Giami, A., Olomucki, H., de Poplavsky, J.

Surveying Sexuality and AIDS :
Interviewers Attitudes and Representations

Originally published in: John Bancroft (eds.). Researching Sexual Behavior.
Bloomington, Indiana University Press, 1997, pp. 61-77.
Reproduced here by permission of the authors.

Contents

State of the Art

Surveying sexuality

Issues in interviewer training

Problematics

Methods

Context: The apparatus of the ACSF survey

Training interviewers

Training program

Content of the Training

Interviewer interviews

Interviewer motivations

Declared interest

Personal interest

The interviewer-respondent relationship

The "hook"

Handling conflict while maintaining neutrality

A double bind situation

Experiencing the interview process

Concerns and embarrassment

Ambivalence and dynamics of the motivation

Fantasies

The Survey-as-Coitus

The Interview-as-coitus

Interviewer representations

Sexuality and age

Difference between genders and communication about sexuality

The sexuality of the French

The evolution of the representations

Projection of the difficulties of the interviewers onto the respondents

Conclusion

References

 

 

            Surveying sexuality is not a simple act. Asking questions about a subject considered to be private can lead to unexpected or irrational reactions from the respondent, public opinion and politicians.  From the beginning, the ACSF[1] team was aware of the difficulties associated with this  work and rapidly recognized the necessity of training, supporting and listening to the problems of interviewers recruited to administer questionnaires used for the ACSF survey.  Completing a questionnaire on sexuality is conceived as a particular way of communicating about sexuality, in a particular cultural context.  This context is determined by psycho-social processes, which is it necessary to understand, in addition to the collection of data.

            This chapter describes and analyses the work of the interviewers who administered the questionnaires and collected data for the ACSF survey (Spira, Bajos and the ACSF group, 1994; Giami, 1996).  The survey is presented, as is the context for the questionnaires.  We then analyze the psychological processes and particularly the counter-transference attitudes (Devereux, 1967) affecting the interviewers during this work.  In particular, we address the interviewers' initial motivations and attitudes to the study, the relationship between the interviewer and respondent, and the representations and fantasies of the investigators concerning the theme of the survey:  sexuality and AIDS.

 

State of the Art

            Classic methodological work on questionnaire-based surveys often considers the interviewer to be one of the major sources of measurement error and bias during data collection, independent of the subject of the survey (Hyman, 1954; Turner and Martin, 1980).

 

Surveying sexuality

            In a text of nearly 40 pages, Alfred Kinsey (1948) defined in detail the ideal personality for an interviewer.  He emphasized the value of tolerance and empathy on the one hand and persuasion and seduction on the other.  He described the central contradiction between "standardizing the questions and adapting the form of the questions" (Kinsey, 1948).  According to Kinsey, the interviewer should be informed of the aim of the question and the field it is to cover, but almost complete freedom should be given in the formulation of the questions while collecting the data [2].  Note that is was only possible in this study because of the very small number of interviewers used to collect data.  The 18,000 people investigated for the Kinsey reports were interviewed by only  9 investigators, all experienced researchers engaged in the project.  Just two researchers, Kinsey and one of his assistants (W. Pomeroy) were responsible for 15,000 of the interviews (85%)! (Pomeroy, 1982)

            More recently, Johnson and Delamater suggested that the "major problem is the attitude of researchers and interviewers to research into sexuality".  As concerns methodology, they state that "concern with the threat of the subject matter is a projection of the interviewer's own discomfort".  They conclude by suspecting having been " insensitive to the important source of sensitivity in research on sexuality -ourselves and our interviewers- and have overestimated the extent to which our respondents are sensitive to these topics and to reporting them." (Johnson and Delmater, 1976, p. 181)[3].

            During the ACSF pilot survey, the formulation of the questionnaire was subject to a linguistic analysis and the interaction of the questionnaire was analyzed from recorded interviews.  This confirmed that the interviewer necessarily put his or her own accent on what he or she said (Richard-Zappela, 1994)and that the relationship between the interviewer and the survey developed as a sort of "conversation" between them (Achard, 1994).  This analysis of the problem of collecting data clearly evidenced the need to train interviewers prior to the ACSF survey.

            On the other hand, some researchers have chosen to ignore this phenomenon. They consequently propose methods to reduce this type of bias.  Sundet et al. (1990) for example opted for an entirely self-administered questionnaire, sent by post to the respondents.  This approach simply eliminates interviewers from the data supply step and leaves the respondent alone with the questionnaire.

