Background and Aims

Previous research has suggested that a word association task called the Kent–Rosanoff task, which measures syntagmatic and paradigmatic associations, can help detect language disorder (LD) in children. We aimed to investigate the performance pattern between the results of the Kent–Rosanoff word association task and other language measures in two groups of children with autism: (1) children with autism and language disorder (ALD) and (2) children with autism and average language ability (ALN).

Methods

Forty-six children aged 6 to 9 years (9 girls, 37 boys) with autism participated in a comprehensive language assessment that focused on receptive vocabulary, receptive grammar and sentence repetition, and nonverbal cognitive ability. The children with autism were divided into two subgroups, namely ALD and ALN, based on language ability while the groups did not differ in nonverbal cognitive ability. The 50-item Kent–Rosanoff list was used to elicit word associations. The children's responses were categorized into four different categories: phonological, syntagmatic, paradigmatic, and other/no response.

Results

Our results did not reveal any differences between children with ALN and ALD on the Kent–Rosanoff word association task. The most frequent category coded for both groups was paradigmatic responses.

Conclusions

Children with ALD did not differ in their semantic associations when compared with the ALN group; thus, sources of diversity in language profiles in preschoolers with autism should perhaps be sought elsewhere or by using alternative measures.

Implications

Our results show that the Kent–Rosanoff word association task did not distinguish between ALN and ALD. Future studies should continue exploring sensitive assessment formats of semantic depth in young children with or without LD and autism.

Search results

Is the Kent–Rosanoff Word Association Task Helpful in Characterizing Language Difficulties in 6- to 9-Year-Old Children with Autism?