Short Time Recall as a Function of Type of Stimulus and Length of Delay Interval

Abstract

We were interested in examining patterns of short-term information recall. We used the Brown-Peterson distractor technique to investigate the effects of stimuli type and delay interval on recall for 17 Ss. Each S was tested under 4 conditions, combined of word triads or nonsense syllables triads, with a short (10-sec) or long (45-Sec) delay interval. S read aloud the visually presented stimulus items, and aurally recalled them after the delay interval, in which S was engaged in counting backwards in threes from a presented 3-digit number. Measures were taken only for recall proportion. Results suggest a significant difference in recall between words and nonsense syllables, with words significantly better recalled. For the delay interval, results show no significant effect, and thus differ from those obtained in previous research. This discrepancy is discussed in terms of technical differences in procedure. No other effect or interaction approached significance.

Short-term Recall As a function of

Type of Stimulus and Length of Delay Interval

Short-term memory (STM) has been vastly studied and tested. One of the popular testing methods for STM has been the short-term recall, in which S is shortly presented with a single or multiple items, later to be recalled. Various researchers have focused their studies on different aspects, while attempting to identify those factors that most strongly influence short-term recall. A short-term memory model based upon a limited capacity to process information suggests that the recall of verbal material can be effected by the activity which consumes the retention interval, referred to as the interpolated activity.

Diverse studies examine the significance of elements involved in the interpolated task. In general, researchers have attempted to prevent Ss from rehearsing, i.e., processing exposed stimuli by keeping the material within the short-term store through the use of rehearsal-preventing interpolated tasks. The most famous of these has been suggested in the Brown-Peterson technique (Peterson & Peterson, 1959) in the form of backwards counting. In the original Peterson studies (1959), Ss attempted to retain aurally presented consonants trigrams while counting backwards by three or four from a three-digit number. This interpolated activity was continued for some predetermined retention interval, immediately after which Ss were asked to recall the originally presented consonants. Using this technique, the Petersons demonstrated a very rapid decline in recall. The probability of recall decreased exponentially with duration of interpolated activity.

Murdock (1960) investigated the effect of the stimulus type and the effect of varying the rate of interpolated activity. Similarly to the Petersons (1959) he found that in all cases forgetting, whether measured by accuracy or latency, increased with the duration of the interpolated activity. In addition, Murdock suggested little difference between the retention of three consonants and three words; consonant syllables and word triads were equally well recalled.

Murdock (Tell, 1971) has shown that short-term recall is effected by whether the presentation is visual or auditory. According to Neisser (Tell, 1971) there is an auditory-storage system which results from auditory input and can function as a source of information in short-term recall. This temporary storage system is referred to as echoic memory. "Echoic memory is passive, continuous, composed of sounds, and decays rapidly" (Tell, p. 150).

However, auditory input may also function as auditory noise. Voiced recall, vocalized irrelevant interpolated activity, or certain types of presentation conditions can mask, erase, or overwrite information available in the echoic memory store. This assumption is especially important when considering the interpolated activity. Tell (1971) suggested that a verbal interpolated task, in addition to its role as a rehearsal-prevention activity, would also produce auditory feedback, which masks information from echoic memory.

Tell and Ferguson (1974) further explored this issue of vocalization by examining the influences of active and passive vocalization on short-term recall. As distinguished by Crowder (Tell & Ferguson, 1974), active vocalization refers to presentation conditions where the to-be-remembered stimulus items are voiced aloud by S as they are visually presented. Under passive vocalization conditions, S listens to E read each stimulus item as it is visually presented. Tell & Ferguson suggested that active and passive vocalization differ mainly at the longer retention intervals, with recall much higher under the passive vocalization condition. This supported the idea that active vocalization could interfere with effective rehearsal or encoding strategies.

Another variable manipulated in search of significant effect on recall is whether interpolated tasks were rewarded or not. Soucar, Walk, and Covert (1971) suggested that retention of syllables is reduced under rewarded conditions. However, productivity and accuracy of the actual interpolated tasks was unchanged by reward.

The present experiment is designed to examine short-term recall as a function of stimulus type and length of delay interval. We hypothesize that both variables will be found to be significant. Specifically, we predict that: 1) recall proportion will be higher for words in compare to nonsense syllables; 2) recall proportion will decrease with the increase of delay intervals.

Method

Participants

Seventeen Hunter undergraduate psychology students served as Ss; twelve were female. Their age range was 18-32. They were demanded to participate in the experiment as part of their Experimental Psychology course requirement.

Apparatus The group of participants was assigned to pairs, who conducted the experiment in small cubicles, so that each pair worked independently of the other pairs. However, cubicles were not soundproof. Each of the participants served both as S and as E, in alternation with his/her partner. Two sets of stimuli were presented: nonsense syllables and words. The class of participants constructed both sets of stimuli. Each E individually selected twenty-four nonsense syllables