Donate

Values and social representations of HIV/AIDS in Central and Eastern Europe:

A multi-method investigation in five nations  

Parts (First section of report):  1 2 3 4

Robin Goodwin^*,

Brunel University,

UK.

Alexandra Kozlova

St. Petersburg University.

Russia

Anna Kwiatkowska

Bialystok University

Poland

Lan Anh Nguyen Luu

Eotvos Lorand University

Hungary

George Nizharadze

Academy of Sciences

Georgia 

Anu Realo,

University of Tartu

Estonia

^* Address for correspondence: Department of Human Sciences,

Brunel University, Uxbridge, UB8 3PH, England. Email: Robin.Goodwin@brunel.ac.uk.

We would like to thank Ahto Külvet and Andu Rämmer for their help during all stages of the project.

We are also grateful to Grazyna Wieczorkowska - Nejtardt. (Polish Academy of Sciences) and

 Shalom Schwartz (Hebrew University, Jerusalem) for their statistical advice.

Contents

Page

Section 1: Theoretical introduction

1.1. Introduction to the work 5

1.2. Incidence and treatment of HIV/AIDS in Central and Eastern Europe 5

1.3. Previous research on factors underlying the spread of HIV and knowledge of HIV 8

1.4. Assessing AIDS beliefs: the theory of social representations. 11

1.5. A culture-specific approach 13

1.6. Additional predictors of sexual activity and perceived vulnerability to HIV 16

1.7. Research Participants 17

1.8. Overview of the research 18

Section 2: Interviews and free associations

2.1. Method 19

2.1.1. Participants

2.1.2. Procedure

2. Results: 20

2.2.1. Interview analysis

2.2.2. Results: Free associations

2.3. Discussion of stage 1 of the research 23

Section 3: Matrices and Questionnaire data

3.1. Method 27

3.1.1. Participants

3.1.2. Measures

3.2. Results 30

3.2.1. Knowledge of HIV amongst our respondents

3.2.2. Sexual behaviour across the sample

3.3. Representations of HIV: Using matrix data 37

3.4. Results from the matrix analyses 38

3.5. Discussion of stage 2 39

Section 4: Media analysis

4.1. Introduction to the analysis 41

2. Hypotheses 41

3. Evaluation of the hypotheses 42

4.4. Individual country patterns 44

4.5. Discussion of stage 3 49

Section 5: Local analyses

5.1. Street and school children in Russia 51

5.2. Gender stereotypes and sexual behaviour in Poland. 53 

5.3. Gender stereotypes in Hungary 54 

5.4. National stereotypes and HIV in Georgia 55

5.5. Personality and sexual behaviour in Estonia 55

Section 6: Overall Conclusions

6.1. Summary 57

6.2. Assessing external validity 57

6.3. Implications of our findings 60

6.3.1. The impact of enduring social representations of HIV

6.3.2. The role of values in promoting safer sexual behaviour

6.3.3. Knowledge and sexual behaviour

6.3.4. The role of the media in health promotion

6.4. Further studies 63

6.5. Concluding remarks 64

References 66

Table 1: Country level scores on Schwartz’s value dimensions 73

Table 2: Stage 1 Participants 74

Table 3: Free Associations with HIV/AIDS 75

Table 4: Stage 2 Participants 76

Table 5: Culture-level data by country 77

Table 6: Correlation matrix for stage 2 variables 78

Table 7: Logistic regressions for stage 2 79

Table 8: Content analysis of Media 82

Table 9: Russian ‘local’ analysis 84

Appendix 1: Interview items from stage 1 90

Appendix 2: Interview items and primary response categories 91

Appendix 3: Commentary of cluster analysis results 99

Appendix 4: Detailed media analysis reports 102

Appendix 5: Shared and country-level matrices 122

Values and social representations of HIV/AIDS in Central and Eastern Europe:

 A multi-method investigation in five nations

1.1 Introduction

The HIV/AIDS epidemic is a relatively recent phenomenon in

Eastern Europe, only beginning in the early 1990s. However, WHO AIDS surveillance

figures indicate a rapid growth in both HIV and AIDS in Central and Eastern Europe

over the past five years (European Centre for Epidemiological Monitoring of AIDS in

Europe, 1999). Between 1995 and 1997 the former socialist economies of Eastern Europe

and Central Asia saw HIV infection rates rise six-fold (Joint United Nations Programme

on HIV/AIDS, June 1998a), with, for example, the 158 people testing positive in the

