Version classiqueVersion mobile
OpenEdition Books

Sexual Behaviour and Risks of HIV Infection

Michel Hubert

II. Qualitative and Quantitative Data Collection Methods in the Study of Sexual Behaviour

Telephone Surveys in Comparison with Other Methods in Psycho-Social AIDS Research

Rob. Tielman

Texte intégral

1Most of the ongoing psycho-social AIDS research makes use of other methods than telephone surveys. In this chapter, a research programme using various techniques will be described to analyse the relative position of the telephone survey method.

  • 1 The psycho-social AIDS research done by the Department of Gay and Lesbian Studies is funded by the (...)

2The Department of Gay and Lesbian Studies of the University of Utrecht started its psycho-social AIDS research programme in 1984 (Tielman & Van Griensven, 1985). The main goals of the programme are to investigate1:

  • how psycho-social factors (such as knowledge, attitudes and behaviour) determine the spread of HIV/AIDS among men with homosexual contacts and in the population in general;
  • to what degree homosexual men and other members of society who are at risk are willing and capable of making their behaviour safer, and how this relates to AIDS prevention activities;
  • whether these AIDS prevention activities diminish at-risk behaviours;
  • what the effects of being tested and of being found to be HIV-antibody or antigen positive are on sexual behaviour and the psycho-social sense of wellbeing of those involved.

3The main parts of the research programme are:

  • an Amsterdam cohort study of homosexual men at risk (by bio-medical research and self-administered questionnaires) that proved that unprotected receptive anogenital intercourse among homosexual men is the most effective way of HIV transmission and that the frequency of such behaviour fell dramatically (Van Griensven et al., 1987);
  • national cohort studies among more representative samples of men with homosexual contacts (by self-administered postal surveys) which showed similar behaviour changes among gay men in the rest of the Netherlands (Tielman & Polter, 1988);
  • small scale studies among participants in AIDS prevention activities directed towards men who have sex with men (using qualitative research methods) which showed the difficulties of some people to have safe sex (Dalles, 1990);
  • national representative sample studies of the general population to investigate knowledge, attitudes and behaviour related to AIDS (by telephone surveys) and the effectiveness of nationwide prevention campaigns (Paalman, 1988);
  • a national survey on sexuality in the general population with an oversampling of people at risk (by face-to-face interviews and self-administered questionnaires);
  • national studies among those being tested at STD-clinics (by bio-medical research and self-administered questionnaires);
  • participation in international comparative studies of men who have sex with men and of the general population;
  • studies of the medical, legal and ethical aspects of AIDS.

4Due to the fact that homosexual men were (and still are) considered to be the main group at risk for HIV/AIDS in the Netherlands, priority has been given to bio-medical and psycho-social research among men who have sex with men. The first problem related to research among homosexual men is the fact that a clear distinction should be made between sexual behaviour and sexual identification. As the Kinsey studies have shown, most men who have sex with men do not identify themselves as homo-or bisexuals. This is influenced by the fact that in most societies homosexual behaviour is stigmatized. Whereas other minorities cannot escape social discrimination (because of their skin or other visible characteristics), men who have sex with men can develop a visible heterosexual life style and at the same time have hidden homosexual contacts. This means that researchers should be very careful in selecting data related to homosexual behaviour, because there is no guarantee that the collected data are representative. Given the fact that the Netherlands has a rather unique tolerant social climate regarding homosexuality (Tielman, 1982, 1988), this created the necessity and opportunity to establish a trustful relationship between researchers and the gay community. This enabled the bio-medical and psycho-social researchers to start in 1984 the ongoing Amsterdam cohort study of men who have sex with men (Van Griensven, 1986). This group consists mainly of men living in or near Amsterdam and who were regular visitors of STD-clinics. Almost all cohort studies of homosexual men have a similar background.

5To create a more representative group of men who have sex with men, two other strategies were used to reach them. First, the researchers made use of the readership of the largest gay magazine in Dutch, De GAY Krant (circulation 50,000) to create a national cohort study. Secondly, the same was done by making use of a list of opinion leaders in the Dutch gay movement. As it turned out that there were no differences in risk behaviour, attitudes and knowledge between readers and opinion leaders, in the ongoing national cohort studies use is made of samples of the readership only. It appeared that there are important differences between the Amsterdam cohort study (showing a history of much more risky behaviour) and the national cohorts. Both the Amsterdam and the national cohort studies showed major behavioural changes (Tielman & Polter, 1988, 1989).

6As a matter of fact, the researchers were very interested in learning more about the behaviour of the unknown number of men who have sex with men without identifying themselves as homo-or bisexuals. It is here where the telephone survey method came into consideration. The main advantages of such a method could be:

  1. the anonimity of telephone conversations could stimulate the willingness to speak frankly about closeted sexual behaviour;
  2. telephone surveys (in a country like the Netherlands) can be done very quickly among representative samples of the population;
  3. regular telephone surveys can be used to give quickly feed back for policy makers on the effects of prevention campaigns;
  4. compared with other survey methods the costs are relatively low.

