Conference details

Training workshop

Research teams

Contributed paper sessions

Conference papers

IUSSP homepage

Chiang Mai Conference on Partnership Networks (February 2000)

Contributed Papers


Michael Bracher, Susan Watkins and Gigi Santow

How important are high-risk groups for maintaining the momentum of the HIV/AIDS epidemic?

HIV/AIDS interventions have typically targeted high-risk groups, which in
sub-Saharan Africa have generally been identified as prostitutes and
long-distance truck drivers.  The logic of focusing on such groups is
two-fold.  First, they are small, visible, and obviously at high risk.
Secondly, targeting such groups for intervention may be more politically
palatable than targeting other groups, such as married couples.  Once the
virus has gained a foothold in the population, however, the existence of
such highly mobile and sexually gregarious groups may be less critical for
the subsequent course of the epidemic than the sexual interaction of members
of the general population.

In this paper we employ microsimulation to examine the rapidity with which
HIV spreads through a population given different patterns of sexual
behaviour and partner exchange, and different sizes of catalytic high-risk
groups.  We use information from surveys of women of reproductive age and
their husbands conducted by Susan Watkins and Eliya Zulu in Kenya and
Malawi.
 
 

R.S. Goyal, Satish Kumar, Sunita Nigam
(Indian Institute of Health Management Research, Jaipur, Inida)

Promoting Sexual Health Through Dialogue Between Men And Women Within Social Networks

This paper presents the findings of an operations research study undertaken to assess the perception, attitude and behaviour of men and women towards sexual health, to understand their information needs on this subject and, to meet these need through dialogue between men and women within same social network.  The project was sponsored by Family Health International, USA.  Against the backdrop of a low literacy level and limited exposure to mass-media, interpersonal communication i.e., `dialogue’ within and between the sexes  could be an effective tool to achieve behaviour change.  The dialogue would be more effective within the members of same social network, because the social pressures for attitudinal and behavioural modifications would be more strong in such settings.

The project was implemented in one rural and one urban localities of Jaipur district (India) in the middle of 1997.  The study showed that both men and women had a very poor perception of their sexual health and sexual health needs.  After initial hesitation, both men and women could frankly discuss sexual health and related issues in dialogue sessions of people of the same sex and of both sexes. The community response to these initiatives was very positive.  The elders in the family (particularly mother-in-low) were initially hesitant to permit their daughters-in-law to participate in the dialogue sessions particularly of both sexes, but they agreed on persuasion.  There was a large and effective diffusion of `dialogue’ in the community at large.

The study shows that dialogue is an effective mode to promote sharing of perceptions and experiences among men and women.  It is also effective in creating awareness and validating scientific knowledge.  It is particularly helpful to people in understanding ones own and appreciating others’ (opposite sex) sexuality.  It also contribute to modifying opinions and attitudes of men and women towards safe sex. More importantly, dialogue has helped in creating a positive environment for adoption of safe sex practises and strengthen the women’s ability to negotiate for less violent and safe sex, within the married unions.
 
 

Simon Gregson, Constance Nyamukapa, Geoffrey P. Garnett, Tom Zhuwau, Violet Matimba-Masuku, Roy M. Anderson and Stephen K. Chandiwana

Sexual Behaviour Patterns and the Rapid Rise in Age-Specific HIV-1 Prevalence among Young Women in Rural Zimbabwe

Background:  A population survey of HIV and associated patterns of sexual behaviour is currently in progress in rural areas of Manicaland Province, Zimbabwe. Early results indicate that HIV prevalence increases very rapidly between the ages of 16 and 25 in women but more slowly and at later ages among men. A part of the explanation for the contrasting age-pattern of infection lies in higher male-female than female-male HIV transmission probabilities. However, differences in patterns of sexual behaviour between young men and young women could be a further contributing factor.
Objectives:  To describe the sexual activity and sexual mixing patterns of young men and women in rural Manicaland. To investigate the extent to which these patterns of sexual behaviour could explain observed gender differentials in HIV prevalence at young adult ages.
Methods:  Quantitative data on HIV infection and sexual behaviour among young adults are being collected in a population-based survey. The latter include data on onset and degree of sexual activity, numbers of partners, concurrent partnerships, and condom use and partner characteristics within recent sexual relationships. These data will be validated and interpreted using information from a parallel programme of qualitative research. Statistical relationships between aspects of self-reported sexual behaviour and the presence of HIV infection will be investigated. Mathematical model simulations will be developed using the empirical sexual behaviour data to test the hypothesis that contrasting patterns of sexual behaviour contribute to the more rapid rise in female age-specific HIV prevalence. 
Results:  Preliminary results indicate that males generally start having sex slightly later than females and are less sexually active at young ages. Young women typically form partnerships with men 5-10 years older than themselves while those young men who are sexually active usually form partnerships with women of a similar age. Firmer and more detailed results will be given in the paper and their implications for HIV prevention will be discussed.
 