Issues in interviewer training

            Bradburn and Sudman (1981) suggest that interviewer training should be central to the preparation of a survey.  They recommend "For surveys involving threatening topics, it would be a good idea to obtain a pretraining measure of the interviewers' expectations about difficulties with threatening questions. Then either those interviewers who expect considerable difficulties should not be assigned to that study, or time should be spent in training sessions to change their expectations and teach them how to handle problems if they should arise." (Bradburn and Sudman, 1981, p. 173).

            Training for interviewers which includes increasing awareness of the relational and psychological dimensions of collecting data has been adopted by most major modern surveys on sexual behavior (Simon, 1972;  Klassen et al., 1989; Laumann, Gagnon, Michael and Michaels, 1994).  In 1970,  the Kinsey Institute during routine preliminary testing of NORC[4] interviewers identified anxiety, embarrassment and a degree of rigidity associated with the research theme, which could potentially result in a large bias during data collection.  These attitudes led to problems when formulating questions, hasty interpretations, induction of replies and an absence of neutrality, all of which reduce the quality of the data collected.  The training program used in this case included, in addition to the technical training, identification and evaluation of potential psychological difficulties the interviewers may have.  The organizers of this survey therefore ran "psychological preparation" sessions during which the interviewers could discuss their difficulties (Shipps, 1989).

            Simon believed it important to "defuse affective and moral resistance, and misplaced curiosity which could result from the subject of the survey, taboo for everybody, and which could have a negative effect on the information to be collected" (Simon, 1972).

            During the preparation of the National Health and Social Life Survey, Laumann, Gagnon and their team obtained finance for a 3 day training session in Chicago for the 220 professional interviewers (mostly women) who had volunteered.  This research team noted that the expected difficulty of this study was seen as a stimulating challenge by many of the interviewers.  Most had volunteered for the survey because it was in the public interest.  The training sessions, animated by psychologists specialized in the field of sexuality, allowed the interviewers to explore their own anxieties and reticence.  Furthermore, participation in these sessions seems to have communicated strong motivation, which was maintained throughout the field phase.  This  research team chose women interviewers who were the most competent, without trying to match sex or race with the interviewees (Laumann, Gagnon, Michael and Michaels, 1994).

Problematics

 

            The use of interviewers is not only a source of bias in the production of data.  The establishment of a interviewer-respondent relationship is also one of the main conditions applying to a survey.  We have tried to identify the conditions affecting the progress of the survey, and particularly the scientific and technical apparatus which was developed and serves as the social and cultural context.  We also assessed the psychological processes and especially the counter-transference attitudes affecting the interviewers.  The significance of these processes was confirmed by the daily monitoring of the interviewers.  On one occasion, during the administration of a questionnaire, the following question was posed by the interviewer: "Have you yourself ever been subjected to pornographic conversations or telephone calls?".  The respondent replied "No, today is the first time".  This shows that meanings can be attributed to the survey and this can help understand the origin of much reticence, both public and private, concerning surveys of sexuality.

            We used as a basis the assumption of G. Devereux  that " An interview about sex even in the case of a scientific interview is in itself a kind of sexual interaction, which can be lived out on a symbolic, verbal and emotional level as shown in the analysis of the sexual transference in psycho-analysis."  (Devereux, 1967, p. 29).  Counter-transference can be defined as "all unconscious reactions of the analyst to the person analyzed and particularly transference from the analyst" (Laplanche and Pontalis, 1967).  This definition has the merit of emphasizing the importance of fantasies developed during this type of communication.

            Completing the questionnaire involves a relationship between two people.  The situation is however asymmetric.  For the interviewer, it is a professional activity, whereas it addresses the private life of the respondent.  For telephone interviews, the interviewer is in his/her place of work, covered by his/her professional status.  The respondents are questioned in their home in a private situation about the most intimate aspects of their life, and not only sexuality.  The questionnaire itself serves as mediator of the communication.  The interview proceeds according to a well defined scenario based on progression and the alternance of questions considered to be banal and those considered sensitive.

            The professional nature of the involvement of the interviewer does not exclude conscious and unconscious manifestations of his/her subjectivity.  We believe it necessary to study the interviewers' representations of sexuality and AIDS to identify both the difficulties with which they expect to be confronted, and spontaneous interpretation of replies which could affect the quality of the data collected.  There is also interference between the interviewer's subjectivity, derived from representations of sexuality and AIDS, and the professional dimension of the interviewer's activity aiming at isolating them.

Methods

 

            The apparatus established for the training of the interviewers gave us the opportunity to perform a qualitative survey among these interviewers to identify the psycho-social and some of the unconscious dimensions of  their  involvement in the survey.  The analyses presented in this chapter are based on all our formal and informal contacts with the interviewers, researchers and the heads of polling institutes throughout the ACSF survey.

