Comparative survey data on sociodemographic predictors of diversity tolerance among selected university students in Ghana and South Africa

This article presents an extensive comparison of survey data on tolerance attitudes of 1758 participants from two public universities in sub-Saharan Africa, the University of Ghana and North-West University. Multi-stage and other sampling procedures were employed to collect the data between 2016 and 2017. Data were analysed using frequencies, percentages and cross-tabulations for each institution separately. Overall, participants expressed a high level of tolerance to others of different racial and ethnic backgrounds, albeit higher for those in the University of Ghana than North-West University. The findings further revealed that participants’ gender, academic level, family socioeconomic status, and parental educational level were significantly associated with tolerant attitudes, depending on the educational institution.


Specifications
Social Sciences, Humanities. Specific subject area Demography, Education, Population Studies, Sociology, Youth Studies Type of data Tables  How data were acquired Information was gathered by administering semi-structured questionnaires to sampled students. A copy of the questionnaire is attached as a supplementary file. Data format Raw Analysed Parameters for data collection Information collected included individual demographic and socioeconomic characteristics, family background, reported attitudes toward foreigners and cultural diversity (see supplementary files). Description of data collection Data were obtained from 912 to 846 students from the University of Ghana and North-West University respectively, using a multi-stage sampling (both stratified and systematic sampling) technique. Data source location The University of Ghana, Greater Accra Region, Ghana. North-West University, Mafikeng, Vaal Triangle, Potchefstroom, South Africa. Data accessibility Data is included in this article

Value of the Data
• Africa is often characterized as a region of rampant intolerance, division and conflict in an increasingly globalized world-whether that tensions be based on ethnicity, religion, nationality, or sexual orientation [1] . This negative image persists despite a paucity of data gaging attitudes towards diversity in the African context. The dataset aims to contribute to the knowledge base and broader debates on tolerance in this regard. • Respect for and acceptance of diversity is crucial in maintaining social cohesion in a population of different cultures, races/ethnicities, and religions [ 2 , 3 ]. The dataset provides vital information on individual background characteristics, parental characteristics and self-reported attitudes toward ethno-racial differences and human diversity (see supplementary file). This can be useful for anyone who has an empirical interest in multiculturalism or interculturality, intergroup relations and group conflict, particularly in the youth population or higher education sector. • Higher education is posited as promoting a more enlightened and cosmopolitan worldview, attenuating prejudice, and fostering intergroup tolerance [4][5][6] . Thus, people participating in the higher education system are exposed to such modernizing values as good citizenship and inclusivity needed for positive social relations [3] . The data, therefore, can be used to test the liberalizing attitudinal effects of education-higher education is associated with more favorable attitudes toward different cultures or groups. • The datasets are primarily focused on two institutions of diverse social contexts in sub-Saharan Africa: The University of Ghana is the premier and largest public university in Ghana, a multi-ethnic and relatively harmonious society; whereas North-West University is a unified multi-campus public university in South Africa, a polarized multiracial society. Given both sample sites are culturally diverse environments, the dataset can be used in descriptive or comparative research into how diverse social backgrounds (educational, racial/ethnic, class, religion and gender) shape individual attitude towards and or acceptance of diversity. Table 1 summarises sample demographic and socioeconomic characteristics for the two groups. Overall, females constituted more than half of the total sample in both universities -University of Ghana (51.8%) and North-West University (54.8%). The average age of participant from the University of Ghana was 21.1 years old, whereas that of north-West University was study, whereas more than a third of participants in North-West University were in their first and final years of study. More than two-thirds (69.4%) of participants in North-West University reported belonging to middle-wealth households, whereas more than two-fifths of participants from the University of Ghana reported belonging to middlewealth (48.4%) or richer (45%) households. Breakdown by parental education shows that more than half (56.4%) of participants from the University of Ghana had fathers with tertiary level education compared to 44.7% of participants from North-West University. Finally, more than a third of participants from the University of Ghana had mothers with secondary-level education or higher, while more than two-fifths of participants from North-West University had mothers with secondary-level education (45%) or higher (43.4%). Fig. 1 shows the tolerance level of students from University of Ghana and North-West University towards members of different racial/ethnic groups; Tables 2 and 3 show the distribution of the responses to the eight items measuring tolerance of diversity. As Fig. 1 shows, on the whole, the level of tolerance or openness to diversity was rather high for both groups, although  Table 3 shows, the mean score for students' tolerance to diversity was significantly higher than the scale midpoint for both University of Ghana and North-West University at different levels. Table 4 shows the relationship between the eight items measuring tolerance of diversity. The students' attitudes were found to be significant positive correlates of toleranceranging from 0.429 to 0.892 for the University of Ghana and 0.283 to 0.881 for the North-West University-indicating that students at the University of Ghana were more tolerant of diversity than counterparts at the North-West University. Table 5 presents the relationship between the patterns of distribution of tolerance to diversity and selected socio-demographic factors for students in Ghana and South Africa, respectively. As Table 4 shows, among University of Ghana students, males (85.4%), students of Ga-Adangbe ethnicity (88.1%), those who belonged to 'other' religions (90%), those who were in their final year of study (87.2%), those whose fathers had secondary-level education (88.8%) or mothers had below secondary-level education (84.8%), and those from poorer families/households (89.8%) were the most likely to tolerate diversity. Conversely, among North-West University students, females (79.1%), Coloureds (80.0%), Muslims (83.3%), students who were in their first year of study (78.6%), those whose fathers had tertiary education (78.7%) or mothers had secondary-level education (78.5%), and those from richer families/households (82.2%) were the most likely to tolerate diversity. Lastly, the bivariate analyses show that the academic level and paternal education were significant for students at the University of Ghana, whilst gender, family socioeconomic status, and maternal education were significant for students at the North-West University.

