Publication: Gender & Behaviour
Date published:
Language: English
PMID: 93409
ISSN: 15969231
Journal code: GNBV

Introduction and Literature Review

According to Social Representation Theoiy (SRT) individuals faced with a frightening and complex new occurrence such as HIV/AIDS will seek to integrate this new concept into their eveiyday cognitions, lifestyles and behaviours. In doing so, the main groups protect their identities by way of group-specific representations (Joffe, 1996, 1999; Marková & Far, 1995; Moscovici & Perez, 1997). The consequence of fitting their understanding of the disease within pre-existing phenomenon allow individuals within the main group to feel less anxious about the disease as well as psychologically distance themselves from pariah groups who are blamed for the incident. The SRT has been applied to the spread of HIV/AIDS (Joffe & Bettega, 2003; Goodwin, Kozlova, Nizharadze & Polyakova, 2004a&b). These studies explored how young people in Zambia and Eastern Europe respectively conceptualize the origin and spread of HIV/AIDS, their risk perception, values, and sexual behaviour. The SRT has also been applied to safe and unsafe sexual practices such as multiple partners, not using a condom etc (e.g. MacPhail & Campbell, 2001).

A major claim arising from this approach is that blaming other marginal groups within society for the spread of the disease allows individuals within the mainstream society to feel relatively safe from the disease, and so they do not take sufficient steps to protect themselves. For example, Joffe and Bettega found that their sample tended to blame out-groups for the spread of the disease, and reported low perceptions of their own risk. They reported that blaming out-groups and low perceptions of own risk co-exist in the same sample and so argued that one leads to the other. Although plausible, this claim has not been directly tested among young people in UK. To test these findings, the theoiy was applied to young male students in the UK.

Aim of Study

The aim was to provide a more direct test of the central claim that blaming out-groups leads to lower risk perception and subsequently to higher risk-taking behaviour.

This was done by exploring the question directly with the participants themselves in the context of their own understanding of the factors that contribute to their assessment of their own risk status, and by statistically testing whether those who more strongly endorse beliefs about out-groups being to blame are more likely to show lower perceptions of risk, and whether lower perception of risk is associated with higher engagement in unsafe sexual practices.

Social Representation Hypothesis

The hypothesis is that distancing strategies result in reduced appraisal of threat, which in turn results in less safe sexual behaviour. In statistical terms, the hypothesis is that there will be a negative association between the use of distancing strategies and safe sexual behaviour that is mediated by a reduced threat appraisal. A mediatoiy relationship, in turn, implies a significant negative correlation between distancing strategies and safe sexual behaviour; a significant negative correlation between distancing and threat appraisal; and, typically though not necessarily, a significant positive correlation between threat appraisal and safe sexual behaviour (Kenny, Kashy & Bolger, 1998).

Ethical Concerns

The approval of this study was granted by the Ethics Committees for the School of Psychology, University of Birmingham. The signed consents of both pupils and their parents or accredited guardians were required before any pupil took part in the study. The main ethical concerns were to protect the identity of the participants; to ensure proper informed consent; to avoid pressurising participants into taking part; and to avoid causing any distress to participants. An opt-in scheme was employed to recruit participants from the secondaiy schools. Both students and their parents or guardians signed consents before any pupil took part in the study. The pupils were free to withdraw from the study whenever they wished. The welfare of the pupil took priority over any other considerations.



One structured questionnaire with four subscales was designed for this study. The questionnaire was designed to measure the following:

* Two distancing variables associated with the social representation approach - specifically, blaming out-groups and endorsement of the idea that HIV spread because of the excessive sexual behaviour of people from Africa.

* The participants' appraisals of how much threat HIV posed to them.

* Personal moral views about the acceptability of various safe/unsafe sexual behaviours

* How safe had been the sexual behaviour of those in the sample who were sexually experienced.

The first two and the fourth of these were measured in order to address the social representation claim that distancing serves to make people feel safe, and they consequently are less cautious in their sexual practice. The third was measured to explore the contribution of personal moral values to sexual behaviour (an issue that tends to have been neglected by both traditional cognitive models, such as the theory of planned behaviour; and by the social representation approach).

Each item on the questionnaire consisted of a statement, to which the participant responded 'true' or 'false' (or on some items, ^es' or 'no'J. Answers were scored with a 0 or a 1, and summed to give a score for each subscale. Higher scores were associated with higher exhibition of the belief, attitude or behaviour being measured. Higher scores on the variable (personal views) represented personal views in favour of safe sexual practices. Higher scores on the safe sex variable represented safer sexual practices.

The questionnaires were devised for the study because social representations about HIV/AIDS and sexual behaviours, and beliefs and attitudes about such matters (and the language used to describe them) are specific to the socio-cultural context (Moscovici, 1984; Joffe, 1999).

Therefore, rather than attempting to use a standardised questionnaire from previous research, it was considered necessaiy to generate a questionnaire appropriate to the specific socio-cultural context. The questions used for the study were based on the findings from a qualitative study with the sample population.