            We asked the interviewers to comment and give their subjective impressions both before and after the survey.  This allowed us to investigate the interviewer-respondent interaction from the angle of its interiorization by the interviewers, rather than from its exteriority.  Our position is thus consistent with that of M. Godelier who states: "Social relations are not only between individuals, they are within individuals (...) Ideology is the inner framework of social relations and this framework is found as much within the individual as in his or her relationships" (Godelier, 1996, p. 28). Our qualitative study of interviewers was conducted through semistructured interviews to evaluate the importance of countertransference attitudes involved in carrying out the survey. About 30  interviewers were interviewed at different moments of the survey. 

Context: The apparatus of the ACSF survey

 

            In our approach, the apparatus of the survey can be considered to be a working situation  characterized by a particular working organization  (Dejours, 1993).  This serves as a framework within which the various participants in the survey act.  The research apparatus consists of all the relationships between the various people involved in the project (the researchers, polling institutes, interviewers) and also the external bodies controlling the research.  This is the context for the relationship between the interviewer and respondent.

            The interviewers are a sort of open window onto the studied population, in this case representative of the French general population.  The interviewer-respondent relationship is mediated by a technical (the telephone) and scientific supports (the questionnaire).  The interviewers are directly supported by the polling institute supervisors and research teams of the ACSF who contributed to the training, the monitoring and control of the survey.  This monitoring is not limited to verifying the quality of the data collected, but allows effective supervision of the interviewers throughout the survey.  Thus, ACSF researchers are available both at moments of interest and excitement and periods of discouragement.  There are also informal contacts between interviewers and obviously they had relationships with their personal entourage which included discussions of the survey.  The discussions between the interviewers and the researchers gave the researchers a clearer and concrete view of the population surveyed.

            The researchers, supervisors and polling institute heads also have programmed meetings.  Within the research teams, the coordinators of the ACSF survey hold a particular position with respect to the researchers, the interviewers and the polling institute heads and external bodies.  The ACSF survey was evaluated by scientists of the ANRS, was assessed by the national AIDS committee as concerns ethics, and was under the administrative control of the CNIL (the national commission for computer technology and civil liberties) to ensure the confidentiality of the data collected.

            INSERM occupies a special place in this structure.  The unit organizing the survey is part of INSERM, and it is the client of the polling institutes and manages the public fundings allocated to research.  The name INSERM is a symbolic guarantee of the scientific quality of the work.  The project would not have been possible except at the request and with the authorization of the government (Got, 1989).  Thus throughout the survey, the interviewers worked in a context particularly loaded with the symbolic dimension of Science, the issues concerning the "fight against AIDS" and with Public involvement. 

Training interviewers

 

            Defining the objectives of the training and establishing training strategies were subjects of extensive discussion among the ACSF staff and managers of the polling institutes.  First, it appeared that providing training for the interviewers would mean a change in the usual work practices of these institutes who are usually responsible for it themselves.  In accordance with their organization culture, the project managers and supervisors wanted to be completely involved in the training process and be technically involved in its management.  As two institutes had been engaged to carry out the survey, the managers of each were reluctant to share a common training program which might reveal their expertise to the other.  The principle of technical training was immediately and unanimously accepted by the ACSF staff, given the complexity of the questionnaire (filters,  precise wording of instructions, selection of questions depending on the respondents' characteristics, etc.).  The introduction of a psychosocial and interpersonal dimension, on the other hand, was subject to more reservations.  The task appeared even more difficult as the team psychologists could not simply define typical ideal conduct and encourage the interviewers to adopt it.  Rather, the interviewers had to be helped to become aware of their own emotional investment in the subject of the survey and to try to avoid defensive attitudes when administering the questionnaire. 

 

Training program

            An initial training program was tested and evaluated at the same time as the pilot survey. The final program had the following objectives: 

           

            Direct Objectives

            1) Acquiring technical competence in the use of the questionnaire: familiarity with the introductory text, the filters, the forms of answers, etc;

            2) Acquiring the interpersonal skills for supporting the respondent in an exercise in which interviewers might be emotionally involved. This skill  involves the personalization of each interviewer-respondent interaction.  The interviewer must find, for each respondent, the right distance between  them.  Contrary to what might be thought, this is not contradictory with the requirement for standardizing the administration of  the questionnaire;

            3) Acquiring specific competence: understanding the survey objectives, knowledge of the structure and logic of the questionnaire.

 

            Indirect Objectives

            1) Clarifying the interviewer's psychological involvement in the research, in  particular interviewer resistance and counter-attitudes to the subject matter.

            2) Motivating the interviewers, giving them a sense of responsibility as they  are participating in a large-scale survey;

            3) Providing the necessary information about AIDS so as to better  understand the questionnaire and answer questions asked by respondents.  

