Problembereich:
|
 |
Applying a screener questionnaire to a group of n people is in essence an experimental protocol involving n experiments conducted in parallel. Pyramiding is an experimental protocol involving x experiments conducted in series. In the case of experiments conducted in parallel, no learning is possible between experiments. In the case of experiments conducted in series, it is possible for an experimenter to incorporate learning acquired from previous experiments into each succeeding experiment in the series. Which of these basic types of experimental protocol is most appropriate depends upon what one is trying to do. In many common marketing research applications, protocols involving parallel experimentation are appropriate. Indeed, a common form of parallel experimentation in marketing research – screening questionnaires - have probably come to be standard operating procedure precisely because parallel experimentation is appropriate for many of the types of questions of interest to marketers. In the case of finding innovators or innovative ideas, it is our contention that a serial experimental protocol, with learning from each experiment in the series incorporated into the next, is more appropriate than a parallel experimental protocol. In this research project, we will compare both search strategies.
|
