Qualitative research models in representative longitudinal studies
The call for "qualitative" research is always heard when the standardised methods which primarily measure quantitative relationships are unable to explain a change in the market or to explain it adequately. In this situation, attention is focused on 5 group discussions or 20-30 in-depth interviews--the findings from which are usually of interest, giving the imagination a greater boost than do the mere figures from broad representative samples. But the shortcoming inherent in this procedure should not be forgotten; the findings do not have a secure statistical base and may thus essentially result from the in part highly artificial conditions of the experiment. While these research techniques are extremely valuable for purposes of assembling hypotheses, they should not determine marketing decisions. Decisions should not be made without a statistically secure base; hypotheses must be tested using sufficiently broad samples and suitable investigative techniques which can be replicated at will by researchers. The Institut fur Demoskopie Allensbach has been working in this area since the sixties with representatively selected personally-interviewed panels, using samples of at least 1,000. The focus is on "qualitative" research, which seeks not only to register but also to explain changes in markets and in society.
- This could also be of interest