 

Content of the Training

            Plenary Session:  Introducing the study, its scientific objectives and its usefulness to society; medical and epidemiological information on AIDS and STD (plenary  session); questions and answers.

             Workshops (small group): a detailed analysis of both the procedure for selecting the respondent and the questionnaire item by item; actual run-through of a questionnaire; analysis of this experience, involving  identification of interpersonal difficulties, role-playing and simulations; identifying technical and linguistic difficulties.

            Individual interview: with volunteer interviewers   

 

Interviewer interviews

            The interviews on which this study was based were primarily at the request of the ACSF psychologists.  However, they were used by the interviewers to express themselves, reconsider thoughts provoked by the training, clarify some of their motivations, reply themselves to some of the questions in the questionnaire, converse with a psychologist and state their opinion of the training.  Note that such interviews were offered to all the interviewers, and that a large proportion volunteered, but that only around 30 were actually interviewed due to available ressources.  Although not all the interviewers were interviewed, these interviews were perceived as an integral part of the support and monitoring of the interviewers as a body.  They revealed that the interviewers needed to communicate their experience of the survey.

 

"An interview is like an analysis, or a psychotherapy ... Well, I feel that its a great opportunity to discuss all this with a psychologist from the beginning to the end.  I really think, and I say this egotistically, only for myself, that is really good.  If everyone had been able to do it, it would have been even better" (Anne)

 

Interviewer motivations

Declared interest

            First, we picked up what might be defined as the interviewers' public and correct discourse associated with the necessity of convincing the research staff and employers of their competence and motivation.  We noted here the importance of participating in a large-scale humanitarian and scientific survey which would mean that human lives can be saved and science is advanced, while also informing the population and making it more aware of the problems of AIDS and the prevention of HIV infection.  Thus, there was substantial support for the objectives presented during the training sessions.  To a lesser degree, some interviewers were skeptical as to the effectiveness of this kind of survey in modifying sexual behavior and contributing to AIDS prevention.  According to this opinion, the survey would give a generalized and vague reflection of sexuality in France.    This dimension of the motivation seems to reflect the discourse expressed  by the staff  in the first session of the training program (presentation of the scope of the survey).  Some interviewers expressed here common views about the survey and research against AIDS, views that might be considered as correct by the staff.

Personal interest

            In this category, motivation refers much more to the personal experience of the  interviewers.  This section of the analysis identified motivations less consistent with the demands of the research team.  A significant number of the  interviewers knew one or more HIV-positive individuals or persons living with AIDS.  Such proximity was a strong motivation: for them, they would be making a personal contribution to the fight against AIDS.  Participating in a large-scale survey gave them a certain pride associated with the desire to be up to the task entrusted to them.  Being able to get respondents to answer honestly seemed to be a challenge which they wanted to take up.  There was also a certain amount of curiosity and desire to learn about the population's sexual practices, with a hope to make commonplace a subject which may be considered as taboo.  The link between the survey subject (sexual behavior) and the prevention of  AIDS was an element enabling them to sublimate positively their resistance to and guilt at exploring respondents' sexuality.  Finally it is important to note that participating in the survey was a paid job. This was also a personal motivation for the  interviewers.  

The interviewer-respondent relationship

            The "hook"

 

            The first contact with the respondent is the most important for establishing the survey relationship.  One of the consequences of the decision to use the telephone is to isolate the voice from the global sensorial perception of the outside world.  The voice is then the vehicle of the communication between the two people involved.  The voice can, according to the moment, be seductive, harmonious, sympathetic or rejecting.  Emotion and affects are also transmitted by the voice.

            The voice is initially persuasive to "hook" the other party.  Once involved in the survey, the individual becomes the respondent, and the voice of the interviewer is as neutral as possible when asking the questions.  The survey relationship is established and the respondent can respond.

            The ideal goal is neutrality, and establishing this is facilitated by styles adopted by the interviewer: according to the case, this could be arguments appealing to thought and logic, altruism or seduction.

 

Handling conflict while maintaining neutrality

 

            The interviewers in some cases became involved in discussion with people expressing opinions which shocked the interviewer.  It is interesting that the interviewers expressed ideological values colored by liberalism and tolerance which were consistent with the social values expressed by the ACSF research team.  Some interviewers were shocked by the number of respondents who did not believe themselves concerned by the AIDS problem.  The necessity of continuing the survey and the assessment of its objectives helped the interviewers to adopt a neutral attitude, overcoming their reactions to the views of the respondents.

