Limitations in research (Sexology)

Biases in sex research (and research in general) come from many sources. The treatment success in an uncontrolled clinical trial may be due to the attention paid to subjects and the power of positive suggestion - scientifically known as the placebo effect - rather than the treatment itself. The attitudes and preconceptions of the researcher influence the study methods, the scientific questions asked, and the interpretation of data. Researchers may fail to "see" events that do not match their theoretical view of the problem or they may "see" nonexistent events that neatly fit their model of thought.
Individual characteristics of the researcher such as personality, sex, appearance, friendliness, age, may also cause distortions in the data. For example, a study of sexual function in young spinal cord-injured men used female nurses as interviewers. Therefore, giving rise to the possibility that subjects might have exaggerated their reports of sexual ability to impress the nurse or not to embarrass themselves. It is possible that different information might have been obtained if those conducting the interviews were male. Another example: subjects interviewed by black experimenters may give different answers to questions about racial prejudice than do subjects interviewed by white experimenters, and vice versa.
Questionnaires are useful from the viewpoints of economics and time efficiency. However, they have several limitations. They can be used only with subjects who can read and write; they must be constructed in such a way that respondents do not lose interest  or become fatigued; and there is no opportunity for in-depth examination of the meaning or nuance of answers. Survey studies in general - whether done by questionnaire or interview  - share another limitation: In many cases, the questions asked elicit opinions, attitudes, or perceptions of behaviour, but these do not necessarily reflect the actual behaviour under study. There is nothing wrong with studying attitudes or perceptions, but scientific accuracy is not served confusing fact and feeling.
Observational and experimental studies are usually more expensive and time consuming than surveys. In both of these methods, the setting of the study may influence the observed behaviour. Even after subjects become acclimated to a laboratory environment and the equipment used to monitor or measure their responses, it is unlikely that they will be as relaxed and spontaneous as if they were at home. All observational research in which subjects know that they are being observed exerts some influences on subsequent behaviour. The same is true for experimental  research; subjects may try to respond in ways they believe the experiment "wants" them to or in ways dictated by how they want the experimenter to perceive them. The change in people's behaviour caused by knowing they are in an experiment is called the Hawthorne effect.
Many reports use statistical tests of significance, with significance in this instance meaning a result not likely to have happened by chance. Most commonly, a probability level of 5 per cent or less ( p s .05) is chosen to define significance: this means that the probability of the research findings arising from chance alone is 1 in 20 or less. Although the formulation sounds impressive, results showing statistical significance are not infallible and absolute. More important than such statistics alone are the questions of whether the research design is appropriate, the types of bias minimized, and whether the findings have been independently replicated. When research findings are closely duplicated by a separate investigator using similar methods , it is much safer to conclude the findings are valid.
One cautionary note about replication is important however. If two independent researchers use the same methods and perpetuate the same biases, the conclusions they come to may be identical but may still be wrong. For example, in 1957 a study of gay people had been repeatedly studied from poorly selected samples  taken from prisoners and psychiatric hospitals. The researchers concluded that gay people were often maladjusted and sick. Later studies, without the same sampling bias and outside prisons and psychiatric hospitals came to very different conclusions about gay people.
A research perspective is only one way of looking at they world. It is not the only way and not necessarily the best way. Research is only an approximate way of getting at facts. Studying sex research can be informative but also is knowledge gained from clinical situations, personal experiences, literature, art, and culture. no single perspective on human sexuality has a monopoly on truth.