Norman R. Ellis

Short-term visual memory in normals and retardates

Abstract In two experiments, three groups of Ss differing in ability level were tested for retention of a six-item array by the partial report method after delays of 0, 1, 6, 10, 15, or 45 seconds. The 0-second delay produced a performance decrement, while delays of 1 and 6 seconds were optimal. Performance declined slightly, but significantly, only when delays were increased to at least 20 seconds. All groups improved with daily practice. No significant group differences were found, but between-group differences were highly consistent across delays and in both experiments. The 0-second performance decrement was discussed in terms of metacontrast effects. The data suggest that while the average number of items retained by Ss may differ with ability level, adequacy of retention is comparable for groups differing in IQ.

Absence of Noise Effects in the Simple Operant Behavior of Defectives

A review of the literature relative to the effects of noise upon performance of a task by human Ss indicates results are equivocal. While it might be expected thac distracting and/or noxious auditory stimulation would serve to interfere with task performance, Kryter (1950), after an analysis of rhe relevant variables, has concluded thac there is minimal evidence to support such a view. Berrien (1946), in a review of noise studies, has pointed out the need for more rigorous control in the investigation of noise effects. This investigator, however, feels that there is sufficient evidence to justify the conclusion that noise does affect the quality and quantity of task performance. In particular, Berrien, who used prolonged noise, suggests that noise may generate emotional instability which, in turn, is responsible for a decrement in output. This interpretation may also be relevant for noise which is not prolonged. Most of the studies conducted with noise as the independent variable have utilized normal Ss. Pascal (1953), however, studied the effects of noise upon simple reaction time in mental defectives. After obtaining a relatively stable measure, a loud noise was presented to S through earphones, and again, reaction time measured. The first effect was to cause a pronounced increase in reaction time, although wide individual variations were noted. In general, low MA Ss seemed most affected by the noise. The present study was designed to provide further information on the effects of noise upon the behavior of defective Ss. A simple lever-pulling task was employed. Since persistence at a task in the face of distracting and extraneous stimulation is generally considered to be an aspect of adaptive behavior with which retarded Ss have difficulty, it might be expected that noise would serve to create at least a temporary decrement in response rate.

Note on Habit Formation in Normal and Retarded Subjects

A previous report (Ellis, Pryer, Distefano, & Pryer, 1960) dealt with the relation between test intelligence and serial verbal learning. The purpose here is to show the dependence of habit formation (increasing frequency of R+) upon reinforcement (correct anticipation of an item) and to provide evidence bearing on Hulls postulate of individual differences ( 1952, p. 13 ) . Data were drawn from the study by Ellis, et al. as well as from a later one by Pryer (1960). The task was that of memorizing a list of 10 words presented in a memory drum. All words were familiar nouns, cat, boy, ball, car, man, doll, cup, book, one, bed, and were approximately equal in frequency of usage according to the Thorndike-Lorge list (1944). The words were presented to S at 2 sec. per item with 20-sec. intertrial rates. The method of anticipation was used. The retarded group consisted of 118 Ss (mean CA = 17.1, SD = 3.76; mean Binet IQ = 64.2, SD = 13.47). All Ss learned a first list; 75 of these learned a second list; and 60 of the 75 learned a third list. The normals were 75 high school students [mean CA = 16.1, SD = 1.33; Terman-McNemar (1949) mean IQ = 102.7, SD = 10.431. All 75 learned the first two lists and 60 of these, a third list. Approximately one month elapsed between learn, ing Lists 1 and 2 but not more than 2 hours between Lists 2 and 3. These data were pooled and subjected to an analysis similar to that described by Underwood (1954). Essentially this consists of determining the frequency of occurrence of R+ following successive reinforcements for each item and for each S. In the present analysis only the initial 12 correct anticipations were included. Fig. 1 presents the results. The initial empirical points are based on 1920 and 2410 observations for normal and retarded Ss, respectively. Number of observations is not equal to the pro-