 

A double bind situation

 

            This mode of communication involving completing a long questionnaire on sexuality cannot be a simple game of question and answer.  From the start, the interviewers developed representations of the study, in which they attributed it meanings other than the simple collection of data.  The way the interviewer-respondent relationship was handled depended among other factors on the ambiguous nature of the instructions given to the interviewers by the research team.  These instructions can be seen as a "double bind"[5]: the respondents must be made to talk, but also prevented from talking such that they do not digress, and so only produce usable data.

     "I do not really react in terms of the instructions however.  But, well, that's me and like a few others who'll be the same. ...  But I think that we risk people hanging up if we are too directive with this sort of questionnaire.  Remember, we'll be asking them not only about health, sex and sexual practices, but also what they think, their attitudes to life, death, how they see the future, and these are big issues..." (Anatole, before the survey).

            Getting reliable answers depends on the ability of the interviewers to manage a complex relationship loaded with their own motivation, reticence, representations of the survey developed from the first days, and the demands, expectations and questions of the respondents.  During the survey, the interviewers were confronted with various demands from the respondents.  As only the answers to the survey questions are considered as data, the interviewers thus had to restrict to a minimum other confidences, digressions and confessions brought on by the situation, so as not to interfere with the correct completion of the questionnaire.  A certain distance was perceived as being necessary for answering intimate questions.  Intimacy risked creating a process of self censorship.

 

Experiencing the interview process

 

            The interviewers were particularly interested in the working conditions, due to the fact that they had to be available and listen to the respondents.  The necessary productivity (linked in one of the two polling institutes to their remuneration) often appeared to be an obstacle to establishing such a relationship.  The time constraints were sometimes perceived as possibly damaging to the climate of confidence, which was difficult to establish in some cases.  However, it must be borne in mind that the interviewers often felt some apprehension about managing difficult cases and the time limits imposed by the needs of the survey were undoubtedly reassuring.

 

Concerns and embarrassment

            The enthusiasm and altruism of the interviewers must be balanced against the legitimate concerns expressed before the beginning of the survey.  Making contact with respondents brought some of these into clear focus, particularly the possibility of meeting with refusal and the problems associated with asking such intimate questions.  Part of this apprehension was projected onto the respondent.  We noted the fear of shocking some people or upsetting them and comments on the crude nature of the questionnaire and its length.  Some situations, related to interviewers' personal experiences, provoked strong feelings of embarrassment, particularly questions dealing  with death, rape and, more generally, sexuality.  The interviewers were sometimes worried about the risk of drifting towards coarseness or obscenity.  All of them were concerned for the absolute respect of  respondents' and their own anonymity.

 

Ambivalence and dynamics of the motivation

            The interviewers' involvement seemed to us to be a complex mixture both of strong social and personal motivations and of an expressed ambivalence composed of enthusiasm and concern.  These motivations and the underlying energy changed during the survey.  Concern decreased significantly as the work progressed and its feasibility was  demonstrated daily.  We noted the strongest involvement during collective summing-up sessions and personal conversations at mid-survey time.  The interviewers then showed themselves to be particularly stimulated by this exchange.  On the other hand, at the end of the survey, weariness and fatigue had set in, due to the progressive difficulty of  finding people to question and the repetitive nature of the responses.

            Throughout the three months that survey took, there were continuing processes of maturation, investment then disinvestment, excitement, and boredom.  It may therefore be necessary to take the time taken by a survey into account, and ensure effective and efficient support for the interviewers so as to maintain their motivation and their competence through to the end.

Fantasies

            Administering the questionnaire led to a number of fantasies among the interviewers.  Some of these fantasies were associated with situations involving communication on sexuality: using the telephone for sexual abuse; religious confessions; obtaining confessions of sexual guilt; or administering psychotherapy.

"I didn't dare go as far as say confession, but, I won't go as far as saying we'll be like priests, or psychologists ..." (Alphonse).

 

            Moreover, some of the words used by interviewers imply that the  interviewer-respondent interaction might have been experienced (on the  level of fantasizing) as a sexual interaction incorporating both excitement and embarrassment. The fantasies can be deciphered from the words and metaphors used by the interviewers.

 

      The Survey-as-Coitus

            Desire: during the training sessions, the interviewers expressed a very strong desire at the same time as expressing their fears linked to the high stakes involved.

            Excitement: At the beginning of the survey there was a fair level of excitement represented by very rich cognitive functioning (construction of  spontaneous theories about the respondents).

            Fatigue and boredom: Towards the end of the survey, the interviewers spoke of fatigue and boredom with the repetitiveness of responses.

 

      The Interview-as-coitus

 

            Carried on a wave of excitement, administration of the questionnaire was in some cases fantasized as a sexual relationship between the interviewer and respondent.  These fantasies were not directly admitted by the interviewers.  They used metaphors for the sexual act.

