The suffix effect and preattentive unit-formation in visual short-term memory.

Abstract

Immediate ordered recall of digits from visual short-term memory was tested, primarily using a long-exposure (approximately 3^ sec) visual presentation during which subjects spoke aloud (articulatory suppression). A to-be-ignored digit placed to the right of the visual presentation, called a suffix, impaired recall approximately one digit, suggesting that the suffix could not be excluded from visual short-term memory. The suffix effect was attenuated when the suffix was segregated from the to-be-recallcd digits by Gcstalt principles of grouping, suggesting that a segregated suffix could be excluded from visual short-term memory. The results suggest that the visual field is organized into units preattentively according to Gestalt principles of grouping and that when a unit (such as a string of digits) is selected, the entire unit enters visual short-term memory. RESUME Le rappel en ordre et immediat de chiffres de la memoire visuelle a court terme est evalue en utilisant une presentation visuelle de longue duree (environ 3-4 sec) au cours de laquelle les participants partent a haute voix (suppression articulatoire). Un chiffre qui doit etre ignore, place a droite de la presentation visuelle, et appelc suffixc, nuit au rappel d'environ un chiffre. Ce resultat suggere que le suffixe ne peut etre exclu de la memoire visuelle a court terme. On parvient a attenuer cet effet en isolant le suffixe des chiffres a se rappeler grace aux principes gestaltiques du regroupement. Les resultats suggerent que le champ visuel est organise en unites de facon preattentive selon les principes gestaltiques de regroupement. Aussi, quand une unite (tellc unc suite de chiffres) est selectionnee, l'unite entiere entre dans la memoire visuelle a court terme.

In search of a strong visual recency effect

“…Immediately following presentation, the subjects recalled the letters in serial order. Frick (1985;Frick & DeRose, 1986) had previously found that preventing subvocalization of visually presented list items tended to induce reliance on a visual code. The subjects in this experiment were therefore required to silently mouth the syllable "blah" during all conditions.…”

Two-component theory of the suffix effect: Contrary evidence

“…Specifically, Greene (1987) failed to replicate the visual suffix effects documented in each of three previous reports (Frick & De Rose, 1986b;Hitch, 1975;Kahneman, 1973). Even had the effects been replicated, it can be argued that the findings documented in the three reports do not seriously challenge bottom-up, overwriting accounts of the suffix effect (see Crowder, 1986;Greene, 1987).…”

“…The to-be-remembered items were the digits 0-9, except for 3, which, following LeCompte and Watkins (1995) and Frick and De Rose (1986b), was reserved for use as the suffix. A separate set of 63 digit lists was constructed for each participant.…”

“…First, Kahneman used such a brief presentation duration that, as some have argued, his findings may be better cast in perceptual, rather than mnemonic, terms. Second, as has been noted by some, although Frick and De Rose (1986b) used a presentation duration more characteristic of auditory suffix experiments, their lists, unlike typical auditory lists, were not of predictable supraspan length. Third, Hitch presented the suffix and no-suffix lists in separate blocks, thereby making the suffix predictable.…”

“…Furthermore, suffix effects should not, according to the PAS theory, occur with static visual presentation, because iconic information is too fleeting to contribute to recall in a typical suffix effect procedure, meaning that recall should be an entirely postcategorical phenomenon. Such suffix effects occasionally have been reported (Frick & De Rose, 1986a, 1986bHitch, 1975), but they tend to be small and not always replicable (Greene, 1987; see also Engle, 1974). In addition, because visual and auditory channels are assumed not to meet prior to the categorization of information into words (Crowder, 1976, p. 59), even an auditory suffix should not produce a typical suffix effect on the recall of visual list items (see Crowder, 1976;Crowder & Morton, 1969).…”