No Evidence That Articulatory Rehearsal Improves Complex Span Performance

Alessandra S Souza,Klaus Oberauer

doi:10.5334/joc.103

Abstract

It is usually assumed that articulatory rehearsal improves verbal working memory. Complex span is the most used paradigm to assess working memory functioning; yet, we still lack knowledge about how participants rehearse in this task, and whether these rehearsals are beneficial. In Experiment 1, we investigated the patterns of naturally occurring overt rehearsals in a complex span task requiring processing of a non-verbal distractor task. For comparison, another group of participants completed a matched simple span task with an unfilled delay in between the memoranda. Time permitting, participants rehearsed the memory list in forward serial order, a strategy known as cumulative rehearsal. The degree of cumulative rehearsal was correlated with recall accuracy in both span tasks. Rehearsal frequency was, however, reduced in complex span compared to simple span. To assess the causal role of rehearsal in complex span, we trained a group of participants in a cumulative rehearsal strategy in Experiment 2. This instruction substantially increased the prevalence of cumulative rehearsals compared to a control group. However, the increase in cumulative rehearsal did not translate into an increase in recall accuracy. Our results provide further evidence that rehearsal does not benefit working memory performance.

Highlights

When asked to briefly hold in mind a random list of digits, letters, or words – such as a phone number or a pin-code – people often repeat this information over and over to themselves, a behavior known as articu latory rehearsal (Baddeley, 1986, 2012; Logie, 2011)
The goal of the present study is to fill this gap by assessing when and how participants rehearse in complex span, and what effect it has on recall accuracy
Participants in the Overt Rehearsal Complex Span group tended to respond slightly less accurately and more slowly than participants in the Silent Complex Span group, but there was not enough evidence in the data to support a difference between groups

Summary

Introduction

When asked to briefly hold in mind a random list of digits, letters, or words – such as a phone number or a pin-code – people often repeat this information over and over to themselves, a behavior known as articu latory rehearsal (Baddeley, 1986, 2012; Logie, 2011). Despite the widespread belief that rehearsal is beneficial, there is limited evidence linking rehearsal causally to WM performance (for reviews see Lewandowsky & Oberauer, 2015; Oberauer, 2019) This is true for complex span, which is one of the most used tasks to assess WM functioning (Conway et al, 2005). Distractor processing in complex span is assumed to disrupt refreshing to the degree that the distractor task requires central attention, but it only disrupts rehearsal if it requires articulation. Other commonly employed manipulations are variations in word length (assumed to vary pronunciation time) or phonological similarity (assumed to increase confusion of articulated words that are phonologically similar) All these manipulations yield worse performance in WM tasks (Baddeley, Lewis, & Vallar, 1984; Larsen & Baddeley, 2003). Word-length co-varies with orthographic neighborhood-size, and this variable was found to explain the effect of word-length observed in previous studies (Jalbert, Neath, Bireta, & Surprenant, 2011; Jalbert, Neath, & Surprenant, 2011)

Objectives

Results

Conclusion