Collective intelligence and social sensitivity amongst students at the University of the Witwatersrand
Date
2019
Authors
Van Staden, Jason Gary
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
The study of groups and their ability to solve problems is referred to as collective intelligence. Woolley, Chabris, Pentland, Hashmi and Malone (2010) demonstrated that measuring a group’s intelligence is similar to the way an individual’s intelligence is measured. This is done by establishing the emergence of a factor referred to as ‘c’, like the factor ‘g’ that emerges at an individual level. This factor predicts the performance of a group in future tasks. It appears from literature that social sensitivity is the most predictive variable of the collective intelligence factor. However, little is known about what determines a person’s social sensitivity. This research investigates social sensitivity to guide and inform academia and practitioners on how to improve collective intelligence.
The most popular measure used in current research for measuring social sensitivity in adults is the ‘Reading the Mind in the Eyes’ test (RMET). This is a test that requires respondents to choose the correct mental state of a person by viewing static stimuli of their eye region. Previous research, however, shows lower scores for cultures that are not Caucasian. It has been suggested that this is due to the stimuli being exclusively Caucasian in the current RMET. This research creates and tests two new RMET instruments, one containing African stimuli, the other, diverse Asian stimuli. Along with culture, gender has shown variations in performance on the RMET. Women tend to score higher than men on the RMET. This research extends literature by assessing gender RMET differences across cultures and different cultural instruments. This serves as Study 1 in the present research. A by-product of Study 1 is valid mental state items that is used to create a shortened multicultural RMET instrument. The new multicultural instrument avoids cultural biases and allows for the efficient measurement of social sensitivity.
Study 2 investigates determinants of social sensitivity by using the multicultural instrument developed in Study 1. The determinants are trait emotional intelligence, personality and individual contextual factors such as culture, gender and family background. Along with direct relationships, mediation and moderation relationships are also tested.
A post-hoc investigation, Study 3, assesses social sensitivity’s importance to collective intelligence in an African context. This is also the first attempt in an African context to validate the emergence of the collective intelligence factor when applied to an African sample. Together, these three studies further academic knowledge of social sensitivity and collective intelligence.
Using a quantitative methodology, the research consists of two designs. The first uses a questionnaire survey administered in person. The first research design was used to create and test new social sensitivity instruments (Study 1) and investigate the predictors of social sensitivity (Study 2). The second research design implements a quasi-experimental method which is used to test for the emergence of the collective intelligence factor in groups. Study 1 tested a sample of 1 306 responses of University of the Witwatersrand students. Study 2 tested 427 responses, while Study 3 tested a sample of 36 students – which were dived into 13 groups. Study 1 follows precedent in literature to create the new RMET instruments and then applies analysis of variance (ANOVA) techniques to assess variance across cultures and genders. Studies 2 and 3 use factor analysis, bivariate and multivariate statistical techniques to assess variables and their relationships with each other. Moderation and mediation techniques are also applied.
Study 1 results produce a contextually valid 21-item African RMET, and a contextually valid 26-item diverse Asian RMET. Assessment of these new instruments plus 30 context-validated items from the original Caucasian RMET show significant cultural differences. Those self-reporting with an African culture score significantly lower on social sensitivity as measured by the RMET, regardless of the culture of the stimuli presented to them. Therefore, there appears to be no advantage in performance on the RMET tests when assessing the stimuli of an individual’s own culture, an assumption which is suggested in literature. A gender advantage, whereby women score higher on social sensitivity, is also present across all cultural instruments. Study 2 results investigate the antecedents of social sensitivity.
The individual contextual factors gender and home language significantly predict social sensitivity in Study 2. Personality and trait emotional intelligence prove to be non-significant predictors of social sensitivity. Interestingly, the relationship between trait emotional intelligence and social sensitivity is moderated by home language. Those who speak an African home language display a significant negative relationship between trait emotional intelligence and social sensitivity. In contrast, those who speak English as their home language show a significant positive relationship. This finding provides insight into the cultural differences observed in the RMET.
Findings suggest that those of an African culture self-report that they are more emotionally intelligent if they perform lower on the RMET. This suggests high performance in mind-reading by identifying facial cues is, potentially, not an important indicator of high emotional intelligence in African cultures. In support of this, literature suggests that less emphasis is placed on facial cues in African cultures, thus cultural norms need to be considered. For instance, in certain African cultures direct eye contact is seen as a sign of disrespect in certain situations. These norms appear to influence African cultures’ performance on the RMET. This finding challenges assumptions in academic literature, particularly views on the differences between cultures’ RMET scores. This study provides culturally important information on different cultural views on assessing mental states.
Study 3 results provide limited evidence that the collective intelligence factor emerges in an African context. However, results show that social sensitivity is correlated with the identified collective intelligence factor in the study. This research provides three new tested RMET instruments to academia (African, diverse Asian and a shortened multicultural instrument). This investigation also shows that African cultures and men perform lower on social sensitivity instruments. In addition, the research also shows African cultures who are performing higher on trait emotional intelligence have lower RMET scores, an opposite effect to English cultures, suggesting differing perspectives on the importance of this particular ability to each culture’s emotional intelligence. Such findings provide interesting implications for both academia and practitioners.
Education and training of both African and other cultures will potentially allow an understanding of each culture’s social interaction norms, and what is seen as the emotionally correct response in different situations. This adjustment may help social interactions between cultures by governing the individual’s use of different emotional skills in a way that promotes cultural inclusiveness and cooperation. This research, therefore, provides resources and knowledge to help academics and practitioners hypothetically improve cross-cultural social interactions. This understanding will potentially improve collective intelligence in teams and groups.
Description
A thesis submitted to the Faculty of Commerce, Law and Management, University of the
Witwatersrand, Johannesburg, in fulfilment of the requirements of the degree of Doctor of Philosophy
Keywords
Citation
van Staden, Jason Gary, 2019, Collective intelligence and social sensitivity amongst students at the University of the Witwatersrand, University of the Witwatersrand, https://hdl.handle.net/10539/29624