Computer Simulation: Some Remarks on Theory in Psychology1
Geoffrey R. Loftus University of Washington
Almost a decade ago, Russell Church delivered the Presidential Address to the Society for Computers in Psychology ( Church, 1983). In this address, Church presented an excellent summary of the usefulness and pervasiveness of computers in all phases of psychological research. He started with the literature search, proceeded through the phases of experimental control, recording of results, storage of data, analysis of results, development of theory, comparison of theory with data, preparation of figures, and ended with the processing of the manuscript. For each of these phases, Church compared the tedium of the precomputer technique with the ease and efficiency of the corresponding postcomputer technique. Presented in this way, the enormous facilitative impact of the computer on our research endeavors was breathtaking: Indeed, as Church's arguments unfolded, it became difficult to imagine how scientific research ever got done in those bygone days before there were computers.
Today, as the computer revolution accelerates, this general view has become yet more ingrained, and it is virtually impossible to envision scientific research carried out in any sort of computerless environment. My own laboratory, for example, is computerized to a degree that far exceeds Church's fantasies ( Stod dard & Loftus, 1988).
Does this mean, however, that more computerization is always better in all phases of research in the social sciences? I argue in this chapter that the answer is, "not necessarily." In particular, there is a danger that letting the computer do our thinking--as opposed to our busy work--for us can potentially deprive the scientific enterprise of the creative and insightful thinking that is the sine qua non of scientific progress.
Consider, as an example, the use of the computer in statistical analysis of____________________