying the balance of negative and positive statements. Furthermore, internal constituency is another measure that was taken into consideration. This notion emphasises the homogeneity of the statements and how far they are measuring the main factor of the survey (DeVellis, 2003). To ensure that this is occurring in the survey, Cronbach’s alpha coefficient (1951) is calculated, which shows interrelated variance between each item, and the variance towards the main factor of the survey, which is being measured. Moreover, to ensure the reliability of the survey, a test-retest measure is taken (DeVellis, 2003). Retesting a smaller sample for the initial set of respondents, may signify the consistency of responses as the results of the second survey produced after ‘cleaning’ it from items which are low in variance may have a higher score, thus having a positive reliability. Therefore, these measures are taken into account to test our prediction that there will be a difference between males and females in the University of Hertfordshire on a measurement of attitudes on cross-cultural relationships.

Method: Participants: The development of the survey was collaboration between five second-year undergraduate Psychology BSc students. The survey was completed by 50 respondents, which consisted of 10 for each researcher. Among these 50 respondents, 25 were male and 25 were female, while the ages were between 18 and 24 with four outliers of ages between 25-33. The sample was randomly selected in the Learning Resources Centres of the University of Hertfordshire, as allocation and completion of the survey was efficient. Out of the 50 participants, 5 males and 5 females were re-approached to complete the final survey as a testretest method, to test external reliability.

Design: The purpose of the survey was to develop scales that would be tested in an analytic survey using the Principle Components Analysis (PCA). The PCA was used to filter out any statements of the survey that were not testing our main factor. This was done to ensure the validity of the survey. Based on our hypothesis, the independent variable (IV) of the study was the gender of the population tested, male and female, and the dependent variable (DV) was the mean total score of the attitude through the final scale produced measuring attitudes towards cross-cultural relationships. After the scale development process through the PCA, a between-subjects design was used to test the hypothesis set for the survey. This was due to the comparison of the categorical variables, male vs. females. The prediction for the survey was that, there would be a difference between males and females, in the University of Hertfordshire, in a survey on cross-cultural relationships.

Materials: Scale Construction Process and Scale Processing: The process began by brainstorming various types of attitudes and discussed the pros and cons of testing each one. Resulting to a relationship oriented item pool; the factor was decided to be cross-cultural relationships, due to the University’s vast amount of international students. In developing 20 optimum statements for the survey, some criteria were taken into account. The statements had to be 10 negatives and 10 positives, and had to be completely unambiguous for the participant to clearly understand each statement. Moreover, the topic being about cross-cultural relationships, there was a fine line of having a survey which did not push any buttons concerning racial differences, leading it to be misinformed as a racism questionnaire. Therefore, the statements were produced carefully considering the feelings and thoughts that may arise with each statement. A base line statement that was on the survey, which has a role of a face valid statement, is for example “I enjoy attending cross-cultural social events”. After developing 20 statements, an R-Matrix of all the items was produced to scan for correlations between the responses of the participants. Any correlations with a value below 0.30 and above 0.50 were not used. It was decided through the correlations of the items, that item 6, 9 and 12 of the survey had to be excluded. These statements were “I enjoy eating food from different cultures”, “I do consider other peoples’ views and beliefs of culture” and “My friendship group is mainly made of different ethnic groups”. These three statements had a cultural orientated context, which proved to have low correlation with any other item in survey as they may have been set to measure a different factor than the formal factor. After these steps, PCA was used in order to extract all the items that were inter-correlating and correlating with the factor 1. This process was done to validate the survey, and increase internal and external reliability and consistency between the items stated and question actually measured in the survey. To ensure this, the PCA was run twice and also a test-retest to confine the findings to perfection.