Russian Federation in 1994 in the Russian Federation rising to 4400 in 1997, and with

5000 new HIV infections identified during the first six months of 1999 (Kalichman et al,

2000). Although the great majority of new HIV infections have been in the drug-injecting

population (Rhodes, Ball, Stimson et al., 1999), this is a primarily young and sexually active

group (Kalichman et al, 2000; Rhodes, Stimson, Crofts, Ball, Dehne & Khodakevich, 1999)

and rates of growth in other sexually transmitted diseases such as syphilis and gonorrhoea

in this population suggests high rates of unprotected sex with non-monogamous partners,

whilst bacterial sexually transmitted diseases (STDs) enhance the transmissibility of HIV

(Axmann, 1998; Bingham & Waugh, 1999; Borisenko, Tichonova & Renton, 1999; Dehne

et al, 1999; Joint United Nations Programme on HIV/AIDS, December 1999; Kalichman et

al, 2000; Rhodes, Ball & Stimson, 1999). As a result, Eastern Europe now shows the world’s

steepest HIV curve (Joint United Nations Programme on HIV/AIDS, December 1999) and

HIV rates are expected to increase substantially in a number of the countries in this region

(Dehne et al, 1999; European Centre for Epidemiological Monitoring of AIDS in Europe, 1999).

1.2. Incidence and treatment of HIV/AIDS in Central and Eastern Europe

In the research reported here, we focus on samples taken from five Central and Eastern

European countries. These countries have had very different experiences of HIV infection

 and have employed different policies for HIV treatment and testing (Joint United Nations

Programme on HIV/AIDS, June 1998b; European Centre for Epidemiological Monitoring of

AIDS in Europe, 1999). The countries also vary both in terms of political complexion of the

leadership, influence and nature of religion practised and levels of economic investment and

growth. This latter is likely to have particularly important implications for the spread of HIV

infection (Borisenko et al, 1999).

Estonia and Georgia were both parts of the Former Soviet Union with low rates of HIV infection.

 Estimated cumulative number of AIDS cases in both countries are less than 100 (27 in Georgia,

22 in Estonia). In Estonia, HIV testing began in 1988 (Poder & Bingham, 1999) and is mandatory

for blood donors and prisoners. Forty-five people were found to be HIV positive between

December 1988 and January 1995 (Joint United Nations Programme on HIV/AIDS, June 1998b)

but as yet infection rates are not associated with drug user (Dehne et al, 1999). However, there

is a potential danger of infection given the relatively large increase in STD incidence, which has

risen from less than 10 cases per 100,000 in 1982-1990 to 70 per 100,000 in 1995. In Georgia,

voluntary testing replaced mandatory testing during 1992/1993, but testing remains mandatory

for blood donors. Nearly all new infections since 1996 have been amongst the drug-using

population (Dehne et al, 1999). Although AIDS cases are few, a high prevalence of injecting

drug users and needle sharing, as well as unprotected sex amongst injecting drug users, raises

the possibility of an HIV epidemic (Dehne et al, 1999; de Jong, Tsagarelli & Schouten, 1999),

as does the risk of infection through sexual relationships with vacation visitors to the Black Sea

 coast (Dehne et al, 1999).

In Central Europe increase in HIV infection is generally relatively slow (Dehne, Khodakevich,

Hamers & Schwartlander, 1999), and in Hungary and Poland, there is little evidence of a marked

increase in AIDS incidence (European Centre for Epidemiological Monitoring of AIDS in Europe,

1999). Much of this might be related to the better economic conditions and low rates of drug

injection and prostitution in this region in comparison with the Eastern countries and Caucus

region (Dehne et al, 1999) Transmission in Hungary has been largely amongst homosexual and

bisexual men. At the end of 1997 an estimated 2000 people were living with HIV/ AIDS in Hungary

with 328 confirmed AIDS cases and rates of STDs such as syphilis remain low (Joint United Nations

 Programme on HIV/AIDS, 1998b). In Poland, 794 confirmed AIDS cases were reported at the end

of 1997, with an estimated 12 000 people living with HIV/AIDS (Joint United Nations Programme

on HIV/AIDS, 1998b). Here, drug use accounts for the majority of cases and there is only limited

evidence of infection outside the drug using population (Bingham & Waugh, 1999; Dehne et al,

1999). Infection amongst the gay population, however, is likely to be underestimated because of

the stigma attached to being gay in Poland (Joint United Nations Programme on HIV/AIDS, 1998b).

