Relationships among conversational language samples and norm-referenced test scores
Article
Owens, RE, Pavelko, SL. (2017). Relationships among conversational language samples and norm-referenced test scores
. 2(1), 43-50. 10.21849/cacd.2017.00052
Owens, RE, Pavelko, SL. (2017). Relationships among conversational language samples and norm-referenced test scores
. 2(1), 43-50. 10.21849/cacd.2017.00052
Purpose: Research demonstrates that many speech-language pathologists (SLPs) do not routinely include language sample analysis (LSA) in their clinical practice because LSA has limited recognition as a valid assessment measure. Limited research suggests that some LSA values obtained from narrative samples correlate with the results of standardized language tests. This research examined the relationship among values obtained from conversational language samples and the results of standardized testing. Methods: This study investigated whether LSA values obtained from conversational language samples shared a relationship with the results of standardized language testing. A total of 16 children ages 43-90 months (M=61.5 months) completed three subtests of a standardized language test and a 15-minute conversational language sample. Fifty-utterance language samples were analyzed for four LSA values including mean length of utterance (MLUS), total number of words (TNW), clauses per sentence (CPS) and words per sentence (WPS). Results: Results revealed that three of the four LSA values (MLUS, TNW, and WPS) demonstrated statistically significant (ps<.006) strong correlations (rs>.65) with the results of norm-referenced language testing. The partial correlations and the zero-order correlations were significant, suggesting age had little influence in controlling for the relationships. Conclusions: Conversational language samples complement norm-referenced tests well. Results support further exploration of the relationships among LSA measures obtained from conversational samples and the results of standardized language testing.
