Presenters: Melanie Bledsoe ; Alyssa Davidson ; Sarah Ericson ; Sarae Niemela
Authors: Melanie Bledsoe, Alyssa Davidson, Hailey Barker, Sarah Ericson, Sarae Niemela, Kristi Jones
Faculty Advisor: Douglas Petersen
Institution: Brigham Young University
Language sample analysis (LSA) is a vital part of the language assessment process; however, Pavelko et al., (2016) found that only 67% of speech-language pathologists (SLPs) use LSA, with the majority citing lack of time as the primary barrier. In response, researchers have investigated whether shorter language samples can be reliably used in place of lengthier samples (Casby, 2011; Gavin & Giles, 1996). The majority of these researchers have analyzed conversational language instead of narratives (stories), which could feasibly be very brief yet still contain all elements of story grammar and language complexity. Furthermore, most studies have included small sample sizes, focused on a narrow band of ages, and did not examine relative and absolute reliability, where relative reliability pertains to correlational data and absolute reliability reflects the extent to which two language samples yield the same results. The purpose of this study was to examine the absolute and relative reliability of long versus short narrative retells with a large sample of students. Participants included 193 kindergarten through sixth-grade students who were asked to retell two short narratives from the Narrative Language Measures, which took approximately four minutes to elicit and score. Participants were also asked to retell a lengthier story from the wordless picture book “Frog Where Are You?”, which took approximately 20-30 minutes to elicit, transcribe, segment, and score. Story grammar and measures of language productivity and complexity were analyzed. Results indicated that there were moderate to strong correlations for nearly all complexity and productivity measures between the short and long narratives. However, there were significant differences across almost all measures, with the shorter narrative yielding higher scores. Hence, short narrative language samples can provide as much if not more information than longer narrative samples, suggesting that language sampling could be clinically feasible when short narratives are used.