 Overall sexual disease rates in Poland have been relatively low and stable (Bingham & Waugh,

1999) but there has been an increase in sexual disease in Western Poland as a result of prostitution

at the borders of the Czech Republic and Germany and in Eastern Poland resulting primarily from

contact with visitors from the Former Soviet Union (Chodynicka, Serwin, Janczylo-Jankowska &

Waugh, 1999).

Alongside the Ukraine, the Russian Federation is the area where there is the most marked escalation

in the epidemic in this region (Ingram, 1996; Kalichman, Kelly, Sikkema et al., 2000; European Centre

for Epidemiological Monitoring of AIDS in Europe, 1999: see figure A below). During the early phases

of the epidemic in the late 1980s and early 1990s, contraction of the virus was predominantly through

contact with foreigners from higher risk regions or through homosexual transmission, and numbers

infected were small (Dehne et al, 1999). As a result of the degeneration of the health care system

following the collapse of Communism, and the fear of stigmatisation associated with STDs, a number

of people with STDs sought treatment outside the state sector, attending poorly qualified private

doctors or institutions or attempting to treat themselves (Renton, Borisenko, Tichonova et al, 1999).

The epidemic began in 1995 primarily amongst drug users (Rhodes, Stimson, Crofts, Ball, Dehne &

Khodakevich, 1999). In 1995 only 7/1062 cases of infection were drug-related: this increased rapidly so

that by the beginning of 1998 almost 80% of cases were drug related (Dehne et al, 1999). An estimated

40,000 people were estimated to be living with AIDS/HIV at the end of 1997 (Joint United Nations

Programme on HIV/AIDS, 1998b). In the towns and cities around Moscow, HIV infection rate increased

five times in the first nine months of 1999, and in some towns this rise has been spectacular. Alongside

this, new cases of syphilis increased 63 fold between 1988 and 1996 (Joint United Nations Programme

on HIV/AIDS, 1998b; Kalichman et al, 2000) and gonorrhoea rates have shown a similar rise (Kalichman

et al, 2000). More recently there has been a growth in anonymous-testing dermatovenereological facilities

in Russia, and testing has now been made ‘voluntary’, although some routine testing does continue

amongst drug and sex workers and mandatory screening of blood remains for professionals exposed to

HIV (Rhodes, Ball et al, 1999b). There is, however, still wide variation in care for STDs received in Russia

with much depending on the ability of the infected individual to pay for treatment (Renton, Borisenko,

Tichonova et al, 1999).

Figure A: New HIV infections in Russia, 1993 – 1997 (data from Borisenko et al, 1999)

3.      Previous research on factors underlying the spread of HIV and knowledge of HIV

Despite the prospect of a marked increase in HIV infection, little research has been conducted into sexual

health and the wider socio-political factors associated with it in this region. Nevertheless it is evident

that the spread of HIV infection is subject to a range of social, cultural and political influences (Lear, 1995

; Rhodes, Stimson et al., 1999). Important contextual factors, including recent market reforms and the

consequent increase in income inequalities, have led to profound changes not only in the practices and

resourcing of health services across the region but also in patterns of sexual behaviour (Borisenko et al,

1999; Kon, 1995; Rhodes, Ball et al, 1999).

Rhodes and his colleagues (Rhodes, Ball et al, 1999; Rhodes, Stimson et al, 1999) identify a number of

features of the ‘macro risk environment’ that act as major factors in sustaining epidemic growth and mediate

 the efficacy of prevention responses. One major factor of course is the rise of prostitution which has

been associated with increasing rates of poverty and unemployment (Bingham & Waugh, 1999; Borisenko

 et al, 1999; Rhodes, Ball et al, 1999), as well as the growing trend in temporary sexual partnerships as a way

 of surviving economically (Renton, Borisenko, Tichonova et al, 1999; Rhodes, Ball et al, 1999). Inequalities

in income have increased dramatically, with wide-scale social disruption linked to the marked decline in life

expectancy (life expectancy for men fell more than 6.7 years between 1989 and 1994: Bobak, Pikhart, Hertzman

et al, 1998).^. A second factor is the expansion of trade and migration with an increased mixing of populations

(Borisenko et al, 1999; Rhodes, Ball et al, 1999), some of which has been directly linked to the break-out of

sexual diseases such as syphilis (Rhodes, Ball et al, 1999). In countries such as Georgia, where HIV rates are

still low, the opportunity for HIV infection through injecting drug use has increased through the development

of drug trafficking routes (Dehne, Khodakevich, Hamers & Schwartlander, 1999; Rhodes, Stimson et al, 1999).

