ABSTRACT
It struck me that, as a social psychologist, this controlled experimental approach-in which methodological precision is valued and all extraneous influences to the one under study are removed or controlled away-was fine provided I was confident that the phenomena I was investigating were important in the course of naturally-occurring human behavior. But neither theory nor the prior literature addresses the issues of the strength, prevalence, or prominence of predicted effects in everyday interaction. Rigorous experimental procedure certainly doesn't answer the question either. If anything, the precision of technique and measurement may mislead us by allowing for the discovery of replicable and statistically significant influences that are so trivial in their size and impact outside of the experimental setting as to never manifest themselves when other factors are allowed to vary. I realized that I was bringing the factors of theory, prior literature, and experimental precision to bear too early in the sequence of my investigation. These were things that would tell me, once I had something important to examine, what exactly it was that I had.
