Reversing the phonological similarity effect

Abstract

The phonological similarity effect-poor retention of order for lists of similar-sounding items-is a benchmark finding in the short-term memory literature. In our first two experiments, we show that the effect actually reverses following relatively brief periods of distraction, yielding better order retention for similar than for dissimilar lists, provided that different items are used on every trial. In Experiment 3, the same items were used on every trial and similar lists produced poorer performance across all three retention intervals. The results are interpreted from a general discrimination framework: Items are viewed as occupying positions in a multidimensional space defined by list and within-list dimensions.In immediate serial recall, lists composed of similarsounding items "ire typically remembered less well than lists of dissimilar items (Baddeley, 1966;Conrad, 1964;Conrad & Hull, 1964;Hintzman, 1967). Increasing the phonological similarity among the members of a to-berecalled set results in more omission errors at test, as well as more item-to-item transpositions in the output order (Coltheart, 1993;Conrad, 1965;Hintzman, 1965). The disruptive effects of phonological similarity are robust and have been replicated with a wide variety of experimental designs and materials, including the use of letters (Conrad, 1964), words (Baddeley, 1966), pictures (Hitch & Halliday, 1983), visual presentation (Baddeley, 1968, auditory presentation (Longoni, Richardson, & Aiello, 1993;Posner & Konick, 1966), and RSVP (Coltheart & Langdon, 1998). It is fair to say that the phonological similarity effect has achieved the status of a "benchmark" finding in the immediate memory literature, and most theories of short-term memory include mechanisms that are specifically designed to account for the phenomenon (e.g., Baddeley, 1986Baddeley, , 1992Burgess & Hitch, 1992;Drewnowski, 1980;Nairne, 1990a;Schweickert, Guentert, & Hersberger, 1990;Shiffrin & Cook, 1978).Nairne's (1990a) feature model, for example, is a member of a class of models that places the locus ofthe effect primarily at the point ofretrieval, where the memory trace is interpreted or "deblurred," Confusions are thought to arise when degraded primary memory traces are compared with candidate members of a secondary memory search set. If list items sound similar (i.e., they share common vowel and/or consonant sounds), it is assumed that their mnemonic representations contain many overlapping features. Feature overlap, in tum, means that the representation of a given item in the sequence is apt to be confused with other list items, leading to incorrect trace interpretations at retrieval. Models that place the locus ofsuch simCorrespondence should be addressed to 1. S. Nairne, Department of Psychological Sciences, Purdue University, West Lafayette, IN 47907-1364 (e-mail: nairne@psych.purdue.edu). 45ilarity effects during the trace interpretation stage have been shown to mimic the empirical data quite well (Nairne, 1990a; see also Lewandowsky & Murdock, 1989).The phon...

Age and redintegration in immediate memory and their relationship to task difficulty

“…However, there are a growing number of instances where phonologically similar items have been better recalled than dissimilar items (Fallon et al, 1999;Nairne & Kelley, 1999;Nimmo & Roodenrys, 2004). If our ideas are correct and we replicate the findings of Experiment 1, a linear relationship between task difficulty and similarity advantage should be apparent, particularly for item scoring.…”

“…For the correct-in-position measure, there was a similarity disadvantage at low levels of task difficulty, which reversed to a similarity advantage at high levels of task difficulty (Nairne & Kelley, 1999). When correctin-position scoring was decomposed into its components, there was a consistent similarity advantage for item scoring and a consistent similarity decrement for order accuracy (Fallon et al, 1999).…”

“…In their terms, similarity narrows the search set in long-term memory. Note that the cuing function is again code independent, as is implied in many accounts of short-term similiarity effects (Fallon et al, 1999;Nairne & Kelley, 1999;Nimmo & Roodenrys, 2004;Poirier & SaintAubin, 1995;Saint-Aubin & Poirier, 1999b).…”

Suppression effects on musical and verbal memory

“…In other models (Brown, Preece, & Hulme, 2000;Nairne, 1990), forgetting is accounted for solely by interference. In the feature model (Nairne, 1990;Nairne & Kelly, 1999), memory traces are represented by feature vectors, and memory relies on the availability of retrieval cues; the feature model distinguishes between modality-independent features (in an abstract representation) and modalitydependent features (those related to presentation conditions, including presentation modality) and assumes that they do not interfere with each other. Auditory traces are thought to have more modality-dependent features than do visual traces.…”

Phonological similarity in working memory span tasks

“…One explanation for this reversal is that rhyming words provide a category cue that facilitates list retrieval (Gupta et al, 2005;Nairne & Kelley, 1999), thus enhancing recall in the same manner as semantic similarity between memory items on memory span tasks (Huttenlocher & Newcombe, 1976;Saint-Aubin & Poirier, 1999).…”