"Well, some people were a bit embarrassed, when asked if they thought that the man was frustrated by not having an orgasm, or for the woman not being penetrated.  But then, once entered into, once started, once we'd really got into the subject, it went better, you know.  It's true that some were a bit prude at the beginning, but there were others that plunged straight into it, you know.  It just happened ...  Anyway, it takes two people, so I don't know why we use more, we emphasized the man more than the woman ... and because I'm a woman it just happened.  It's true that it was easier when they were alone."  (Amélie)

 

            These exerts from an interview show the use of sexual metaphors for describing the survey.  Administering the questionnaire was experienced unconsciously as an active sexual relationship.  The interviewers were thus required to act in a contradictory way, but this seemed to work for the survey.  The necessary restriction on confidences served as a reassuring censure for the interviewers (involving asking pre-established questions and not drifting into intimate curiosity which could awaken various fantasies).

            The survey relationship required that the interviewers use a rational strategy to establish and maintain contact throughout the questionnaire, by handling their own motivations, the demands of the respondents, their personal reactions, and constraints imposed by the scientific structure.  The issues they faced included their own fantasies brought on by the situation of the survey on sexuality, reinforced by what they were saying and hearing.

 

Interviewer representations

 

            During interviews prior to the survey the interviewers described their fears and expectations of the survey.  These expectations are the expression of the interviewers' initial representations.  They were linked both to the interviewer's previous professional experience, and also to their representations of sexuality.  Identifying these representations was useful for two reasons.  During the survey they helped elucidate some of the difficulties felt by the interviewers (and this was the subject of work involving role-playing during the training sessions).  They also helped identify some of the representations of sexuality in the general population.  Obviously, the interviewers express representations likely to be found among people of their age and social category.

 

Sexuality and age

 

            Respondents more than 20 years older than their interviewer (i.e. a one generation age different) are considered here to be in an older age group.  Individuals in this group were perceived as not being concerned by sexuality.  Thus, the interviewers felt that discussion of their past sex life risked upsetting these individuals.  Older persons were perceived as belonging to a different cultural world in which talking about sexuality to a younger person is taboo.

 

Difference between genders and communication about sexuality

 

            The gender difference was raised when the interviewer seemed to perceive possible eroticism in the conversation.  Male interviewers presumed that women would have more difficulty replying to a man, and the female interviewers believed that men would have problems confiding in them.  Both men and particularly women interviewers emphasized the seriousness of the questionnaire and its scientific context.  The involvement of eroticism in the situation appeared to the interviewers as a potential source of difficulty.

 

The sexuality of the French

 

            Most of interviewers had a representation of a "well-behaved and  conformist France":  "the French always have the same classical  practices";  "they do not have that much intercourse";  "traditional gender roles are still dominant".  However, a few atypical cases were reported as shocking by  the interviewers: "very large number of sex partners"; "sexual violence and sexual abuse"; "sexually active catholic priests".

            Interviewer surprise evidenced the existence of normative representations, which filtered the respondent replies.

 

The evolution of the representations

Projection of the difficulties of the interviewers onto the respondents

            The perception of difficulties with communication are frequently projected onto the other party rather than leading to recognition of one's own difficulties with this other party.  Interviews with a woman interviewer before and after the survey reveal this type of process.

            Prior to the survey:

     "Some of the questions, which are crude, like 'do you suck your partner's penis' ...   I couldn't see any other way of saying it ... 'fellatio' was suggested but some people won't understand, so we said it more simply, more, well, directly.  It's true that I thought we could have found different words.  I said to myself that it's crude and a bit vulgar.  It's true, 'suck', well, it's vulgar ... I think that I blushed there, but, me, I'm sure I would have blushed if I'd been asked that question like that." (Agathe)

            After the survey

"I had a North African man who spoke fluent French, but I remember that there were the questions about his sexual activities during his life, what was it... 'have you licked your partner's sex', he didn't understand 'licked'.  I had the impression he didn't understand ... finally, I was tired, so I said to him 'listen, you lick an ice cream, it's the same only it's your partner's sex', and he understood.  Well, perhaps it was the situation that he didn't understand..." (Agathe).

            During the first interview, Agathe expressed her embarrassment about a question concerning fellatio.  During the post-survey interview she described a difficulty with the same question, but which she attributed to the respondent.  Thus, the interviewer initially could not attribute the problem to anyone else, and was forced to recognize it in herself.  Subsequently, it is striking that she chose the same item to report one of the difficulties she had during the survey.

