There are several ethical considerations involving a study of this nature. The study obviously has to be a confidential one. The information on the participant will remain always need to remain private. Informed consent needs to be gained and the participants must feel no pressure to answer questions. However they must understand the importance of the study and not answer the questions incorrectly, or feel they have to lie about certain questions as they are in doubt as to whether their information does not remain private, and so their trust must be gained.
The design of the experiment must reflect the type of data needed to be collected because it is a sensitive subject but a great deal of information is sought. A face to face style interview allows a report to be built between the interviewer and participant, and allows the questions to be put in a certain manner, and any confusing issues to be clarified. It also allows the participant to ask questions if there is something that they are unsure of or do understand fully what is being asked of them.
There may also however be biases as the requires may lead the participant to answer a question purposely wrong, as they feel this is the answer the interviewer requires, and as there is no anonymity when answering the questions the participant may not feel inclined to answer truthfully. The other possibility would be to conduct a self-completion personal interview. This would allow for more privacy when answering the personal questions, and relieve some of the discomfort that might be felt when an interviewer had to ask the participant very personal questions. In the national survey, the participants were selected by the small user address postcode file, in which small residential properties are listed, and which is constantly being updated.
There was a sub-strata systematic selection in which the interviewer selects one person at random from a household. This means that there is an uneven chance of selection due to differing household sizes and so the data has to be re-weighted. This is because the sample was from a registrar of addresses, and there were differing number of occupants in each. It is obvious that a lone occupant has more chance of being selected than a multi-populated residence.
The location of the respondent is also important because an even response is wanted across all geographical areas. By re-weighing the data, all the results are a product of the regional weight and the household size scaled to the sample size which in turn shows a more balanced picture. It was decided that they would use a combination of face-to-face interviewing with a self-completion section. Questions of a more personal and sensitive nature were asked in a booklet which was sealed by the respondent after they had answered the questions, and it was only identifiable be a number.
Some difficult issues in measuring data about sexual attitudes is that people have to understand what point you are trying to make, but not feel that they are being patronized. Every participant will have had different experiences and have differing levels of knowledge. The questions have to be expressed in a manner that the point of it is clear, but it does not read so basic that it is not taken seriously, or that people feel they are being talked down to.
By putting the entire sensitive questions in a booklet, such as in the national survey, and by putting a glossary at the back people are able to look up any parts they do not fully understand and avoid any unnecessary embarrassment. Also, by allowing the participants the opportunity to answer using multiple choice such as by for example answering a, b, c or d, they do not feel that they are as pressured to answer, and as the answers are all ready laid out they don’t have to worry that their answer is a socially unacceptable one and which they feel they can not write out. By seeing their answer written down, in gives an increasing confidence to answer truthfully.
Reliability and validity are very important when wanting to find out information and present the findings to the public. When examining sexual behaviours there are not many opportunities to check the data provided is correct because sexual behaviours are mostly private affairs. No other sources can be used, and it is not likely that anything can be verified with disrespecting the respondents’ privacy, so it is important that the survey appeals or the respondent to be honest with their answers.
They respondent needs to be reassured constantly that their data is going to remain private and that the data collected is very importance, and what the implications of the study might be in helping the AIDS cause and improving the public’s general health. In search for validity it is however possible to see how consistently one participant answers all their questions. By checking that their answers were constantly along the same line, it suggests they are answering truthfully. In the Wellings study inconsistency checks were carried out on all the questionnaires and it was found that around 80% had no inconsistencies at all (Wadsworth, Field, Johnson, Bradshaw, Wellings 1993).
Surveys that break barriers and report information that has not been collected before because of there controversy, are often hard to perform due to all the difficulties it entails. The public are often not willing to participate if they feel their private life is being intruded and the questions are too personal. There may be problems finding funding and backing for surveys from the companies that the data concerns if they believe the survey will not be successful and they won’t want their name attached to failed work. Such surveys though, if researched and carried out successfully could be highly publicized and fulfill most of the agenda set at the start with ease.
Cochran, W.G. (1977), Sampling Techniques (3rd edition). Wiley.
Cochran, W. G., F. Mostelier, and J. W. Tukey (1953) “Statistical Problems of the Kinsey Report”, Journal of American Statistical Association, 48, 673-716
Gagnon, J. H. (1988). “Sex research and Sexual Conduct in the Era of AIDS”, Journal of AIDS, 1:593-601
Kinsey, A.C., W. B. Pomeroy, and C.E Martin (1948). Sexual Behavior in the Human Male. Philidelphia: W.B. Saunders
Kinsey, A.C., W. B. Pomeroy, C.E Martin, and P.H Gebhard (1953). Sexual Behavior in the Human Female. Philidelphia: W.B. Saunders