Recruitment of Participants

Sixty-five male students between the age of 15 and 17 were randomly selected from a single-sex private school in a large city in the U.K. to participate in the study. The figure of 65 has been suggested as an adequate basis for small to medium models (Cohen, 1988; Howell, 2002). Furthermore, this sample size was considered adequate because power analysis indicates that with three predictors, 54 participants will provide adequate power to detect a moderate effect size in relation to a beta coefficient (power = 0.8; alpha = 0.05; moderate effect = squared correlation of 0.13) (Green, 1991).

Justification for use of Young People

Young people were used for the study because reports from around the world suggest that negative outcomes of HIV/AIDS threaten the lives of people in their youth more than any other age group (Lamptey, Johnson, & Khan, 2006; UNAIDS, 2008; UNICEF, 2002; USAID, 2005).

Procedure for Data Collection

The surveys took place on the school's premises. Questionnaires were distributed as pupils sat at their desk in their classroom. Each questionnaire had a unique code on it. The participants were asked not to write their names on any part of the questionnaires, but rather to take note of their unique codes and to contact the researcher or a named teacher of the school within 24 hours with their code if they wish their questionnaires to be destroyed. They were informed that their participation was voluntaiy; that they could choose not to answer any question that made them uncomfortable; and that they could withdraw from the study at anytime. They were asked to complete the questionnaires honestly and without any discussion with fellow pupils. They were also informed that their answers were confidential and that there was no possibility of their answers being traced back to them.

Participants first completed a short demographic questionnaire ascertaining age, sex, class/form, religion and whether or not they know somebody who is HIV positive. Then they completed the main questionnaire. However, out of the 65 participants who were selected from those interested to participate in the study, only 39 completed the questionnaires. The remainder returned the uncompleted questionnaires because they saw the questions as "intrusive".

Data Protection

Completed questionnaires were kept in a locked filing cabinet when not in use. No record was kept of the names of the participants other than their signature on the consent forms. The only record of their real names was on the signed consent form. The consent form did not contain the code, and so there was no way of connecting the consent forms to completed questionnaires.

Data Analysis

Data were analysed using SPSS 16.0 for Windows (SPSS Inc., 2008). Kendall's tau (τ), a non-parametric correlation were used to test for the relationships between distancing variables (outgroups) and threat appraisals, personal views about sexual behaviours and actual safe sex. Non-parametric correlations were used because the variables were not normally distributed. Also Kendall's tau was used rather than Spearman's product moment coefficient correlation because of the small data set with a large number of tied ranks (Field, 2005).


Participant Characteristics

Sixty-five male students were randomly selected to participate in the study. However, only 39 completed surveys were returned. The remaining 21 were returned to be uncompleted because the students reported that that the questions were intrusive. Those who completed the surveys were between 15-17 years of age with a mean age of 16 years (standard deviation of 0.98). Christians constitute the majority (27%) followed by Sikh (24%), with Hindus being the minority (5%). The fathers of 63% were professionals, 8% were self employed and the remainder did other jobs. Fifty percent of mothers were professionals, 8% self-employed and the remainder did other jobs. Asian British were in the majority, constituting (45%); followed by White British, (39.5%) and then Indian (10.5%). The least represented group were Black British who made up 3%. The remaining 3% were of mixed race.

Only 2 of the participants knew someone with HIV. Twenty seven percent (N = 9) of them reported ever having sex (at least once). One person did not provide an answer to that question.

Descriptive Data

Table 2 (below) shows the range, mean and standard deviation of the subscales used for the study. Higher scores on threat appraisal represented a perception of greater risk; higher scores on distancing variables represented higher blaming of out-groups and a stronger endorsement of the idea that African hyper-sexuality is to blame for the spread of HIV in Africa. Higher scores on the variable actual safe sex represented safer sexual practices. Higher scores on the variable personal views represented personal views in favour of safe sexual practices. The distribution of some of the scores (personal views of sexual behaviours and African sexuality) were not normally distributed and the internal reliability (Cronbach's alpha) for some of the subscales was below the 0.7 usually considered as the benchmark for adequate internal consistency (Kline, 2000). Consequently non-parametric correlations were used to analyse the data. Kendall's tau was used rather than Spearman's product moment coefficient correlation because of the small data set with a large numbers of tied ranks (Field, 2005b; Howell, 2006).

Social Representation Hypothesis

The data offered partial support to the social representation hypothesis. There was a significant correlation between blame out-groups and African sexuality, suggesting that these variables may be sensitive to a broader construct of distancing. There was also a significant correlation between personal moral views and actual safe sex, suggesting that a person's views about the moral and social acceptability of behaviours do influence actual behaviour.

However, there was no effect to be mediated (i.e. no significant negative correlation between the distancing variables and actual safe sex); there was no significant negative correlation between the supposed causal factor (distancing) and the mediator (threat appraisal); and no significant positive correlation between the mediator (threat appraisal) and the outcome (actual safe sex). Since only 10 participants had actually had sex, the correlations involving this variable should be considered as having low reliability.

