A tutorial on the psycholinguistic analysis of word-final repetitions

Purpose: We aim to provide a tutorial describing how atypical disfluencies can be assessed from a psycholinguistic perspective. We specifically focus on word-final repetitions (WFRs), e.g., “school-ool”. Methods: We identify the type of data that can be used to assess whether WFRs arise from difficulty at each stage of the word production process. WFRs produced by an adolescent with autism spectrum disorder are used to assess difficulty at each of these stages. Results: WFRs tended to occur on content rather than function words, and the portion repeated seemed to be driven by the syllabic rather than morphological structure. The pause time between the end of a WFR-containing word and the beginning of the repetition (e.g., between “school” and “ool”) was at least 298 ms. In contrast, there was only an audible pause between the repetition and the next word 42% of the time; the other 58% of the time, the repetition led directly into the next word. WFRs also tended to occur in longer utterances and in utterance-medial and -final positions. Conclusion: Our analysis suggested that the best candidates for loci where the production process breaks down to cause WFRs are (a) external channel monitoring of WFR words, which is consistent with the long pauses that occurred between the WFR word and the repetition, or (b) syntactic planning, which is supported by the occurrence of WFRs away from utterance-initial words and in longer utterances. We discuss the limitations of data collected through language sampling and make suggestions for future research.