 

Laura Koehly, Martina Morris and Steve Goodreau

Random Graph Models for Contact Networks:  Loglinear models, Logit models and P*

Morris (1991) introduced log-linear models as a method for representing selective and proportional mixing in partner contact matrices.  These models assume dyadic independence, and thus ignore micro-level relational structure.   The p* models recently described by Wasserman and Pattison (1996), provide a more general framework for modeling partnership structure.  This approach allows for more sophisticated modeling of the social interactions among network members, while simultaneously accounting for partnership attribute composition.  The loglinear models can be shown as special cases of the more general random graph models represented by p*.  The common exponential form makes it possible to use this modeling framework to investigate the role of network microstructures and aggregate mixing biases on disease transmission dynamics.
 
 

Hans-Peter Kohler, Jere Behrman and Susan Watkins

The Determinants of Partners for Conversations about AIDS

In this paper we consider the selection of partners for conversations about AIDS, using data from Kenya and Malawi. We do this by comparing the determinants of network characteristics for AIDS and family planning conversational networks. We develop a theoretical model of these determinants, and estimate it using data from household surveys conducted in rural areas (approximately 800 women and their husbands in Kenya, and 1600 women and their husbands in Malawi) as well as qualitative interviews (some of which will be conducted in June-July 1999).

In both countries, high proportions of both men and women have talked with others about both family planning and AIDS (for example, in Malawi 94% of men and 84% of women had talked with other people about AIDS, with a slightly lower proportion talking about family planning).  To a considerable degree, network partners are homophilous: they are local,  and approximately 50% in both types of networks are confidants (with the rest being friends or, a smaller proportion, acquaintances).  And both networks are dense: the network partners know each other.  AIDS networks are, however, somewhat larger and younger than family planning networks, and the network partners have a slightly higher level of education.

The qualitative data show that to a considerable extent these conversations occur casually rather than purposefully, in the course of conversations which may cover a range of topics.  Yet the differences in the network characteristics and the qualitative data suggest that some women and men choose at least some of their network partners deliberately, either because they are expected to have relevant experience or knowledge, or because the respondent wants to gauge their opinion (e.g. of the morality of using modern family planning or condoms).  To the extent that network partners are deliberately selected, it would appear likely that different selection criteria come into play.

AIDS and family planning networks appear to be similar in that for both what is relevant is social learning and social influence about an innovation (modern contraception and aids prevention behavior): the individual is seeking an answer to the question "what do you do given that you intend to limit/space births (in fp) or prevent AIDS with a potentially effected partner.  They differ, however, in that an important additional aspect of AIDS conversations appears to be the evaluation of risk groups within a community.  Qualitative interviews conducted in rural areas in both countries suggest that for many neither strict marital fidelity nor regular condom use are acceptable; rather, those considering extramarital sex and/or condom use turn to others in order to attempt to estimate the likelihood that a particular potential partner is infected, and thus whether the partner should be avoided or whether condoms should be used, and for how long.  There appears to be general agreement of levels of risk: for example, women from the large cities are risky, as are bar girls, but village women are less likely to be infected.  If these expectations are correct, we should find that the respondents' mental models for choosing conversational partners for discussions of AIDS should be different from their models for discussions of family planning.
 
 

James Moody
Department of Sociology, The Ohio State University

Indirect Connectivity and STD Infection Risk: The Importance of Relationship Timing for STD Diffusion.

Sexual networks have a peculiar temporal feature: while each dyadic relationship is symmetric, the indirect relations implied by partners of partners are not symmetric.  Indirect asymmetry follows from the fact that infections can only flow forward through time-ordered paths.  Thus, one's past partners are not at risk to one's current infection while one's future partners are.  This temporal ordering is a key element needed to understand disease diffusion in the population.  Depending on the timing of relations, the same contact network may result in dramatically different sets of people exposed to the STD.  In this paper, I first show how the timing of sexual relations determines potential exposure and provide a method for deriving upper and lower bounds for indirect contact within networks of a known structure.  I then apply the method to data on romantic and sexual networks in the United States, showing how relationship timing both limits the ultimate exposure within the network and creates pools of people who are mutually connected indirectly, thus forming pockets where an STD could circulate.  This work implies that understanding STD diffusion requires collecting data on the timing as well as the pattern of sexual relations. 
 
 

Soori Nnko, Ties Boerma, Mark Urassa, Basia Zaba

Sexual networking in Kisesa, Tanzania

In  1994 a cohort study was initiated in Kisesa community in northwest Tanzania. A survey including all adults 15-44 years was carried out in 1994/95 and repeated again in 1996/97. Both surveys included a series of questions on marital and sexual behaviour of the respondent, including number of sexual partners in the last month and in the last year. During the second survey a sexual networking module was added to the general questionnaire and was administered at the end of the interview. The questionnaire was adapted from the UNAIDS module for tracing sexual networks. Questions about the partnership characteristics were limited to the last five partners during the last 12 months. Data were also collected on HIV status in both survey rounds.

