TIME FOR READING AND MEMORY

This paper on silent reading is concerned with the rate of comprehension, selection while reading, and the rigidity of pacing while reading. Subjects studied 144 short statements under one of three degrees of pacing: ( a ) a time was set for studying the whole series of statements (whole condition), ( b ) each complete statement was presented in turn by itself for a set time (statement condition), and ( c ) part only of a single statement was presented at a time, and there was a period after each statement when no print was visible (segment condition). In different subconditions 1.0 min. was allowed for an average of every 293, 146, 73 or 37 words. In one variation of condition ( a ) the subject was instructed to pay attention only to the 18% of statements connected with geography (selection condition). In another variation, the time for study was fixed at 3.0 min., and the number of statements was reduced accordingly (shortened condition). After reading, the subject had first to estimate from memory the percentage of statements which could be classified under certain specified topics. He was then given a test of memory for seventy‐two of the statements. Half of them had to be completed (recall), and each of the remaining half had to be discriminated from a paired statement which resembled it in one of six standard ways (recognition). An identical test of memory for the remaining seventy‐two statements was normally given unexpectedly 1 week later. A control group of subjects took these tests without having read the statements. In the whole condition when 1.0 min. was allowed for every 293 or 146 words, most of the subjects reported that they had read practically all the statements once and only once. Reading at about 146 words per minute in this condition Was found to give significantly better comprehension of the individual statements than reading at about 293 words per minute. In contrast, in this whole condition when 1.0 min. was allowed for every 73 or 37 words, the subjects reported that they had re‐read some or all of the statements at least once. Between the rates of presentation of 37 and 293 words per minute, substance memory varied approximately as the logarithm of the time available per statement. After the 3.0 min. of study in the shortened condition, about the same number of statements could be remembered in substance or recognized whether 293 or only 73 words had had to be studied per minute. Selection tended to improve memory for the selected statements, and to reduce memory for the remainder. But when 293 words had to be studied per minute, only recognition soon after study benefited significantly from selection. When 1.0 min. was available for every 146 words, the rigid pacing of the segment condition gave significantly better memory soon after study than the loose pacing of the whole condition. But when 1.0 min. was available for every 37 words, pacing had no effect upon memory. There were no significant differences between the degrees of pacing used in the statement and whole conditions.