Initially through PCA, a Scree plot (see Appendix A) was used developed with the use of the responses in the survey. The Scree plot would indicate the different factors that are being measured in the survey and would enable us to choose the factors with the highest eigenvalue, thus the extraction method. The eigenvalue is basically the amount of information processed by each factor (not needed if > 1). The Scree plot consisted of 17 different factors, where 6 had an eigenvalue higher than 1. Due to the fact that there were too may factors to take into account, the Cattell Method (1966) was used. A choice of 2 factors was made at the point where an ‘elbow’ was created on the scree plot’s curve from the horizontal to vertical axes which may explain the total variance of the items. Table 1: Rotation of Sums Squared Loadings, which shows the total variance of items in each factor.

Cross Cultural Relationships

Total Eigenvalue

% Of Variance

3.30

19.39

3.19

18.75

(Factor 1) Cultural characteristics (Factor 2) % Of Cumulative Variance

38.14

A rotation of the sums was performed after running the PCA the first time. As shown above in Table 1, the two factors extracted among the 17 factors with the highest variance were the first statement, which was the main goal of the survey and also another issue regarding the liking and disliking diverse cultural characteristics. This was to ensure the maximum variance of the items on factor and at the same time minimizing variance in other factors. The Orthogonal Varimax Rotation was used in this case because of the fact that 2 factors were to be extracted it ensures a very clear interpretation of the variance in the 2 significant factors involved in the survey and in addition the correlation occurring between the items splitting themselves up in the two factors.

Table 2: Rotated Component Matrix, which presents the eigenvalue and allocation of each item within each factor.

Items Crosscultural relation ships I like meeting new people from diverse cultures. (Item 1) I enjoy listening to music in different languages. (Item2) I don’t like going on holiday to countries other than my homeland. (Item

Factors Preference of diverse cultural characteristics

0.52 0.73 0.64

3) I have an interest in learning other languages. (Item 4) I prefer choosing close mates that are from the same ethnic group. (Item

-

0.56 -

5) It’s irritating to be around people that speak a different language to mine.

0.59

(Item 7) 0.79

I don’t like watching movies in other languages, even if there are subtitles.

(Item 8) I enjoy attending cross-cultural social events. (Item 10) I don’t feel comfortable when being approached by someone of a different culture. (Item 11) I like socializing with people who have similar skin colour. (Item 13) I automatically turn down a member of the opposite sex, if they are not the same skin colour as me. (Item 14) I don’t take relationships seriously if the couples are of different ethnic groups. (Item15) I would help charities that help developing countries. (Item 16) I would be in a relationship with someone of a different culture. (Item 17) I wouldn’t make a close bond with someone from a different culture that my friends won’t like. (Item 18) I would rather not share my secrets with a member of the opposite sex that isn’t from the same culture as me. (Item 19) I wouldn’t marry someone from a different culture, even if I love them. (Item 20)

0.47 0.50 0.51

-

0.64 0.52 0.70 -

-

0.58 0.63

Table 2 presents all of the items’ correlation (except items 6, 9 and 12 which were excluded before) with each factor. As it can be seen, Cross-cultural relationships had the majority of the items, thus permitting the usage of the items 1,3,10,14,15,16,17,19 and 20 in the new refined survey, which would be used to retest 10 participants to test external reliability. Moreover, to test the internal consistency of items that they are indeed measuring the same attitude, Cronbach’s alpha was used. The overall alpha of the refined survey was considerably good, a=0.76. It was not necessary to remove any items from the item pool in order to increase the alpha, thus the internal consistency. Out of the 50 participants, 10 were selected to go forward in the test-retest part of the survey in order to test external reliability. Basically, 5 males and 5 females were chosen to see whether their views might have changed

over a period of 2 weeks. To test external reliability of the survey, a correlation test was done between the results of the 10 people chosen for the refined survey and their initial results from the first survey 2 weeks earlier, which was at 0.7. In an overall outlook, the 50 participants had a mean of 4.35 and Standard Deviation of 0.52. The direction of scoring each statement varies due to the nature of it. A higher score of participant may suggest that one has a positive attitude towards the topic of cross-cultural relationships, and a low score shows that there is a negative view upon the matter.