7Starting in 1987, every half a year a telephone survey of a representative sample of the Dutch population was conducted every six months (De Vroome et al„ 1988). The telephone surveys enabled the researchers to evaluate the effects of AIDS prevention and condom promotion activities in the Netherlands (Paalman et al.,1988) Based upon earlier research it was assumed that about 5 to 10% of the population would be willing to answer that they had recently had homosexual contacts (Tielman, 1982). During the seven surveys it became clear that on average only 2 to 4% of the representative samples were willing to answer that they had had homo/bisexual contacts recently (De Vroome et al., 1988,1990). There is no difference between older and younger people (Verhouden et al., 1990). Another comparison could be made in 1989 by using an ongoing media cohort study. Whereas in the seven telephone surveys each new sample was drawn up on a household basis (which means that no names had to be given to the interviewer), in the media cohort study the interviewers knew exactly whom they were interviewing. It happened that the expected percentage of homo/bisexuals was exactly the same percentage of those who refused to answer the question on sexual orientation, the rest all stating they were heterosexual (Tielman, 1989). The researchers are now waiting for the results of the representative survey using face-to-face interviews and self-administered questionnaires (mentioned above) to see what effects those survey methods will have on reporting sexual orientation and behaviour. In any case it became clear that the first reason mentioned in favour of telephone surveys (frank answers) did not work out as hoped for in interviewer-initiated research. There is, however, another method of telephone survey which has been used so far only in a few countries like the Netherlands, namely, respondent-initiated research. In countries with well functioning AIDS telephone help lines, many people call regularly with specific AIDS related questions. Due to the anonymity of these help lines, people tend to be very frank in asking questions. Whilst we do not get a representative picture of the general population from the statistically relevant data obtained (like sex, age and risk behaviour), we will certainly have a clear impression of the most frequent uncertainties with which the AIDS prevention activities should deal. In the Netherlands, the phoned questions most frequently asked are published regularly (with the answers given) to provide quick feedback for the general population on existing lack of knowledge.

8The second reason (representativity) turned out to be problematic in dealing with men who have sex with men, given the fact that many men did not identify their homo-or bisexual behaviour. On the other hand, it worked out very well for other purposes related to knowledge, attitudes and behaviour in the general population (De Vroome et al., 1989). Nevertheless, there are some problems here as well. This can be illustrated easily with an example less controversial in the Netherlands than sexuality, namely, life stance. About 50% of the Dutch population has no religion, and about half of them (25%) identify themselves as humanists (Tielman, 1987). Both the telephone survey of the general population (De Vroome et al., 1990) and that of young (Verhouden et al., 1990) indicate a similar percentage non-believers (49%, resp. 55%), but a much lower percentage of people identifying themselves as humanists (3%, resp. 1%). This difference can be explained by the way in which the question was asked. In the afore mentioned telephone surveys published in 1990, the question was: "What is your present life stance or religion?", without mentioning possible answers (Catholic, Protestant, Humanist, etc.). When these categories are presented to the interviewees, we see more differentiation in the answers of non-religious people. This indicates that traditional identifiers (like Catholic or married) are mentioned much more spontaneously than non-traditional ones (like Humanist or homo/bisexual). To be able to check and improve the representativity of telephone surveys, it is essential to include in the questionnaire some traditonal items (like sex, age) and some non-traditional, potentially controversial items (like life stance, political and sexual orientation), because this enables the researchers to situate the sample in its socio-cultural context (including the willingness to be frank on controversial issues). If specific categories in the general population tend to be underrepresented in telephone surveys (like the poor, students, homo- and bisexuals, and people who are not at home regularly), extra attention should be given to reaching them in sufficient ways to correct the underrepresentation statistically.

9The third reason for telephone surveys (quick feedback on campaign effects) is also problematic for stigmatised target groups like homosexuals. Due to the fact that 100% anonimity in the perception of the interviewee is required (the first reason), it is virtually impossible to have a cohort study to measure campaign effects on the same people who were interviewed earlier. The non-response rate in cohort studies of the general population is in general very high but in AIDS research on homosexual men it is very low (Van Griensven, 1989). However, these cohort studies offer participants advantages like regular free medical check-ups and feedback on the results of the ongoing cohort study. As a matter of fact, the effects of prevention campaigns on the general public can also be measured by a series of representative telephone surveys. They can be very helpful for AIDS prevention policy makers (De Vroome, 1990), but for more detailed information on the mechanisms of prevention and behavioural effects more qualitative research is needed. Nevertheless, telephone surveys can be used as an additional instrument in ongoing cohort studies, but only if the participants have been asked in advance about their willingness to be approached by phone. Given the results of the media cohort study mentioned above, it is probably not a very reliable instrument for sensitive questions.

