On the use of "theory" and the usefulness of theory.

The three preceding articles (see PA, Vol 76:10520, 10513, and 10523) question Greenwald, Pratkanis, Leippe, and Baumgardner's (see record 1986-20964-001) advocacy of result-centered research methods. This response notes points of agreement with the three comments and indicates how some disagreements can be traced to differences in use of the term "theory" and consequent differences in conceiving the relation of theory to data. A more substantial disagreement concerns Greenwald et al.'s point that result-centered methods may often advance theory more effectively than do theory-centered methods. That point is developed further here by (a) summarizing arguments for concluding that theory testing is not guaranteed to lead to increased precision or scope of theories, and (b) explaining how result-centered methods can be efficient in selecting among competing theories. (PsycINFO Database Record (c) 2010 APA, all rights reserved)