Materials: The final scale which was used in the retest section was cropped and refined through the PCA is presented below. The list includes the items that intercorrelate towards the measure of the attitude towards cross-cultural relationships (Factor 1).

Strongly Agree I like meeting new people from diverse cultures. (Item1) I don’t prefer going on holiday to other countries, other than my homeland. (Item 3)

Agree







Neutral

Disagree













Strongly Disagree 

I enjoy attending cross-cultural social events. (Item 10) I automatically turn an opposite sex down if they are not the same skin colour as me. (Item 14)

 







I don’t take relationships seriously if the couples are of different ethnic groups. (Item15)









I would help charities that help developing countries. (Item 16) I would go in a relationship with someone of a different culture. (Item17)

















I would rather not share my secrets with an opposite sex that isn’t from the same culture as me. (Item 19)









I wouldn’t marry someone I love from a different culture (Item 20)













Procedure The place where the participants were allocated was in the Learning Resources Centre (LRC) of the University of Hertfordshire. Due to the fact that many people are studying in the LRC it may be hard to approach them and disrupt them from their business. Some participants were allocated outside the LRC, in the smoking area as it is much more casual and easier to approach someone. A simple opening line was used, i.e. “Hello, my name is Chris and I am a Psychology student. If you could be so kind to help and fill out a survey on views upon crosscultural relationships.” Allocating participants while they didn’t look like they were into their studying so much was easy as a minimal number of people refused to participate. The majority of participants responded positively. After giving them a brief introduction and elaborating a bit, they would be given the consent form and then the survey, which they filled out for as long as they needed, although they were told it would only take 15 minutes. While the participants were filling out the survey, the coordinator would sit in a comfortable distance and not look intensely at the participant to avoid stress and pressure. As soon as the participant completes the survey, he or she is given a debrief form with an email on it, and he or she was told that further communication may occur for further testing.

Results: As the survey was refined and the allocation of the data was completed it is important to say that all the scores of the 50 participants were coded and the scores of the 10 selected participants were chosen to fulfil the test-retest procedure. As a participant having a mean score of 4.53, shows a positive attitude towards the attitude object of the survey. On the other hand having a score of 3.00 means there is more negative view on the main attitude object.

Table 3: Descriptive Statistics on Attitude Scores of the Compared Groups

Gender

Number of

Mean Attitude

Standard

Participants

Score

Deviation

Male

25

4.31

0.53

Female

25

4.39

0.53

The main prediction of the survey was that there would be a difference between males and females in the survey measuring attitudes towards cross-cultural relationships. Having said this, with the data shown above it Table 3, an independent samples t-test was used to compare the two groups, males and females and their mean scores. The mean score of Male participants in attitude survey was 4.31(S.D. 0.53) and the means score of female participants was 4.39 (S.D. 0.53). An independent samples t-test showed that the difference between the two scores was not statistically significant: t (48)= -0.504, p= 0.617, twotailed. As a result, the statistical analysis rejects the prediction of a difference between attitudes of cross-cultural relationships. This means that the attitude towards cross-cultural relationships does not vary between males and females within the University of Hertfordshire, and indeed is it possessed by a positive view towards it than a negative one.

Discussion: The survey’s purpose was to measure the attitude towards cross-cultural relationships within the University of Hertfordshire. After developing the initial questionnaire that was completed by 50 participants, 25 male and 25 female, the survey was refined and tuned by using the PCA. The prediction of the survey of finding a difference in the attitude towards cross-cultural relationships between males and females was rejected. This means that the attitude towards crosscultural relationships was very similar between males and females. The participants chosen for the survey are students of the University of Hertfordshire and thus taking this into account it enhances the University notion of being multicultural and multinational. Moreover, the majority of the participants had a

