Recognition memory for words presented at a slow or rapid rate

with those of Lev & Locascio (1969). who found that within levels of M, RT of stimulus terms. in forward anticipation learning. had no effect on PAL, but in a backward recalJ test, RT of response terms (stimulus terms in the forward anticipation learning task) had a significant effect (i.e., short-latency RT terms were recalled more frequently than long-latency terms). Again, M was a significant variable in both learning tasks. The Familiarization by M interaction provides additional evidence that within levels of M, the role of RT in PAL is different from that of M. Consistent with the predictions of the present study, learning the low-M response pairs was facilitated by familiarization trials (findings consistent with those of Schulz & Martin, 1964), whereas familiarization did not interact significantly with RT. The results of the present study are consistent with the hypothesis put forth by Ley & Locascio (I 969), viz, recognition or pronuncia bil i ty or Martin's (I 968) perceptual encoding are all indexed by M and are processes separate from recall. Further, since, within levels of M, the effect of RT is limited to the term of the PA that is required to be recalIed, it is conceivable that RT, although highly correlated with M, is the variable underlying recall, i.e., short-Iatency RT units may be recalled more readily than long-Iatency RT units. REFERENCES BATTIG, w. F., & SPERA, A. J. Rated association values of numbers from 0-100. Journal of Verbal Learning & Verbal Behavior, 1962, I, 200-202. GOSS, A. E. Comments on Professor Noble's paper. In C. N. Cofer and B. S. '.Iusgrave (Eds.), Verbal behavior and learning problems and processes. New York: McGraw-Hill, 1963. Pp. 119-155. GOSS, A. E., & NODlNE, C. F. Paired-associates learning: The role of meaningfulness. similarity, and familiarization. New York: Academic Press, 1965. LEY, R. Associative reaction time, meaningfulness, and presentation rate in paired-associate learning. Journal of Experimental Psychology, 1968, 78, 285-291. lEY, R., & ANDERSON, L. Associative reaction time of response terms in palred-associate learning. Journal of Experimental Psychology, 1969,79,378-380. lEY, R., & lOCASCIO, D. Associative reaction time and meaningfulness of stimulus terms in paired-associate learning and stimulus recall. Paper presented at the Psychonomic Society meeting, St. louis, Missouri, November 1969. lEY, R., & lOCASCIO, D. Associative re action time and meaningfulness of CVCVC response terms in paired-associate learning. J oumal of Experimental Psychology, 1970, 83,445-450. MARTIN, E. Stimulus meaningfuiness and paired-associate transfer: An encoding variability hypothesis. Psychological Review, 1968,75,421·441. SCHUlZ, R. W., & '.IARTIN, E. Aura! paired-associate learning: Stimulus familiarization, response familiarization, and pronounciability. Journal of Verba! learning & Verbal Behavior, 1964, 3, 139-145. TAYlOR, J. Meaningfulness of 320 words and paralogs. Unpublished doctoral dissertation, Duke University, 1959. UNDERWOOD, B. J" & SCHUll, R. W. Meaningfulness and verbal learning. Chicago: Lippencott, 1960. Recognition memory for words presented at a slow or rapid rate* ARTHUR 1. SCHULMAN and EUGENE A. LOVELACE University of Virginia, Charlottesville, Va. 22901 Fast presentation (about I word/sec) impairs subsequent recognition of both common and very rare words. This presentation rate, slow enough for accurate perception but probably too fast for much more elaborate information processing, also sharply reduces the variability of recognition memory scores. In studies of the ability to recognize words recently seen, investigators have overlooked the importance of the rate at which the words are originally presen ted. *The research described in this paper was first reported at the 1967 meeting of the Psychonomic Socicty. Psychon. Sci., 1970, Vol. 21 (2) They have chosen their rates more or less arbitrarily, so that a word's exposure duration alm ost always has fallen within the range of 2-5 sec. Over this range, later recognition performance seems to vary only slightly and unsystematically (Egan, 1958; Schulman, 1967; Shepard, 1967). It is as if normal processing of the words presented could be completed within 2 sec llr :;0, additional exposure time being supertluous. If sufficiently less than 2 sec we re allotted each word, however, performance in recognition memory would surely decline. Brief (but not tachistoscopic) exposures, long enough for accurate perception but too short for much more elaborate information processing, should impair recognition memory to the extent that this memory for recent experience depends upon such processing. Moreover, sufficiently brief exposures may be expected to curtail opportunities for idiosyncratic encoding and, consequently, to reduce variability among Ss. The present experiment shows that recognition memory indeed suffers and that variability is reduced when a suitably fast presentation rate is used. METHOD Two presentation sequences were prepared. One consisted entirely of common English words, the other entirely of very rare English words. A group of Ss was confronted with one of these sequences, at either a slow or a fast rate, and then was tested for recognition. Two Kodak Carousel slide projectors were connected in tandem and externally timed to produce the fast rate, which exposed each word for 1.00 sec, or the slow rate, which exposed each word for 4.25 sec. There was about 0.75 sec of change-time between slides. Two hundred common words were drawn selectively from the 1,000 most frequently occurring English words, as listed by Thorndike & Lorge (1944); 200 rare words, a majority of which had probably never before been seen by any of our Ss, were drawn se1ectively from Cieutat (1963). Within each group of 200 words, structural similarity was avoided, and only words of two or more syllables were allowed. 1 Of the 200 words in each group, 100 at random were selected for the presentation sequence. These "old" words were then interspersed among the 100 remaining "new" words on the test of recognition memory that followed. The Ss knew that the presentation sequence was to be followed by a recognition test. Only a few minutes for instructions separated the two phases of the experiment. The recognition test was in booklet form, and S worked at his own pace. Alongside each of the 200 test words, S recorded his confidence, using a 6-point scale, that the word was "old." The 2 x 6 data matrix so determined-Le., the six responses distributed over the event categories "old" and "new" -may be represented as pairs of increasing, accumulated proportions and plotted as points to which an operating characteristic (OC) may be fitted [see 99 Z. under both presentation rates. In arecent experiment, Schulman (1967) found comparable differences in recognizability, , """ -1 0.141 despite the fact that his "rare" words were ,., much commoner than ours, occurring once " ,-' (...truncated)