 

            Performing the survey allowed the interviewers to express some of their representations of sexuality, particularly the relationship between age and sexuality, relationships between the sexes, the place of sexuality in the couple, and the opposition between situations considered to be typical and atypical.  The production of objective data due to the questionnaire allowed them to see the difference between prior representations and the overall impression gained through this experience.  The interviewers seemed to conduct, in their own way, personal research based on the confrontation between initial hypotheses and research data.

 

Conclusion

            Our analysis shows that conducting a survey on sexuality, a particular form of communication about sexuality, is far from being a natural act[6] which can be reduced to the simple materialism of the results presented to the scientific community, and society in general through the media.

            The interviews collected allowed us to identify representations and fantasies, or metaphors used by the interviewers to interpret their experience.  We tried to understand how and why the interviewers, overall, gave a meaning to their involvement and subjectively implicated themselves in the ACSF.  We do not know of any analysis of the subjective position of interviewers, during a quantitative survey by questionnaire, that has used this approach (see Catania et al., 1995).  Our work is similar to that of G. Herdt and R. Stoller using an ethnographic approach to eroticism (Herdt and Stoller, 1990). 

              The training and support from the research team made it easier for the interviewers to handle the interviewer-repondent relationship in the heat of the action, the difficulties they imagined at the start of the survey, and those encountered during the survey.  Collecting data as complex as these depends on establishing a scientific, technical and relational structure developed for the occasion.  The structure restrained the subjective reactions (both psychological and ideological, conscious and unconscious) raised by the subject surveyed.  It also channeled eroticism involved in all forms of communication about sexuality.  The interviewers were thus is a double bind situation, or subject to ambivalence, which showed itself to facilitate formulation of the questions, while feeling protected by the scientific context.

            Using the telephone, thus hearing, seems to have favored expression of the imagination of the interviewers.  The technical language of the survey was not sufficient to contain the unconscious representations which should have used different words and metaphorical strategies, for expression. Note that a situation determined by focus on the voice and listening was represented using visual terms (implying a voyeurist position) or active terms (implying genital type fantasies).  The evolution of the psychological investment of the interviewers during the study revealed how a degree of eroticism associated with the situation, which could be seen as a risk of divergence, allowed the survey to develop, proceed and come to a conclusion.  Understanding this dimension could allow better management of secondary effects, particularly embarrassment and guilt, which could impair the progress of scientific work - the perfect subliminatory activity.

            The main aim in the development of the questionnaire was the collection of coherent,  homogeneous and quantifiable data.  This involved minimizing the personal expressions of the interviewers and respondents.  A pilot survey was used to compare face-to-face with telephone interviews (ACSF investigators, 1992).  A study of the interviewer effect evidenced the importance of the gender  on the data collected but only for some questions on opinion (Firdon and Laurent, in press).  The questionnaire and structure used can be considered to have minimized the effects of the interviewers on the production of data.  These effects were not however eliminated.

            It is surprising that despite the psychological factors at work on the interviewers, quantitative analysis of the interviewer effect particularly that associated with the interviewer's gender, did not evidence any large bias in the replies obtained.  It is possible that the production of reliable data from surveys may only be possible with the participation of trained interviewers who understand the difficulties associated with the relationship they establish with the respondent.  The data can only be standardized by appropriate use of flexibility and specificity for each interview.

           

            Our work helps illuminate the complexity of communication about sexuality in a professional context.  It mobilizes personal and subjective involvement, representations and fantasies in the professional interviewers. In addition to the information we collected about the act of surveying, our study opens new fields for work on communication about sexuality.  We obtained information about taboos and guilt arising from the curiosity involved, and about the satisfaction, possibly mixed with embarrassment associated with the discovery of details of the sexual life of the other.  We learn how erotic investment and disinvestment processes work in the individuals in this strictly defined context aimed at describing sexual behavior.  Desire finds indirect ways of manifesting itself other than the direct and crude expression of sexuality, which can be compared to certain forms of pornography.  The notions of confession, confidence, fault and surprise described by the interviewers show the extent to which communication about sexuality, whether during a survey or within a relationship between partners, is structured by imaginary significations attributed by the individuals.  The voyeuristic and sexual metaphors are evidence of the erotic character of the energy necessary for this work.

 

References :

 

Achard P. (1994), Sociologie du langage et analyse d'enquêtes. De l'hypothèse de la rationalité des réponses, Sociétés Contemporaines, 18/19, pp. 67-100.

ACSF Investigators, (1992), Analysis of sexual behavior in France : A comparison between two modes of investigation, telephone survey and face-to-face survey,  AIDS  6, pp. 315-323 

Adorno T.W., Frenkel-Brunswick E., Levinson D., Nevitt Sanford R. (1950) :  The Authoritarian Personality.  New-York, Harper and Brothers.