Experimental Design, Materials and Methods
The data forms part of an exploratory investigation under the Religion and Positive Youth Development Project, an initiative of the University of Ghana and the North-West University (Mafikeng Campus). Data collection employed multi-stage cluster sampling procedures in selecting the final respondents from the two universities. First, respondents were divided by faculties using a probability sampling method, stratified sampling. This was achieved through proportional allocation to size based on the student population of each faculty. Departments were then selected from each faculty and each department, core courses at all levels (year of study) were selected for inclusion. Consequently, the student population of each faculty determined the proportion of respondents selected for inclusion in the final sample. Individual faculty samples were Table 4 Intercorrelations between participants' attitudes and diversity tolerance.

University of Ghana
North-West University  further disaggregated by year of study; trained field assistants surveyed the number of students in each faculty, department, year of study and gender as determined a priori through the stratified random sampling procedure. Using the triangulated sampling procedures, a total of 846 and 912 individuals was obtained from North-West University and the University of Ghana, respectively. Trained field assistants administered semi-structured questionnaires in English to participants in the respective institutions. The data collection occurred in two phases: data from North-West University was collected between September 2015 and April 2016, whilst that of University of Ghana took place between September 2016 and March 2017. Participants completed a series of questions related to biographical information (age, gender, religion, nationality, year of study, faculty, marital status), family socioeconomic conditions, parental characteristics, perceptions of immigration and foreign nationals, as well as attitudes towards diversity.
The outcome variable of interest in the present analyses was tolerance of diversity . Because tolerance is a complex [2] , multidimensional and context-dependent construct rather than a single universal construct, it is conceptualized and operationalized here in several ways. First, it is broadly understood to mean positive social perceptions and behaviors-acceptance of others, such as relating to individuals from different racial, ethnic, and/or cultural backgrounds [7] . Secondly, indicators developed by the Sociology Department at the University of Johannesburg, South Africa (through the Student Communities Project) were adapted for the data landscapes by the researchers, enabling them to measure students' attitudes and opinions about other peo-ple/groups of diverse backgrounds. The original research instrument from which the survey items were adapted was a social distance scale about willingness to accept different degrees of social proximity with members of various groups. The original instrument itself is essentially an adaptation of the Bogardus social distance scale, used to measure attitudes and prejudice, to suit an African landscape. The original scale measures varying degrees of closeness in people towards other members of diverse social, racial or ethnic groups.
Accordingly, the measure of tolerance of diversity was computed as a summated score of the following eight items: All eight items were measured on a five-point Likert scale ranging from 1 = 'to no extent' to 5 = 'to a very large extent'. Higher means indicated greater levels of tolerance and higher willingness to accept differences, and vice versa. The internal consistency of the instrument showed Cronbach's alpha of reliability coefficients of 0.927 for the University of Ghana sample and 0.885 for North-West University, consistent with a previous study that used the questionnaire in a sample of university students in Johannesburg [3] . However, for ease in crosstabulation in the bivariate analyses, i.e. Table 3 , the responses were dichotomized by splitting them at the mean into tolerance and intolerance of diversity. Thus, a score lower than "3 represented intoleran ce or low tolerance and a score of 3 or higher represented ( high) tolerance . The data was then descriptively and inferentially analysed into tables using frequencies, percentages and cross-tabulations, as well as Pearson Chi-Square tests to determine the relationship between tolerance and socio-demographic factors. All computations were done using SPSS version 25.
It is worth noting that owing to differences in the social stratification of the societies where the two institutions are situated, race and ethnicity are used interchangeably as proxy indicators for culture to allow for easy comparability of data. Also, because the questionnaires were anonymous and the sensitive nature of the questions, some respondents may have been reluctant to respond to some of the items, which led to missing values in some of the variables. There is also a strong likelihood of social desirability in the responses. Finally, because the data was collected from only two public universities, its generalizability to the larger high education sector in the respective countries is limited. Limitations notwithstanding, the trends are substantively and theoretically important-the instrument can be adopted or adapted for use in higher samples or other contexts.

Ethics Statement
The ethics approval was obtained from the Faculty of Health and Social Sciences of the North-West University (Mafikeng Campus). Participation was voluntary and anonymous; data was treated with optimum confidentiality.

Declaration of Competing Interest
None.