THE EARLY SEX RESEARCHERS: SOME METHODOLOGICAL ISSUES - THE ACTUAL MEASUREMENT DEVICES USED IN SURVEYS

Another major problem concerns the actual measurement devices used in surveys. A researcher should always question whether the questionnaire is both reliable and valid. In discussing reliability, we are referring basically to the consistency or stability of a test. Let us suppose we ask 1,000 students at what age they began to masturbate. One week later we ask the same question of these same 1,000 individuals. If the questionnaire is reliable, the responses should be the same on both administrations. This is an example of what is termed test-retest reliability, a measure of stability over time. Other types of reliability are also important, such as measures of inter-item reliability, or the internal consistency of a questionnaire. These types of reliability reflect whether the questionnaire as a whole is reliable or if certain parts or questions are likely to be unstable.

Validity is another important concept related to the specific assessment techniques utilized. In very simple terms, a test is valid if it accurately measures what one really wants it to measure. A test cannot be valid if it is unreliable; on the other hand, a test can have good reliability and still not be valid. Many of the sexual surveys are inadequate in terms of both reliability and validity. Investigators usually make up their own questionnaires, a fact that yields almost as many questionnaires as there are studies. Many, if not the majority, of these questionnaires lack any published data as to their reliability and validity and thus must always be interpreted with great care.

*71\265\8*

Men's Health-Erectile Dysfunction