Three studies are reported, in which a correlational approach was used to investigate the nature of the relationship between maximum speech rate and auditory memory span in children.
A two-part correlational study examined the relationship between maximum speech rate and auditory memory span. Study 1A represented an empirical investigation of the measurement sensitivity of speech rate and memory span, in a group of 152 kindergarten, second- and fourth-grade children. Through a comparison of the strength of correlations it was shown that there was little difference in sensitivity between two alternative methods of scoring memory span. Speech rate was more sensitively assessed using single short-word repetitions.
Study IB, which included an additional sample of 53 sixth-grade children, examined the suggestion that maximum speech rate for word groups may itself involve memory demands, contaminating the
correlation between speech rate and memory span in younger children. Analyses using composite speech rate and memory span measures showed that speech rate for word triples shared variance with memory span that was independent of speech rate for single words. Following from these results, further studies reported here used single-word speech rate measures and memory span scored for the number of trials correctly recalled.
Study 2 further examined the association between speech rate and memory span, within the context of a global processing speed model of development, which proposes that all developmental changes in cognitive abilities can be linked to global increases in processing efficiency. An independent sample of 142 children from first- to seventh- grade completed tests of processing speed, single-word speech rate, auditory memory span, and word-level reading. Hierarchical multiple regression and structural equation modelling analyses established that
changes m processing speed with age explained unique variation in both speech rate and memory span. However, the relationship between speech rate and memory span was only marginally significant, and not all age changes in speech rate were accounted for by processing speed, suggesting that some revision of the model was necessary. In addition, word-level reading was correlated with memory span, independent of all other processes. The results of Study 2 suggest that auditory memory span and word-level reading share underlying processes that were not measured in the current study.
Following from these results, a two-part correlational study examined an elaborated global processing speed model of memory span development, hi Study 3A, 220 first- to seventh-grade children were tested on measures of processing speed, single-word speech rate, vocabulary, phonological awareness, memory span and word-level reading. Hierarchical multiple regression and structural
equation modelling analyses supported two alternative versions of an elaborated global processing speed model, with either long term memory (operationalised as receptive vocabulary), or phonological processing skill (operationalised as phonological awareness) predicting independent variation in auditory memory span performance. Maximum speech rate did not predict unique variation in memory span, independent of vocabulary or phonological awareness. A series of parallel regression analyses also showed that word-level reading no longer accounted for unique variance in memory span after controlling the effects of phonological awareness. In contrast, vocabulary, speech rate, processing speed and age did not adequately explain the underlying association between memory span and word-level reading.
Study 3B used a reading-level design to further examine the source of auditory memory span deficits in reading-delayed children. Twenty-nine less-skilled readers were
individually matched to 29 younger normally-achieving readers, with results showing that, across all the experimental measures in Study 3 A, only phonological awareness was deficient in less-skilled readers compared to reading-level matched controls. This finding highlights the need to include phonological processing skills in models explaining the development of both auditory memory span and word-level reading.
Altogether, these studies provided some support for an overall association between measures of maximum speech rate and auditory memory span across a wide age range in children, but did not show a consistent relationship between these measures within smaller age ranges of children. Moreover, this association was not independent of other processes such as long-term memory, or phonological processing skill. General support for an elaborated version of the global processing speed model of auditory memory span development was observed, although the present
results suggest that covert rehearsal speed, as indexed by maximum speech rate, is not a key determinant of children's auditory memory span performance.