Bradburn N., Sudman S., (1974) :   Response effects in surveys   Chicago, Aldine Publishing

Bradburn N. , Sudman S., (1981) :  Improving Interview Method and Questionnaire Design .  San Francisco,  Jossey-Bass Publishers.

Catania J. A., Binson D., van der Straten A., Stone V., (1995) : Methodological Research on Sexual behavior in the AIDS  Era, Annual Review of Sex Research,  6, pp. 77-125

Dejours, C., (1993) : De la psychopathologie à la psychodynamique du travail, pp. 205-253 in Travail, Usure mentale   Paris, Bayard éditions.

Devereux G. (1967) : From Anxiety to Method in the Behavioral Sciences. The Hague, Paris, Mouton & Cie.

Firdion, J.-M., Laurent, R. (In press) :   Effet du sexe de l'enquêteur. Une enquête sur la sexualité et le sida.    

 

Giami A. (1996) : The influence of an epidemiological representation of sexuality: The ACSF survey questionnaire.  In M., Bozon, & H., Leridon, Eds. Sexuality and Social Sciences-An analysis of the French Sexual Behavior Survey. London: Dartmouth, in print, 1996.

Godelier, M. (1996). Sexualité et société.  In J. Cournut (Ed.), Psychanalyse et sexualité. Questions aux sciences humaines  (pp.27-40). Paris : Dunod.

Got Cl. (1989)  : Rapport sur le sida . Paris, Flammarion.

Herdt G., Stoller R., (1990) :   Intimate communications -  Erotics and the study of culture .  New York, Columbia University Press.

Hyman, H., (1954) :  Interviewing in Social Research  

Johnson W., Delamater J. (1976), "Response effects in sex surveys" Public Opinion Quaterly , 40, pp. 165-181. 

Kinsey A., Pomeroy W., Martin C. (1948) :  Sexual Behavior in the Human Male. Saunders, Philadelphia.  

Klassen A., Williams C., Levitt E. (1989) :  Sex and morality in the U.S. An empirical enquiry under the auspices of the Kinsey Institute. Middletown Conn. Wesleyan Univ. Press.

Laplanche J., Pontalis J.B. (1967) :  Vocabulaire de la psychanalyse  Paris, P.U.F.

Laumann E., Gagnon J., Michael R., Michaels S. (1994) :  The social organization of sexuality.  Sexual practices in the United States   Chicago, The University of Chicago Press.

Pomeroy W. (1982):   Dr. Kinsey and the Institute for Sex Research.  New Haven, Yale University Press.

Richard- Zappela  J.  (1994) Linguistic analysis of the questionnaire. In A. Spira, N. Bajos, & ACSF Group. Sexual behaviour and AIDS. Aldershot: Avebury, 1994, pp.  33-37.

Shipps J. (1989) : Training the NORC Field Supervisors, pp. 295-302 In  A. Klassen, C. Williams, E. Levitt  (eds.) :  Sex and morality in the U.S. An empirical enquiry under the auspices of the Kinsey Institute. Middletown Conn. Wesleyan Univ. Press.

Simon P., Gondonneau J., Mironer L., Dorlen-Rollier A.-M. (1972) : Rapport sur le comportement sexuel des français. Paris, Julliard, Charron.

Spira A., Bajos N., and the ACSF Group,  (1994) :   Sexual behaviour and AIDS. Aldershot: Avebury, 1994.

Sundet J., Magnus P., Kvalem I., Bakketeig L. (1990):  Self-Administered Anonymous Questionnaires in Sexual Behaviour Research : The Norwegian experience. In M. Hubert ed.  Sexual Behaviour and Risks of HIV Infection. Proceedings of an international workshop supported by the European Community. pp. 79-96. Bruxelles, Publications des Facultés Universitaires Saint-Louis.

Tielman, Rob, (1990) :  Telephone Surveys in Comparison with Other Methods in Psycho-Social AIDS Research pp. 97-103 In M. Hubert ed. Sexual Behaviour and Risks of HIV Infection - Proceedings of an international workshop supported by the European Communities  Bruxelles, Publications des Facultés Universitaires Saint-Louis.

Turner, C., Martin, E, (eds.) (1984) : Surveying Subjective Phenomena. Russell Sage Foundation, New York.

Wellings K., Fields J., Johnson A., Wadsworth J., Bradshaw S. (1990), Notes on the Design and Construction of a National Survey of Sexual Attitudes and Lifestyles, in : M. Hubert ed.  Sexual behaviour and risk of HIV Infection . Proceedings of an International workshop supported by the European Communities.  pp. 105-133.  Bruxelles, Publications des Facultés Universitaires  Publications des Facultés Universitaires Saint-Louis.