The mechanisms underpinning the successful use of visual supports with children with an autism spectrum disorder (ASD) have not been systematically investigated. One component of lexical processing which may evidence a modality-based advantage is the efficiency of word comprehension. Therefore, the aim of this thesis was to investigate the efficiency of lexical access in children with an autism spectrum disorder, specifically as a function of modality. In particular, the present thesis employed a series of semantic priming paradigms to examine the speed and accuracy of lexical decisions of single written words in comparison to single, spoken words.
The heterogeneity present in the language abilities of children with a confirmed diagnosis of an ASD is well documented. Chapter Two details the recruitment of 35 children aged 9 – 16 years and the grouping of these children according to whether they had a confirmed diagnosis of ASD (N = 20) or typical development (TD; N = 15). Traditional group comparisons revealed comparable group performances on standardised clinical assessments of language ability, reading ability and nonverbal intelligence. In contrast, a hierarchical cluster analysis on the combined groups identified two clusters of children. The first cluster (N = 6) comprised only children with an ASD who had significantly poorer language abilities and more severe autistic symptomatology than those in the mixed, second cluster (N = 29). The small number of participants in the first cluster precluded the inclusion of cluster-based comparisons using quantitative analyses in Chapters Three to Five. However, the performance of these two clusters is described in Chapter Six.
The limited literature on semantic priming using spoken word targets in children with typical development indicated that an initial, separate analysis of the performance of the TD children was warranted. Accordingly, Chapter Three details a study involving TD children aged between nine and 16 years (N = 15) who completed two experiments incorporating speeded lexical decision tasks. In the first uni-modal spoken prime word – spoken target word experiment, a main effect for relatedness was identified in both response latency and accuracy. In the second experiment (written prime – spoken target), no main effect for relatedness was identified. The results of the second experiment are considered in light of attention-driven processing demands of the cross-modal priming task and costs arising from shifts between stimulus modality.
Chapter Four compared the performance of children with an ASD (N = 18) and children with TD (N = 14) on primed, speeded lexical decisions on spoken targets. Children involved in Chapter Four experiments had also participated in Chapter Three experiments. The groups had comparable oral and written language skills and nonverbal cognitive abilities. In the first, uni-modal group comparison (spoken prime – spoken target), significant priming effects were found in both groups. When age was included as a covariate, a significant effect for age was identified. Post-hoc analysis demonstrated that this effect for age resulted from a decrease in response speed, and not from greater priming, with age. While an effect for age was identified in the second cross-modal experiment of Chapter Four (written prime – spoken target), no effect for relatedness or group was found. This result is interpreted with regards to the attentional capacity required for access to the lexicon via written stimuli within the developing semantic system.
Chapter Five utilised a similar semantic priming paradigm to Chapter Four but employed written word targets. The same cohort of children from Chapter Four participated in this study (ASD, N = 18 and TD, N = 14). In the first, cross-modal experiment with a spoken prime, no priming effect was identified in either group but, again, a significant effect for age was revealed. In contrast, results from the second uni-modal written target experiment included an age x group x relatedness interaction and each group underwent an age-based median split. Analysis of these age-based subgroups revealed priming only in the younger participants with an ASD and not in younger or older TD participants. Distinctions between the two experiments based on modality are discussed, as are the results of the written prime – written target experiment in terms of the lexical tuning hypothesis and the Dual Route Cascading Model.
In total, this series of priming experiments demonstrated that priming as a function of modality occurred comparably in the participant group with ASD and the ability matched group of peers with typical development. Both groups of participants demonstrated modality-shift-costs. However, some divergence in performance was evident in the uni-modal processing of written words: only young children with an ASD primed in the uni-modal written word experiment. An advantage to the efficiency of lexical access arising from written words by these children was one potential factor considered to underpin the different priming between participants. The contrasting lack of priming in older children with an ASD and children with typical development is also considered with regard to the efficiency of processing written words. With regards to the aims of the Thesis, the findings provide preliminary evidence suggesting that any processing advantage offered by written words to children with an ASD is influenced by their age, oral and written language competency, and attentional capacities. Further to the group comparisons, the clusters of children identified in Chapter Two who were clustered according to their oral and written language abilities, were also described and compared in Chapter 6. These descriptive comparisons suggest value in future investigations into the use of visual supports in children as a function of language ability using alternative methodologies.