The influence of distinctiveness on stimulus selection

Abstract

Paired-associate lists with compound stimuli composed of three adjectives which varied along a distinctiveness (D) dimension and single digit responses were presented to 18Ss. The Ss were given 12 learning trials and then a transfer task on which they were asked to recall the co"ect response when each word was presented alone. The results indicated that selection occu"ed along the D dimension.Several investigators have found evidence for cue selection in paired associate learning with compound stimuli. For instance, Jenkins (1963) andPostman &Greenbloom (1967) examined response recall when only one letter from the stimulus trigram was presented and found significant amounts of recall, particularly for the ftrst element. Underwood, Ham, & Eclcstrand (1962) used trigrams and color and found selection of color to be dominant.The basis for selection of the effective stimulus cue in verbal learning situations is important in determining expected responses and the course of learning. This study involves examination of the effects of a variable labeled distinctiveness (D) on cue selection in the paired-associate situation with three words as the compound stimulus.Since words as units differ considerably from letters or trigrams in their attributes it is possible that different factors will be· involved in the selection process. The basis for this study was the hypothesis that a more distinctive word will be selected more frequently than a less distinctive word. D is defmed here as a function of the results of a short term recosnition memory experiment of the sort performed by Shepard & Teghtsoonian (1961). The difference between hits (old words called old) and false alarms (new words called old) was the value taken for D. Furst (1967) showed D to be a potent variable on the stimulus side of paired associated lists when the lists were heterogeneous with respect to the D variable. Since D did not have a significant effect in comparisons across lists which were homogeneous with respect to it the assumption of a selection factor is reasonable. The compound stimulus technique provides one way of testing this.Other attributes of words which have been shown to affect learning are meaningfulness (m) and frequency of usage (u). Values of these attributes for the words used were obtained using Noble's (1952) production technique for m and subjective ratings for u. METHODThe Ss were 18 introductory psychology students at the University of Georgia.They were required to learn six paired-associates using the standard anticipation technique. Each of the six stimuli consisted of three words with one word from each of the D categories. The six high D words (TARDY, ARCTIC, ZEALOUS, JUICY, UNDONE, and RABID) had values ranging from 43 to 45, all six medium D words (DEMURE, LOYAL, ALIVE, OUTER, MASSIVE, and NIMBLE) had the value of 38, while the six low D words (LUSTY, EXTREME, SIMPLE, JAGGED, VALID, and FATAL) had values ranging from 24 to 32. The mean m values for the high, medium and Psychon. Sci., 1968, Vol. 13 (6) low D words were...

Cue effectiveness and abstraction in paired-associate learning.

“…In cases where there were fairly obvious differences in component difficulty, it has been found that the less difficult component is the more effective cue following compound-response learning. This has been demonstrated for components differing in formal similarity (Cohen & Musgrave, 1966;Newman & Taylor, 1963), in meaningfulness Lovelace, 1968;Spear et al, 1964) and in distinctiveness (Jacobus & Leonard, 1968). The comparison of the compound with the components is more interesting.…”

Stimulus coding and selection as a function of response relatedness

“…This selection process would thus entail the intennediate step of coding FML as FeMaLe. That Ss can be influenced to use such a code when responses are trigrams and related words are stimuli (i.e., response coding) was demonstrated by Podd & Spear (1967).…”

“…Other factors that have been shown to influence the stimulus-selection process are intralist similarity of elements (Cohen & Musgrave, 1966), "distinctiveness" of words (Jacobus & Leonard, 1968), and spatial position of the various elements in a compound (Cohen & Musgrave, 1964Postman & Greenbloom, 1967). Solso (1968) has demonstrated that characteristics of the response member of a pair can also influence selection of particular elements in astimulus eompound.…”

“…List R pairs were drawn from materials used by Podd & Spear (1967). list U consisted of the same stimuli and responses used in List R, but they were inappropriately paired so as to be unrelated.…”

The effects of imagery on the paired-associate learning of adjectives

“…However, since some studies have shown D to be a factor in stimulus selection (Jacobus & Leonard, 1968;Leonard & Jacobus, 1969) in PA learning, it was decided to include this control. Additional controls were i mposed on the level of meaningfulness (m) and on the level of frequency of usage (u).…”