In Poland, sexual disease infection in general is strongly related to sexual relations with foreigners, particularly

from the former Soviet Union, with the increase most marked in the Eastern sector of the country (Chodynicka

et al., 1999). Third, Rhodes, Ball et al (1999) point to an increasing individualisation and change in community

values that emphasise greater sexual freedom (see also Goodwin, 1998, on the erosion of traditional support

networks and trust and Walberg, McKee, Shkolnikov, Chenet & Leon, 1998, on social cohesion and mortality

in Russia). Fourth, persisting gender-role socialisation has led to some acceptance of sexual coercion against

women, with sexual violence a further increase risk factor for HIV (Horne, 1999; Kalichman et al, 2000). Finally,

Kalichman et al (2000) point to the sense of hopelessness and fatalism that have accompanied economic

instability and rising poverty. Such a sense of hopelessness and fatalism they claim has helped promote risky

behaviours.

Despite considerable efforts dedicated to screening for sexual disease, tracing of sexual contacts and enforced

hospitalisation in the years prior to the fall of Communism (Renton, Borisenko, Tichonova et al, 1999) the limited

investigations that have been conducted have demonstrated a great deal of misinformation about sexual matters

and AIDS prevention (Axmann, 1998; Visser & Ketting, 1994: see for example, Stark, Sieroslawksi, Wirth &

Godwood-Sikorska, 1995, in Poland; Attwood, 1996, in Russia). Thus for example, Lunin, Hall, Mandel, Kay &

Hearst (1995) found that only 29% of Russian 16 year olds thought condoms should only be used once. Only

16% of commercial sex workers studied in Moscow understood HIV to be sexually transmitted, and a further 30%

believed oral contraceptives and intrauterine devices were effective against HIV (Kalichman et al, 2000). Some of

this ignorance can be traced to the continuing taboos associated with the discussion of sexual matters in both

former Communist and post-Communist times (Goodwin et al, 1999). In addition, of course, there have been

continuing problems in funding health promotion and health education (Renton, Borisenko, Tichonova et al, 1999).

HIV represents a highly politically charged syndrome rooted in cultural and political constructions (Joffe & Dockerell,

1995; Nelkin, Willis & Parris, 1991). In Georgia, an article in Akhali Taoba (December 3^rd 1996) reporting official

numbers of AIDS cases was contradicted a week later by the Department of Social and Economical Information

which reported there were no AIDS cases in Georgia (Kavkasioni, December 11^th 1996). The official Communist

Party view was to see those with STD infections as morally undesirable, and infected persons in Russia who

refused treatment, or had sex after being notified of their infection, were prosecutable under Article 115 of the

Criminal Law (Renton et al., 1999). More recently, a wide-spread belief that HIV is an ‘outsiders’ problem associated

the ‘decadent West’ (Transitions, 26th May, 1995) contributed to the controversial AIDS law requiring the compulsory

 testing of foreigners visiting Russia for more than three months. In Poland, the strongly influential Catholic Church

 has blocked a number of attempts to encourage discussion about sexuality in the country (Danziger, 1996). Here

the church's influence has become particularly pronounced in the aftermath of the last general election won by a

coalition of those integrated around so-called "Christian values". This has led to fears of further restrictions in

the provision of sex education.

1.4. Assessing AIDS beliefs: the theory of social representations.

Most of the existent research on AIDS uses survey instruments which attempt to link individual knowledge,

attitudes and beliefs to sexual practices. However, there is only limited evidence that AIDS-related knowledge

and attitudes shape AIDS-related practice (Joffe, 1996), and these theories explain little of the ‘moral panic’ that

allows individuals to distance themselves from high risk groups and expose themselves to risk (Lear, 1995). One

particular problem with this paradigm is the assumption that individuals make considered, rational health choices,

 whereas sex is emotional and involves a high degree of arousal and interpersonal negotiation (Abraham & Sheeran,

1994). As a result traditional models fail to capture the social and non-conscious dynamics that influence the

individuals' beliefs and practices (Breakwell, Millward & Fife-Schaw, 1994; Joffe, 1996; Rhodes, Stimson., et al,

1999). There is mounting evidence that sexual decisions are based upon a number of factors including the degree

of commitment to the relationship, friendships and wider social relationships (Flowers et al., 1996; Joffe, 1997;

Rhodes, Stimson et al., 1999). In addition, even professional groups have a notion of with whom it is ‘safe’ to

have sexual relationships (Kitzinger, 1991), while the predominant image of an AIDS patient as a haggard,

painfully thin, skeletal figure renders the person with AIDS non-human and distant. Such perceptions can

undermine campaigns which stress the lack of difference in appearance between those who are and are not

infected (ibid.).

