Judgments of Alphabetical Order and Mechanisms of Congruity Effects.

The congruity effect is a highly replicated feature of comparative judgments, and has been recently found in memory judgments of relative temporal order. Specifically, asking "Which came earlier?" versus "Which came later?" facilitates response times and sometimes error rates on judgments toward the beginning or end of the list, respectively. This suggests memory judgments of relative temporal order may be part of a broader class of comparative judgments. If so, the same congruity effect should also be found with the English alphabet, despite the alphabet being a longer, semantic-memory list, with forward directional encoding. A large-sample study (N = 340) produced a clear congruity effect in response time and even error rate (when controlled for response time). The large number of serial positions afforded by the alphabet enabled us to test a repertoire of mathematical models instantiating four distinct mechanisms of the congruity effect, against the empirical serial-position effects. The best-performing model assumed a response bias toward a discrete set of letters conceived of as "early" versus "late," respectively, an account that had previously been ruled out for typical comparative-judgment paradigms. In contrast, models implementing congruity effect mechanisms supported for conventional comparative judgment paradigms (based on reference-point theory or positional discriminability) produced quantitatively poorer fits, with more curvilinear serial-position effects that deviated from the data. The congruity effect thus extends to long, highly directional semantic-memory lists. However, qualitatively different serial-position effects across models suggest that, despite the superficial similarity, there are probably several quite different mechanisms that produce congruity effects, which may, in turn, depend on specific task characteristics. [ABSTRACT FROM AUTHOR]