Error type and lexical frequency effects: Error detection in Swedish children with language impairment.

A.E Hallin, Christina Reuterskiold

Research output: Contribution to journalArticlepeer-review


Purpose: The first aim of this study was to investigate if Swedish-speaking school-age children with language impairment (LI) show specific morphosyntactic vulnerabilities in error detection. The second aim was to investigate the effects of lexical frequency on error detection, an overlooked aspect of previous error detection studies. Method: Error sensitivity for grammatical structures vulnerable in Swedish-speaking preschool children with LI (omission of the indefinite article in a noun phrase with a neuter/common noun, and use of the infinitive instead of past-tense regular and irregular verbs) was compared to a control error (singular noun instead of plural). Target structures involved a high-frequency (HF) or a low-frequency (LF) noun/verb. Grammatical and ungrammatical sentences were presented in headphones, and responses were collected through button presses. Results: Children with LI had similar sensitivity to the plural control error as peers with typical language development, but lower sensitivity to past-tense errors and noun phrase errors. All children showed lexical frequency effects for errors involving verbs (HF > LF), and noun gender effects for noun phrase errors (common > neuter). Conclusions: School-age children with LI may have subtle difficulties with morphosyntactic processing that mirror expressive difficulties in preschool children with LI. Lexical frequency may affect morphosyntactic processing, which has clinical implications for assessment of grammatical knowledge.

Original languageEnglish (US)
Pages (from-to)2924-34
JournalJournal of Speech, Language, and Hearing Research
Issue number10
StatePublished - Oct 2017


Dive into the research topics of 'Error type and lexical frequency effects: Error detection in Swedish children with language impairment.'. Together they form a unique fingerprint.

Cite this