more positive view towards cross-cultural relationships and indeed there is no significant difference between males and females in terms of the attitude. No significant difference was found between the two groups, as it may be that being a part of a University environment, there is a high possibility of change in lifestyle as this environment enhances social awareness and open-mindness. However, the main goal of the survey was the actual setting up of the survey, and refining the scale to the optimum level to cover some main issues, which would determine the efficiency and accuracy of the survey. By starting off, an item pool was selected after a brief discussion of different ideas that might be interesting for students. The development of statements was done carefully to cover the main criteria of a survey. There were two main issues to be covered, validity and reliability, in order to confirm the accuracy of the survey. Through the use of the PCA, the extraction method was used to remove all unnecessary factors that were being measured in the survey and two main factors were kept, which were rotated to get the most refined variability of the items in each of the two factors. Through the scree plot and the table of Rotated Sums of Loadings Squared and the Rotated Component Matrix (Table 1; Table 2), this refined the scale in terms of concurrent validity and ‘face and content validity’ (Oppenheim, 1997). Moreover, to cover the issue of reliability two tests were done (DeVellis, 2003). Firstly, Cronbach’s alpha was calculated to ensure internal consistency, as it was at an efficient level, which established the inter correlations of the items in terms of the factor being measured. Also, to cover external reliability, a test-retest criterion was set to the survey were 10 people were retested to see the correlation between their responses. The survey had a low external reliability (0.7) as the correlation between the 10 participants from week 2 and week 4 was low. This flaw in the survey may have been caused by an attitude change of the participant, such as an end in a relationship or even a deeper personal issue that may not be discussed. In addition, the refined test had much more negative statements than positive statements, which may have pushed the participant to complete the survey much differently than how it was initially completed. Moreover, the survey produced being a Likert scale (Likert, 1931), may have led participants thinking that it was a racism survey, and may have disoriented the point of the survey. This suggestion may come in relation to the theory of

planned behaviour (Fishbein and Ajzen, 1974), as due to the subjective norms of each individual and action or intent may occur, thus the fact that the survey was referring to different ‘skin colour’ and/or ‘marriage’ may have led to this misinterpretation. Nevertheless, the survey had an overall efficient outcome and may be further used in many different ways. For example, it may be used in social and dating websites, where individuals using these sites may be in need of a certain guideline or suggestion or even database matching to other individuals with the same views in cross-cultural relationships.

References: Allport, G. W. (1935). Attitudes. In C. M. Murchison (Ed), Handbook of social psychology (pp. 789-844). Worcester, MA: Clark University Press. Cattel, R. B. (1966). The scree test for the number of factors. Multivariate Behavioral Research, 1, 245-276 Cronbach, L. J. (1951). Coefficient alpha and the internal structure of tests. Psychometrika, 16, 297–334. DeVellis, R. F. (1991). Scale Developing: Theories and Applications, Applied Research Methods Vol. 26, London: Sage. Eagly, A. H., & Chaiken, S. (1993). The psychology of attitudes. Fort Worth, TX: Harcourt Brace. Eiser, J. R. (1986). Social Psychology: Attitudes, cognition and social behaviour. Cambridge, UK: Cambridge University Press. Fishbein, M., & Ajzen, I. (1974). Attitudes toward objects as predictions of single and multiple behaviour criteria. Psychological Review, 81, 59-74. Himmelfarb, S., & Eagly, A. H. (Eds.) (1974). Readings in attitude change. New York: Wiley. Katz, D. (1960). The functional approach to the study of attitudes. Public Opinion Quarterly, 24, 163-204. Klopfer, B., (1946). The Rorschach Technique: A Manual for a Projective Method of Personality Diagnosis, World Book Co, Yonkers-on-Hudson.