10The fourth reason (low costs) is also partly true. The least expensive way of doing large-scale research among homosexual men has been the use of the readership of gay magazines like De GAY Krant in the Netherlands and Gay Pied in France. Due to the fact that this is the nearest you can get to representative research among homosexual men, this has been an important instrument for psycho-social AIDS research on this target group (Tielman & Poller, 1988, 1989). Telephone surveys based upon the subscribers lists is in fact impossible because it will be considered by the target group as a serious violation of their privacy. For the general population, however, these considerations are of less importance (although there is growing sensitivity on the issue of privacy (Tielman & Hagendoorn, 1989). In countries where enough telephone connections are available, computerised sampling, questioning and data analysis techniques can be used, and telephone survey data can be completed and/or corrected by data compiled by other research methods related to the specific characteristics of the individuals and/or topics to be investigated, such surveys can be very reliable and cost efficient In summary, telephone survey methods may be described as necessary but not sufficient instruments for AIDS research. They are necessary to provide policy makers and AIDS prevention workers quickly and relatively cheaply with the needed information on knowledge, attitudes and reported behaviour in the general public and non-stigmatised sub-groups but they are not sufficient. Other research methods are needed to provide data to check representativity and to understand the mechanisms of social behaviour related to sensitive topics.



Dalles M. (1990), Onveilige seks bij homomannen, Interfacultaire Werkgroep Homostudies, Utrecht

De Vroome E.M.M., Sandfort T.G.M., Paalman M.E.M., Tielman R.A.P. (1988a), AIDS and condom use in the Netherlands, Poster presented at the International AIDS Conference, Stockholm.

De Vroome E.M.M., Sandfort T.G.M., Paalman M.E.M., Tielman R.A.P. (1988b), De ontwikkeling van kennis houding en gedrag inzake veilig vrijen en AIDS, Interfacultaire Werkgroep Homostudies, Utrecht.

De Vroome E.M.M., Paalman M.E.M., Sandfort T.G.M. (1990), AIDS in the Netherlands: the effects of several years of campaigning, in International Journal of STD & AIDS, 1 (in press).

Paalman M.E.M., Tielman R.A.P., De Vroome E.M.M. (1988), Condom Promotion in The Netherlands, Poster presented at the International AIDS Conference, Stockholm.

Tielman R.A.P. (1982), Homoseksualiteit in Nederland, studie van een emancipatiebeweging, Boom, Amsterdam.

Tielman R.A.P. (1987), Humanisme en sociologie, een paradox als paradigma, Socrates, Utrecht.

Tielman R.A.P. (1988), A worlwide inventory of the legal and social situation of lesbians and gay men, in Tielman R.A.P. (ed), Second ILGA Pink Book, Interfacultaire Werkgroep Homostudies, Utrecht 183-242.

Tielman R.A.P. (1989), Het Monster Trotteldrom, Beeldvorming op de Nederlandse beeldbuis, in Schedler P. (ed), Homoseksualiteit in beeld, Veen, Utrecht/Antwerpen, 152-167.

Tielman R.A.P. (1990), Evaluation of Dutch AIDS Prevention Policy, Utrecht (in press).

Tielman R.A.P., Hagendoorn H. (1989), STD/AIDS Prevention and Policy Making, in Promoting Safer Sex, STD-Foundation, Utrecht, 97-102.

Tielman R.A.P., Polter S. (1988), Evaluation of the Dutch AIDS Prevention Campaign among Homosexual Men, Poster presented at the International AIDS Conference, Stockholm.

Tielman R.A.P., Polter S. (1989), Safe and Unsafe Sex among Homosexual Men in the Netherlands, Poster presented at the International AIDS Conference, Montreal.

Tielman R.A.P., Van Griensven G.J.P. (1985), Sociaalwetenschappelijk AIDS onderzoek, in Sociologische Gids, 32: 416-430

Van Griensven G.J.P., De Vroome E.M.M., Tielman R.A.P. (1987), Seksuele gedragsverandering in een groep op HIV-antistoffen getoetste homoseksuele mannen, in Tijdschrift voor Sociale Gezondheidszorg, 65: 345-350.

Van Griensven G.J.P., De Vroome E.M.M., Tielman R.A.P. (1988), Veranderingen in anogenitaal seksueel contact en condoomgebruik afhankelijk van relatiepatroon en HIV-ab serodiagnose, 66: 157-160.

Van Griensven G.J.P., Tielman R.A.P. (1986), Risikofaktoren en prevalentie van LAVIHTLV III antistoffen bij homoseksuele mannen in Nederland, in Tijdschrift voor Sociale Gezondheidszorg, 64: 100-107.

Verhouden J.A.A.M., Sandfort T.G.M., De Vroome E.M.M., Tielman R.A.P. (1990), Veilig vrijen en condoomgebruik onder jongeren in Nederland, Utrecht.


1 The psycho-social AIDS research done by the Department of Gay and Lesbian Studies is funded by the "Praeventiefonds", the Dutch Department of Welfare and Public Health, the Global Programme on AIDS of the World Health Organization and the University of Utrecht


Department of Gay and Lesbian Studies, Rijksuniversiteit Utrecht, Utrecht.

© Presses de l’Université Saint-Louis, 1990

Conditions d’utilisation :