Discussion and Conclusions

The main aim of this study was to subject the Social Representation Hypothesis to a more rigorous empirical test by means of investigating individual differences among young males in UK. The claim of social representations approach to HIV/AIDS is that distancing oneself from the disease generally allows the dominant group to feel relatively safe from the disease, and so they do not take sufficient steps to protect themselves. A number of studies have offered evidence in support of this hypothesis (e.g. Marková & Far, 1995; Joffe, 1996, 1999; Joffe & Bettega, 2003; Goodwin et al., 2004). Most of these studies found that their sample distanced themselves from HIV/AIDS, had lower risk perceptions and reported higher risk-taking behaviours and so concluded that distancing oneself from the disease generally allows them to feel relatively safe from the disease and so took sexual risks.

However, this study did not provide particularly encouraging support for the social representation hypothesis, but the results suggested the value of pursuing the investigation of personal moral attitudes as a potential factor in sexual behaviour. The idea about personal moral attitudes about the sexual practices may be associated with safer sex.


Field, A. P. (2005). Kendall's Coefficient of Concordance. In B. S. Everitt & D. C. Howell (Eds.), Encyclopedia of Statistics in Behavioral Science, me 2, (pp. 1010-1011). Chichester: John Wiley.

Goodwin R., Kozlova A., Nizharadze, G., & Polyakova, G. (2004a). HIV/AIDS among adolescents in Eastern Europe: Knowledge of HIV/AIDS, Social representations of Risk and Sexual Activity among School Children and Homeless adolescents in Russia, Georgia and the Ukraine. Journal of Health Psychology, 9(3), 381- 396.

Goodwin, R., Kozlova, A., Nizharadze, G., & Polyakova G. (2004b). High-Risk Behaviours and Beliefs and Knowledge about HIV Transmission among School and Shelter Children in Eastern Europe, Sexually Transmitted Diseases, 31(11), 670-675.

Green, S. B. (1991). How many subjects does it take to do a regression analysis? Multivariate Behavioural Research, 26, 499-510.

Howard, L. (2006). SouthAfrican Perceptions of Risk and the Social Representations of HIV/AIDS. Psychology Masters Research Thesis, University of the Witwatersrand, School of Human & Community Development, Johannesburg, South Africa.

Joffe, H. (1995). Social representations of AIDS: Towards encompassing Issue of Power. Papers on Social Representations texts sur les répresentaions sociales (1021 - 5573) 4(1) 1 - 40.

Joffe, H. (1999). Risk and 'the Other'. Cambridge University Press

Joffe, H., & Bettega, N. (2003). Social representation of AIDS among Zambian adolescents. Journal of Health Psychology, 8(5), 616-631.

Kenny, D. A., Kashy, D. ?., & Böiger, N. (1998). Data analysis in social psychology. In D. Gilbert, S. T. Fiske, & G. Lindzey (Eds.). The handbook of social psychology (4th ed., Vol. 1, pp. 223-265). New York: McGraw-Hill.

Kline, P. (2000). Handbook of Psychological Testing (2nd edition). London: Routledge.

Lamptey, P. R., Johnson, J. L., & Khan, M. (2006). The Global Challenge of HIV and AIDS. The Population Reference Bureau, Population Bulletin, 61(1).

MacPhail, C., & Campbell, C. (2001). 1 think condoms are good but, aai, I hate those things': condom use among adolescents and young people in a South African township, Social Science and Medicine, 52, 1613 - 1627.

Markova, I., & Farr, R. (eds). (1995). Representations of Health, Rlness and Handicap. Switzerland: Harwood Academic Publishers.

Markova, I., & Wilkie, P. (1987). Representations, concepts and social change: The phenomenon of AIDS. Journalfor the Theory of Social Behaviour, 17, 389-409.

Moscovici, S. (1963). Attitudes and opinions. Annual Review of Psychology, 14, 231-260.

Moscovici, S. (1973). Foreword. In R. M. Farr & S. Moscovici (Eds.). Social representations (pp. p. xiiij. Cambridge: Cambridge University Press.

Moscovici, S. (1984). The phenomenon of social representations. In R. M. Farr & S. Moscovici (Eds.). Social representations (pp. 3-69). Cambridge: Cambridge University Press.

UNAIDS/WHO. AIDS Epidemic update, 2008. Geneva: UNAIDS; 2008

United Nations Children's Fund, Joint United Nations Programme on HIV/AIDS and World Health Organization. (2002). Young people and HIV/AIDS - opportunity in crisis. New York, USA.

United States Agency for International Development (USAID). (2005). USAID Project Profiles: Children Affected by HIV/AIDS (4th Edition). Januaiy 2005. Washington, DC: Population, Health and Nutrition Information Project. USAID Project Profiles: Children Affected by HIV/AIDS, Fourth Edition - January 2005 [PDF, 2MB]

Author affiliation:

Dinah Baah-Odoom

Schoo of Public Health, College of Health Sciences

University of Ghana

The use of this website is subject to the following Terms of Use