What did I just hear? Phonological similarity as an index of short-term memory precision for words and nonwords

The concept of short-term memory (STM) precision has been defined as the resolution at which items are maintained and represented in STM (Joseph et al., 2015). It has to be distinguished from STM capacity, which refers to the number of items that are recalled in a STM task. So far, the concept of STM precision has received very little interest in the verbal STM domain. In two studies, we assessed the sensitivity to different degrees of phonological similarity between memory and probe items as a potential index of verbal STM precision at a lexical and sublexical level. In Study 1, 60 young adults had to maintain auditory lists of 6 words. In Study 2, 20 young adults had to maintain single four-syllable nonwords. After a delay, a probe stimulus was presented, and participants had to decide whether it matched one of stimuli in the list (Study 1), or whether it matched the stimulus to-be-maintained (Study 2). For each study, negative probes showed different degrees of phonological proximity with the target. Using Bayesian ANOVA, we observed robust evidence for an influence of phonological proximity on probe recognition performance: the more similar the negative probes to the target item, the higher the rate of false recognition and the slower the response times. Critically, we observed significant interindividual variability in sensitivity to phonological proximity. Overall, the variation of target-probe phonological similarity along a continuum may be a promising experimental procedure for studying the concept of STM precision in the verbal domain.