One alternative perspective to the dominant knowledge and beliefs paradigm is the theory of social representations

 (Moscovici, 1984), an approach which plays an increasingly major role in British and Western European social

psychology. This approach has used a number of methods to examine the myths, beliefs and values held by a

group, and the consequences of these for social practice. Research informed by this approach is particularly

concerned with the way in which different groups protect their different identities by way of group specific

representations, both minimising their own perceptions of risk and allowing them to perpetuate existing prejudices

 against others. Studies of HIV using this perspective have began to tap into the way in which people `make sense'

of an otherwise abstract scientific concept, anchoring it within existing models of disease and out-group

discriminations (Joffe, 1996; Kitzinger, 1995). This is particularly pertinent for the situation in Central/ East Europe,

in which the recent increase in HIV/AIDS has been accompanied by widespread discrimination against people with

HIV/AIDS (Danziger, 1994).

Social representations of HIV operate on a number of conceptual levels, helping protect culture and sub-group

identities as well as serving the individual in affirming his/her actions and beliefs and justifying individual experiences

(Breakwell, 1986; Stephenson, Breakwell & Fife-Schaw, 1993). In order to examine social representational processes

 in the context of both group processes and individual experiences, social representations research increasingly

involves a diversity of methodologies (Breakwell & Canter, 1993; Doise, Clemence & Lorenzi-Cioldi, 1993). In this

 work we reflect the call for a greater diversity of methodologies in the field of sexual disease by questioning

participants from a number of different cultures using questionnaires and interviews to provide information on

individual beliefs and experiences and group membership and perceptions (see Axmann, 1998; Rhodes, Stimson

et al, 1999). We also use free associations and matrices from a multi-dimensional scaling task to examine

 individual-level differences in representations and relate these to sexual behaviours. Finally, we use media

analyses to investigate cultural variations in representations of HIV in the five countries.

1.5. A culture-specific approach

Despite the growth of epidemiological research into HIV infection in Eastern Europe we have little understanding

of the social, environmental and cultural factors which influence the spread of HIV (Department For International

Development, 1999; Rhodes, Ball et al, 1999). Existing data suggest clear differences between the countries of the

Former Soviet Union and Central Europe in the rates of HIV infection and in methods used to tackle the spread of

infection (Dehne et al, 1999; Rhodes, Stimson et al., 1999). One question therefore is the extent to which enduring

values in a society contribute alongside other socio-economic factors to increase or restrain the spread of the

epidemic (Dehne et al, 1999).

From an ecological perspective, different historical circumstances can be seen as leading to the development of

different individual and cultural values which can have potentially profound impacts on interpersonal behaviour.

Defining values as "desirable transsituational goals… that serve as guiding principles in the life of a person or other

social entity" (Schwartz, 1994, p. 21) Schwartz and his colleagues (Schwartz, 1992; Schwartz, 1997; Schwartz & Sagiv,

1995) have developed an elaborate circumplex model of ten individual values which allows for a sophisticated

mapping of values and their impact on social behaviours. The values were derived from the need to meet three

requirements: biological needs, the requirements of successful social interaction, and requisites for the functioning

and survival of groups (Schwartz, 1994). These 10 value-types have demonstrated a consistent relationship with

one another in more than 70 cultures (Schwartz & Sagiv, 1995), and can be organised along two dimensions:

(i) Openness to Change (Independent thought and action) versus conservatism (Conformity, Tradition and Security)

 and (ii) Self-transcendence (universalism and benevolence) versus Self-enhancement (achievement and power).

A final value, Hedonism, is related to both Openness to Change and Self-Enhancement. These values are the primary

values used for analysis in the present study.