In total 3,700 respondents were asked the questions on sexual partnerships. 58% of sexuall active men and 22%  sexually active women reported at least one non-marital non-cohabiting partnership. Among those who reported at least one partnership the average number of partnerships was 2.1 for men and 1.2 for women.

The analysis will describe the sexual networking patterns in Kisesa. In the first part the patterns of mixing are described: by age, by marital status, by place of residence, etc. Measures of concurrency will be presented.  In the second part, special attention will be paid to the assessment of the validity of data on multiple partnerships reported by men and by women. Even though women report much lower frequencies of multiple partnerships preliminary analysis indicates that women's partnership patterns are potentially a more sensitive indicator of sexual networking in the population than the corresponding male data. In particular, the assessment will focus on changes in partnership patterns by sex of the respondent as an indicator of changes in sexual behaviour and the risk of HIV.
 
 

I.O. Orubuloye
Centre for Population and Health Research, Ado-Ekiti, Ekiti State

Researching sexual behaviour and identifying sexual networks in the shadow of AIDS in Nigeria.

Before the advent of the AIDS epidemic, there was very limited research on sexuality and sexual relations. Earlier social science researchers were reluctant to investigate pattern of sexual behaviour because of the sensitivity of the topic and the fear of hurting their respondents and thus jeopardizing future research. The AIDS epidemic in Africa has given rise to a need for research into sexual behaviour ans sexual networks because of the heterosexual nature of the epidemic. Since 1989  the West African Research Group on Sexual Networking consisting of researchers from Ghana, Nigeria and Uganda has been investigating how many different sexual partners men and women have both inside and outside marriage and how such relationships are distributed overtime. The paper will report on how the studies were conceived, how the survey instuments were developed and tested and the various methodological approaches adopted in investigating these relationships among the general population and specific high risk groups such as sex workers, truck drivers, women traders, prison population, police etc. The problems and methodological advances in the social context of the AIDS epidemic will also be discussed.
 
 

Alex Weinreb
Population Studies Center, University of Pennsylvania, USA

Interviewer Effects in Network Studies

This paper examines the effect of interviewers knowing their respondents in network studies on social relationships and personal behavior.  Three findings emerge.  First, insider-interviewers were much more successful in getting data on network partners.  Second, a substantial amount of response variance on critical dependent variables was associated with the difference in interviewer-stimulus (ie. insiders vs. strangers, men vs. women).  And third, the joint effect of these two factors in analysis leads to the estimation of radically different coefficients on certain network characteristics when the contribution of different types of interviewers to the data is empahasized.  This implies that results are contingent on the type of interviewer that is used, not a surprise but certainly worth bearing in mind given the norm of using stranger-interviewers in survey research. 
 
 

Helene A.C.M. Voeten1, E.I. Meester1, O.B. Egesah2, J.D.F. Habbema1
1 Department of Public Health, Erasmus University Rotterdam, The Netherlands
2 Nyanza Provincial Medical Office/Faculty of Medicine Eldoret University, Kenya
Financial support STDSIM project: European Commission DGVIII

Possibilities and impossibilities of method triangulation in the study of sexual networks

The Department of Public Health of the Erasmus University Rotterdam in the Netherlands has developed a micro-simulation model called STDSIM, as a tool to support decision-makers and health planners in the prioritisation of STD/HIV interventions. Nairobi and the rural province of Nyanza in Kenya have been first sites of application. To fill in gaps of knowledge and to quantify the model, applied research has been set up in Nyanza beginning of 1999, focusing on sexual behaviour patterns and networks of young adults aged 15-29, commercial sex workers and their clients. The following multiple methods were used:  1) household based study of young adults: questionnaires followed by in-depth interviews with a sub-set of the same respondents; 2) CSWs: focus group discussions followed by either a questionnaire or an in-depth interview, which were again followed by a 2-weeks sexual diary keeping; 3) CSW clients: informal discussions mainly in bars. Data were/are being analysed with Epi-Info, SPSS and NUD*IST (package for qualitative analyses). In addition, qualitative data (focus group discussions, in-depth interviews and informal discussions) have been translated into quantitative data-sheets, to enable comparison of data.

The aim of using multiple methods to investigate the same population on the same/overlapping issues (method triangulation), was to gain thorough insights in the sexual networks and to validate the self-reports from surveys.  A number of difficulties were encountered in the CSW data-analyses process, and these will be discussed here. Notwithstanding these difficulties, method triangulation can be useful in sexual network research:

  • It will give insights in the perceptions and interpretations of the study populations, which will help to interpret the data found in survey-research.
  • It will guard against a too straightforward/simplistic interpretation of (erroneous) results.
  • By trying to bring results in accordance with each other, researchers become aware of the complicated nature of their study material, which will trigger new research questions/further analyses.