Aparametric investigation of multiletter matches

Abstract

Proctor (1981) presented a theoretical framework that distinguishes factors contributing to disparities in time to classify physical-same, name-same, and different letter pairs as a function of three variables: method of presentation (simultaneous vs. successive), case relationship (samecase vs. opposite-case pairs), and blocking (blocked vs. mixed presentation of same-case and opposite-case pairs). He also argued that these variables were critical in multiletter matches, and performed a between-study comparison of existing data to support his contention. Because comparison across studies is always a tenuous process and because the absence of several relevant conditions precluded a complete analysis of predictions, a within-experiment manipulation of the three relevant variables was desirable. The present study reports such an experiment. In general, the factorial manipulation of variables supported predictions of Proctor's framework and indicated that many phenomena of multiletter matching, including the widely studied fast-"same" phenomenon, are attributable primarily to differences in the rate at which component letter pairs are matched.Numerous studies during the past decade have examined same-different matches of multiletter strings. Such studies have been popular not only because they allow precise quantitative tests of the manner in which multielement stimuli are compared (e.g., Bamber, 1969; Taylor, 1976b), but also because they provide a baseline for examining the influence that various structural constraints (such as orthographic regularity, phonological regularity, and word structure) have on processing (e.g., Bruder, 1978;Bruder & Silverman, 1974).In the most common version of the multilettermatching task, two letter strings are presented that are of the same length, and subjects are required to respond "same" if all letters in corresponding positions are identical (e.g., MDT B; MDT B) and "different" otherwise. The general pattern of results obtained in all studies of this type is that the latencies of both "same" and "different" responses increase as a function of string length and that, within a given string length, the latency of "different" responses increases with the number of positions on which the two strings match.In the first study of multiletter matching, Bamber (1969) determined that the "different" reaction-time functions were accurately described by an analyticalWe would like to thank Peter Harzem for allowing us to conduct the experiment on his equipment, Deborah Grantham for assistance in conducting the experiment, Maria Ekstrand-Hurst and Otto Arnoscht for assistance in analyzing the data, and Janet D. Proctor for helpful comments on the manuscript. Requests for reprints should be sent to

An examination of response bias in multiletter matching

“…Despite this lack of a difference in error rate, responses to same pairs (M = 632 msec) were reliably faster overall than responses to different pairs (M=691 msec) [F(1,15)=20.5, p < .001]. EXPERIMENT 3: MINIMIZING ERRORS experiments, which found faster responses to same pairs with simultaneous presentation (Eichelman, 1970;Henderson & Henderson, 1975;Proctor & Hurst, 1982;Silverman, 1973;Silverman & Goldberg, 1975), most likely lies in the relatively high proportion of false-"same" errors made in previous studies.' The proportion of false-"same" errors increases greatly with string length, whereas the proportion of false-"different" errors remains relatively constant.…”

“…For simultaneously presented letter strings, the fast-same phenomenon that is usually obtained (e.g., Eichelman, 1970;Proctor & Hurst, 1982;Silverman, 1973) appears to be attributable entirely to response bias. With the bias manipulation used in Experiment I, correct "different" responses were faster overall than were correct "same" responses, even though false-"same" errors outnumbered false-"different" errors.…”

“…Because "same" responses are faster than "different" responses for single-letter pairs, the reaction-time functions are approximately parallel, with the function for same pairs lying below that for different pairs (Bamber, 1969;Proctor, 1981;Proctor & Hurst, 1982;Taylor, 1976).…”

“…Only four-letter strings were used as stimuli, with same pairs being identical at all positions and different pairs differing at only one position. Different pairs were restricted to this type because these one-different pairs provide the most appropriate baseline against which to evaluate responses to same pairs (Proctor & Hurst, 1982).…”

“…Having obtained evidence that response bias, alone, cannot account for the fast-same phenomenon obtained with singleletter pairs, the next step was to determine whether such is also the case for multiletter pairs. The potential role for response bias is much greater in multiletter matching than in single-letter matching, because false-"same" errors predominate (e.g., Proctor & Hurst, 1982;Silverman, 1973;Taylor, 1976). The relatively large number of errors of this type that occur for strings that differ at only one position has been interpreted as indicating that the search process is often prematurely terminated, and a "same" response made, before all comparisons are completed (Taylor, 1976).…”