What is the nature of a 'control' group?

Hey , does anyone know how to draw up a demographic or 'control group' perameters for socialogical experminentation?
Before the groans and/or worries, it is merely based on mores and specifically norms that we all know govern our progra... I mean actions. Anyway, anybody got their Phd in sociology? i would appreciate any help at all. Cheers.

Say I had a theory that sex with multiple women at once was more pleasurable then just one.

so in this case to get base line results, my control group would be the sex with the single woman, versus the experimental group with multiple women at once.
get it?

Other disciplines also use control groups (particularly for intervention/treatment research):

"Control groups are usually used to address threats to internal validity such as history, maturation, selection, testing, and others. Control of these and related threats is accomplished by ensuring that one group in the design shares these influences with the intervention group but does not receive the intervention or experimental manipulation. If the intervention and control groups are formed by random assignment and assessed at the same point(s) in time, internal validity threats are usually addressed. In clinical research, several control groups are often used" (Kazdin, 1998, p. 124).

Kazdin, A.E. (1998). Research design in clinical psychology (3rd ed.). Boston, MA: Allyn and Bacon.

How to construct a control group depends entirely on the comparison group. Ideally the 2 groups will be equal on all variables besides that which you are measuring.


[Edited on 11-9-2003 by MKULTRA]