Schwartz and his colleagues have also examined these values as a set of seven, related cross-cultural indices. These

values are now given new names and formed into three new dimensions. The egalitarian-hierarchy dimension

examines the manner in which societies assure responsible social behaviour: egalitarian cultures encourage equality,

 helping others and feeling concern for everyone’s welfare whereas hierarchical cultures emphasise the fulfilment of

ascribed roles and authority and the obedience of ‘the rules’. A second dimension, harmony-mastery, contrasts the

extent to which cultures emphasise ‘fitting in’ with the physical environment (a harmonious culture) or attempt to

master the environment through self-assertion. A final dimension, autonomy-embeddedness (conservatism) concerns

 the relationship between the group and the individual and is similar to the contrast between individualist (autonomous)

vs. collectivist (embedded) cultures (Smith & Schwartz, 1997). In cultures high on intellectual autonomy cultures

individuals are encouraged to pursue their own ideas, whereas the pursuance of internal feelings is stressed in high

 affective autonomy cultures. Conservatism stresses the importance of traditional groups (such as the family) in

structuring the life of those individuals in that society.

Cultural-level scores using Schwartz’s indices have been compared across Europe, and provide the most

comprehensive data set to date of the countries under examination in our study (see table 1)(Schwartz & Bardi,

1997). Differences between Western, Central and Eastern Europe were attributed to differences in the degree of

exposure and response to Communism. In Eastern Europe, conservatism values were more important than they were

in Western Europe, as were values of hierarchy. Egalitarianism, Autonomy and Mastery were more developed in

Western Europe than the East. Those countries most heavily exposed to Communist rules were most likely to have

developed conservative and hierarchical values, whilst those in Central Europe, which were exposed to the Communist

 system for a shorter time period, were more likely to develop egalitarianism, intellectual autonomy and mastery values.

On the basis of the culture-level scores reported above, we first formulated a number of hypotheses about the likely

ranking of the nations in our sample when examined at the cultural-level. First, we anticipate that Georgians and Poles

 should be the most Conservative (Hypothesis 1). Russians are likely to the most Hierarchical (Hypothesis 2) and

Georgians and Estonians the most Egalitarian (Hypothesis 3). We would expect the Poles to be the highest on

Mastery (Hypothesis 4) and the Estonians the highest on Harmony (Hypothesis 5).

How might Schwartz’s values relate to individual sexual behaviour? At the individual level, those particularly Open

 to Change, who stress a pleasurable, varied and daring life, might be expected to have more sexual encounters and

have had a greater exposure to sexual disease than those emphasising the Conservatism dimension (Conformity,

Tradition and Security) (Hypothesis 6a). Those high on this dimension are also less likely to have used contraception

(Geis & Gerrad, 1984; McCormick, Izzo & Folcik, 1985) (Hypothesis 6b). Sensation seekers seek intense and novel

sensations and experiences and are willing to take risks for the sake of such experiences (Zuckerman, 1994; Donohew,

Zimmerman, Cupp, Novak, Colon & Abell, 2000). Donohew et al (2000) found that those high in sensation seeking

were more likely to have unwanted sex under pressure and when drunk. Using the related value of Hedonism, we

would predict that those high on Hedonism, who seek pleasure and sensual gratification for themselves, would also

be more sexually adventurous (Hypothesis 7). The relationship between Self-enhancement / transcendence and

sexual behaviour is however less clear.

In this study we also include one further and related measure of culture values, collectivism (see Smith & Schwartz,

1997; Singelis, Triandis, Bhawuk & Gelfand, 1995; Triandis, 1995 for a discussion of the overlap between these

cultural dimensions). Collectivism is a widely employed concept in cross-cultural psychology and has been assessed

at both the individual and group level (Triandis, 1995). Collectivists "have a tendency to accept authority, honour

tradition, and are generally conservative...they find change difficult and prefer to stick with tried and true" (Realo,

Allik, & Vadi, 1997, p. 113). We would thus expect collectivists to be more ‘traditional’ and less likely to experiment

with multiple sexual partners (Hypothesis 8a). In particular, we might hypothesise that those high on familism will

be the most sexually conservative, as "family security, honoring parents and elders, respect for traditions serve as

guiding principles in familists' life" (Realo et al., 1997, p.110) (Hypothesis 8b).

1.6. Additional predictors of sexual activity and perceived vulnerability to HIV

The final measures used in our study were added as additional individual-level predictors of sexual behaviour.

These reflect well-established concepts in the analysis of social behaviour which have not been previously

applied to the study of sexual disease in this region.

