Test-Retest Reliability Relational Phonological Analyses for 2-year-old Late Talkers

Toddlers identified as ‘late talkers’ are 2-year-old children with late-developing expressive vocabularies, but are not exhibiting characteristics of known developmental disorders that may cause language delay. Late talkers also exhibit delays in speech sound acquisition (e.g., inability to regularly produce sounds like “w” or “b” accurately) compared to peers. This is because speech and language acquisition influence each other in early development. Researchers theorize that children who use more words produce more sounds, whereas young children who use fewer words are not able to produce a large number of different sounds. Previous studies supported this theory. Planning appropriate intervention for late talkers begins with comprehensive evaluations of speech and language skills. To accomplish this, speech-language pathologists (SLPs) often collect conversational speech samples from children to analyze speech and language skills. Speech sounds are usually analyzed with informal measures such as relational phonological analyses (e.g., percent consonant correct; place-manner-voice). If these measures are inaccurate, SLPs may be misinformed about children’s true abilities, reducing the likelihood they will provide the most effective intervention. The present project with late talkers (n = 3) was an expansion of a study conducted by Morris (2009). Morris’ younger participants (18- to 24-month-olds) exhibited similar language skill proficiencies as late talkers. The project purpose was to investigate test/re-test reliability of two relational phonological analyses. In support of Morris’ findings, present findings indicated measurement discrepancies for samples obtained one week apart under identical conditions. Use of relational analyses may yield inconsistent results for 2-year-old late talkers