Likert, R. (1932). A technique of the measurement of attitudes. Archives of Psychology, 22, no. 140, 44-53. Oppenheim, A. N. (1992) Questionnaire Design, Interviewing and Attitude Measurement. New York: Basic Books Inc. Osgood, C. E., Suci, G. J. & Tannenbaum, P. H. (1957). The measurement of meaning. Urbana, IL: University of Illinois Press. Rosenberg, M. J., & Hovland, C. I. (1960). Cognitive, affective, and behavioural components of attitude. In M. J. Rosenberg, C. I. Hovland, W. J. McGuire, R. P. Abelson & J. W. Brehm (Eds.), Attitude organisation and change: An analysis of consistency among attitude components. New Haven, CT: Yale University Press. Thurstone, L. L. (1931). The measurement of social attitudes. Journal of Abnormal and Social Psychology, 26, 249-269. Vaughan, G. M., & Hogg, M. A. (2011). Social psychology (6th ed). Frenchs Forest, NSW: Pearson Australia.

Appendices:

Appendix A: Scree Plot produced to extract the number of Factors measured by the 20 statements produced.

Appendix B: The full copy of the original survey and consent from produced that was completed by 50 participants. Also in red, there is an indication of the item being positive re negative statement. Section 1: Attitudes Towards Cross-Cultural Relationships

Please read through the following statements and circle the dot for each statement that corresponds to your opinion towards it: strongly agree, agree, neutral, disagree and strongly disagree.

I like meeting new people from diverse cultures P I enjoy listening to music in different languages. P I don’t like going on holiday to countries other than my homeland. N I have an interest in learning other languages. P I prefer choosing close mates that are from the same ethnic group. N I enjoy eating food from different cultures. P It’s irritating to be around people that speak a different language to mine. N I don’t like watching movies in other languages, even if there are subtitles. N I do consider other peoples’ views and beliefs of culture. P I enjoy attending cross-cultural social events. P I don’t feel comfortable when being approached by someone of a different culture. N My friendship group is mainly made of different ethnic groups. P I like socializing with people who have similar skin colour. P I automatically turn down a member of the opposite sex, if they are not the same skin colour as me. N I don’t take relationships seriously if the couples are of different ethnic groups. N I would help charities that help developing countries. P I would be in a relationship with someone of a different culture. P I wouldn’t make a close bond with someone from a different culture that my friends won’t like. N. I would rather not share my secrets with a member of the opposite sex that isn’t from the same culture as me. N I wouldn’t marry someone from a different culture, even if I love them. N

Strongly Agree

Agree

Neutral

Disagree

 

 

 

 

 











 

 

 

 

 

 

 

 

 

 











  

  

  

  

  











 

 

 

 

 











 

 

 

 

 































Section 2: Please state your.... Gender:…………..

Age:…………..

Anonymity code: Please give your first two letters of your Mothers’ maiden name and your house number, for example if your mother’s maiden name is Rose and you live at number 12, your code will be RO12. ………………

Strongly Disagree

Attitudes towards cross-cultural relationships

To all taking part in the following survey, we are a group of undergraduate psychology year 2 students. We are interested people’s attitude towards ‘crosscultural relationships’ (including friendships). This survey will take only 5 minutes of your time, which will be valuable for our findings. Please read each of the statements carefully and tick the boxes to the extent to which you agree or disagree. For all participants taking part in this mini survey, your identity will stay completely confidential and will only be used for the purpose of this survey you have agreed to take part in. You may withdraw yourself from this survey at any time without giving a reason for your decision. Contact Christos Kassapis for any enquires for withdrawal or any ethical issues that may have occurred during/after the survey on [email protected]. Thank you in anticipation of your help.

Participant Consent form; I _________________________________________________________ give full consent to take part in the following survey and understand that my identity will stay completely confidential and I will be able to withdraw from this survey at any time during or after the survey is completed without giving any reason. In order for us to contact you for future information you may choose to leave your email address or your contact number here; _______________________________________________________________________________________

Researcher copy I _________________________________________________________ give full consent to take part in the following survey and understand that my identity will stay completely confidential and I will be able to withdraw from this survey at any time during or after the survey is completed without giving any reason. In order for us to contact you for future information you may choose to leave your email address or your contact number here; _____________________________________________________________________________________