Fatalism assesses individual's beliefs in their ability to control their general fate. In earlier research in Central and

Eastern Europe fatalism proved to be a strong predictor of the ability to establish close relationships (Goodwin,

1998, Goodwin et al, 1999). Kalichman et al (2000), commenting on the rise of HIV in Russia, cite US research

demonstrating that fatalism promotes HIV risk behaviour in several populations. Here we hypothesise that fatalism

scores will be correlated with a willingness to engage in unsafe sexual activities (Hypothesis 9).

The concept of optimistic bias examines the beliefs people hold about their susceptibility to harm (Weinstein, 1987,

p. 481). Weinstein (1980, 1982) anticipates that such optimism might lead to a failure to practice appropriate

risk-reducing strategies, and Taylor, Kemeny, Aspinwall, Schneider, Rodriguez & Herbert (1992) report that

dispositional optimism was associated with fewer worries and concerns about AIDS and a perceived lower risk

of AIDS. Using the well-established optimistic bias scale (Weinstein, 1987) we asked respondents to compare

their own risks with relevant others on a range of hazards (e.g. contraction of cancer). Scores on this scale can

then be related to the sense of invulnerability to HIV/AIDS reported by our respondents, with those high on

optimistic bias less likely to report themselves at risk (Hypothesis 10)(Bauman & Siegel, 1987; Taylor et al, 1992).

However, the relationship between optimism and actual behaviour is less clear, with Weinstein’s (1982) assertion

that dispositional optimism might undermine effective health behaviour challenged by the empirical findings of

more specific research on HIV-related indices of optimism (Taylor et al, 1992).

Finally, we consider the notion of interpersonal trust. Levels of interpersonal trust have been widely held to be

low in the post-Communist regimes of Central and Eastern Europe (Hedlund, 1999). Although interpersonal trust

is a central aspect of sexual behaviour (Lear, 1995; Stephenson, Breakwell & Fife-Schaw, 1993; Willig, 1997) the

implications of a sense of trust in others for sexual behaviour has not, to out knowledge, been previously explored.

 We thus leave the relationship between dispositional optimism, trust and sexual behaviour as open research

questions to be explored in our studies.

1.7. Research Participants

Our participants were from two groups, business people and medical workers (doctors and nurses), in the five

countries. Business people were selected as a highly mobile group likely to visit sex workers (Wellings et al.,

1994) a group with a relatively high incidence of HIV infection in this region (Towianska, Rozlucka & Dabrowski,

1992). Medics have direct contact with people with HIV/AIDS and act as a vital conduit of information about the

disease, making their representations potentially highly influential in the broader society. All our respondents in

this work were aged between 25 and 57 (mean age 33.2, SD 7.1), falling into the groups from which there is the

greatest proportion of reported AIDS cases in the five countries studied (Joint United Nations Programme on

HIV/AIDS, June 1998b).

In the first stage of the research, 50 medical workers and 50 business people were recruited in each of the five

countries (total N for stage 1= 511). For the second stage of the research, a fresh sample was collected, although

just over one-third (35%) of the former participants also participated in the new data collection (total N for stage 2= 503).

Data was principally collected in capital cities (Budapest, Hungary; Tblisi, Georgia; Moscow, Russia) but some

additional data was also collected in the relatively high infection areas of Eastern Poland (Bialystok: Chodynicka

et al., 1999), the towns of Kutaisi and Batumi in Georgia and in St. Petersburg, Russia (Kalichman et al, 2000).

Data from Estonia was mainly collected in Tartu, the site of the only medical school in the country and in Tallinn,

the capital city.

1.8. Overview of the research

The research described in this project was divided into four parts, with a triangulation of methods aimed at

providing both internal and external validation for our findings (see Lear, 1995, Goodwin, 1999, for a discussion

on triangulation methods in HIV/AIDS research). In stage 1 we conducted semi-structured interviews and free

association analyses to investigate how different groups of interviewees represented the origin and spread of

the virus. In stage 2, we built on the free-associations collected to provide our respondents with matrix tasks

which allowed us to formulate cultural, occupational and individual-level representations of HIV amongst our

different populations. Through the collection of additional questionnaire data we related these representation

scores to sexual behaviour, and examined the relationship between values, unrealistic optimism, fatalism, trust

and knowledge and reported sexual activity.

