Not so fast! Talker variability in serial recall at standard presentation rates

Area of Science:

• Cognitive Psychology
• Human Memory Research
• Speech Perception

Background:

• The talker-variability effect (TVE) demonstrates impaired serial recall performance with alternating male and female voices.
• Previous research (Hughes et al., 2009, 2011) used TVE as evidence for a perceptual-gestural account of serial recall.
• The generalizability of TVE was questioned due to non-standard, fast stimulus presentation rates in prior studies.

Purpose of the Study:

• To investigate the generalizability of the TVE to standard serial recall task parameters.
• To compare recall performance in talker-variable conditions at both fast and slow stimulus presentation rates.
• To examine the influence of item properties on the TVE and challenge existing theoretical explanations.

Main Methods:

• Experiment 1: Systematic replication of Hughes et al. (2009) using a novel stimulus set.
• Experiment 2: Tested TVE generalizability and item property influence with a new stimulus set.
• Direct comparison of recall performance across fast and slow stimulus presentation rates in talker-variable conditions.

Main Results:

• A robust TVE was observed at the fast, atypical presentation rate in both experiments.
• The TVE was unreliable at the slower, standard item presentation rate.
• Error analysis indicated that item recall contributes to the TVE, contradicting the perceptual-gestural account.

Conclusions:

• The TVE's generalizability is dependent on stimulus presentation rate, challenging prior claims.
• Findings necessitate a re-evaluation of the perceptual-gestural account of serial recall.
• Alternative explanations focusing on the resource cost of categorical speech perception in talker variability are proposed.