Stages 3 and 4 of the research attempted to contextualise these findings within their wider social and cultural

environment. In stage 3 we examined how the newspapers in these five societies portray HIV/AIDS. In particular

we were concerned with the way in which there were systematic differences in the representations of HIV/AIDS

in the newspapers in the different nations, and in those newspapers read by different sections within these

societies. In the final stage of the research (stage 4) we collected additional data to examine sex roles, stereotyping

 and personality issues in these countries. We also describe an additional study of schoolchildren and shelter

residents in Russia aimed at providing additional insight into the wider implications of our findings amongst some

particularly vulnerable sectors of the population.

Section 2: Research stage 1: interviews and free associations

2.1. Method

2.1.1. Participants

The sample that participated in this stage of the study is outlined in table 2. The majority of our business

respondents were well-educated although the number attending University ranged from 34% in Estonia to

90% in Georgia. Respondents were primarily observant Christians and were predominantly Catholic and

religious in Poland (71% described themselves as observant Christian believers) and in Hungary (62%

described themselves as Catholics). In Russia, 70% described themselves as ‘Christians’ or ‘Orthodox

Christians’ and in Georgia all respondents were Orthodox Christians. In Estonia, levels of religious affiliation

were the lowest, with only 21% of the sample describing themselves as religious believers (with the majority

of these describing themselves as Lutherans).

2.1.2. Procedure

Respondents were asked a total of seventeen questions (see appendix 1) derived primarily from the work of

Joffe (1996). Questions were grouped into three clusters: four questions about the origin and spread of

HIV/AIDS, seven questions about the ‘type of person’ likely to be infected and six broader questions about

 the government’s role in caring for those infected by HIV and the ‘sexual morality’ of people in that country.

Respondents were also instructed to write

"everything that comes into your mind when I say the word AIDS. Please write down as

many things as possible. Please be as unrestricted as possible" (after Doise et al., 1993).

Respondents were requested to write at least six words or phrases. Finally, respondents completed a short

demographic questionnaire ascertaining age, sex, level of education and religion. To aid the media analysis

described below (stage 3), respondents also indicated the newspaper/s they read and the frequency of their

reading. Interviews in the five countries were conducted during early 1999.

2.2. Results:

2.2.1. Interview analysis

Because our primary aim in this first stage of the research was to provide a broad overview of perceptions

of HIV across our samples, interview responses for each questions were collated across the five countries

by all the project participants during a project workshop in Kutaisi, Georgia. Broad themes were identified

for responses to each item from the complete sample. A full list of codings for each question, and the number

 of respondents citing any one response, is provided in Appendix 2.

From this analysis a number of patterns emerged. Russian respondents were generally most optimistic about

the ‘AIDS problem’ in general, whilst information about HIV/AIDS was sourced from different information

sources dependent on the group questioned (unsurprisingly medics were more likely to have learnt about

HIV/AIDS in their educational environment or at work). Polish respondents (and, on a later question,

Georgian business respondents) were most likely to refer to the ‘socially maladjusted’ as a high risk group.

 Polish respondents were also more likely to view ‘modern women’ as the likely carriers of condoms, again

a response not characteristic of our other respondents. Responsibility for dealing with those who are HIV

infected was seen as more widely shared by the Hungarian and Georgian business people, whilst charities

were seen as having a greater role in helping those infected by the Georgian business participants. Russian

 respondents were the most sceptical about the ‘moral health’ of their country.

Cluster analysis

Cluster analyses were conducted by Dr. Kwiatkowska on the interview responses to aid us in grouping

responses and to identify differences across sample groups. Cluster analysis was used to classify objects

 into categories, with the 5 (nation) x 2 (professional groups) clustered on the basis of similarity between

categories of answers to each interview question. Objects were combined using Agglomerative Hierarchical

 Clustering where clusters are formed by grouping cases into bigger and bigger clusters until all cases were

 members of a single cluster. We used the method of Average Linkage Between Groups, which defines the

distance between two clusters as the average of the distances between all pairs of cases in which one

member of the pair is from each of the clusters. Distance/similarity measurement was using the Phi-square

measure – this is for frequency-count data, and is based on the chi-square test of equality for two sets of

frequencies. As phi-square is the measure normalized by the square root of the combined frequency its

value does not depend on the total frequencies of the two items whose distance is computed.

A listing of cluster results for each question is included in Appendix 3. Two and three dimensional cluster

summaries are reproduced below (figure B):

Figure Bi: Summary solutions

Two-cluster solution