Differentiating Children With and Without Language Impairment Based on Grammaticality

Eisenberg, Sarita; Guo, Ling-Yu

doi:10.1044/0161-1461(2012/11-0089)

Cited by 68 publications

(102 citation statements)

References 42 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Following Sackett (1991), a cutoff score for PGU was determined by first using a receiver operating characteristic (ROC) curve to generate pairs of sensitivity and specificity rates for a range of cutoff scores and then choosing the cutoff that maximized diagnostic accuracy. A PGU score of 58.32% yielded sensitivity at a 100% level and specificity at an 88% level for differentiating previously diagnosed 3-yearold children with and without LI (Eisenberg & Guo, 2013). These data supported the use of PGU for measuring children's grammaticality.…”

supporting

confidence: 60%

“…A PGU score of 58.32% yielded sensitivity at a 100% level and specificity at an 88% level for differentiating previously diagnosed 3-yearold children with and without LI (Eisenberg & Guo, 2013). These data supported the use of PGU for measuring children's grammaticality. …”

supporting

confidence: 60%

“…However, children with LI have also been reported to make other grammatical errors, such as argument omissions , pronoun errors (Loeb & Leonard, 1991;Moore, 2001), and omissions and errors on other grammatical morphemes (e.g., articles, plurals) more often than typically developing children (Leonard, Eyer, Bedore, & Grela, 1997;Polite, Leonard, & Roberts, 2011;Watkins & Rice, 1991). Eisenberg and colleagues (Eisenberg & Guo, 2013;Eisenberg, Guo, & Germezi, 2012), therefore, suggested using percent grammatical utterances (PGU) as a broader measure that incorporates a variety of properties that can affect grammaticality rather than focusing exclusively on verb tense marking.Eisenberg and colleagues examined PGU in 3-yearold children with and without LI using picture-elicited language samples (Eisenberg & Guo, 2013;Eisenberg, Guo, & Germezi, 2012). In those studies, language samples were collected by asking children to describe 15 pictures one at a time in response to four elicitation questions (see Appendix A).…”

mentioning

confidence: 99%

“…In addition, four items (i.e., pictures and/or interview topics) are needed to calculate reliable percent correct information units in an item-based task for adults (Brookshire & Nicholas, 1994). However, the numbers of utterances or items that are needed in order to compute reliable PGU scores in children remains an open question, even though studies have shown that PGU scores differentiate children with and without LI (Eisenberg & Guo, 2013;Fey, Catts, Proctor-Williams, Tomblin, & Zhang, 2004;Guo & Schneider, 2014;Souto et al, 2014). …”

mentioning

confidence: 99%

“…We used number of pictures as the measure of sample size because this coincides with clinical practice. That is, clinicians would administer a specific number of pictures to elicit the language sample rather than basing task length on reaching a specific number of utterances.The picture description procedure used to collect language samples for calculating PGU (Eisenberg & Guo, 2013) had included 15 pictures. The average number of C-units produced by the children in the picture-elicited language samples was 72 (standard deviation [SD] = 18) for 3-year-old children with typical language and 62 (SD = 15) for children with LI.…”

mentioning

confidence: 99%

See 4 more Smart Citations

Sample Size for Measuring Grammaticality in Preschool Children From Picture-Elicited Language Samples

Eisenberg

Guo

2015

LSHSS

Self Cite

View full text Add to dashboard Cite

Purpose: The purpose of this study was to investigate whether a shorter language sample elicited with fewer pictures (i.e., 7) would yield a percent grammatical utterances (PGU) score similar to that computed from a longer language sample elicited with 15 pictures for 3-year-old children. Method: Language samples were elicited by asking forty 3-year-old children with varying language skills to talk about pictures in response to prompts. PGU scores were computed for each of two 7-picture sets and for the full set of 15 pictures.Results: PGU scores for the two 7-picture sets did not differ significantly from, and were highly correlated with, PGU scores for the full set and with each other. Agreement for making pass-fail decisions between each 7-picture set and the full set and between the two 7-picture sets ranged from 80% to 100%. Conclusion: The current study suggests that the PGU measure is robust enough that it can be computed on the basis of 7, at least in 3-year-old children whose language samples were elicited using similar procedures. L anguage sample analysis (LSA) is an important part of assessing a child's language. Some LSA provides a fine-grained analysis of a particular aspect of language, such as use of verb tense markers or noun phrase elaboration, for the purpose of determining strengths and weaknesses and selecting treatment goals (e.g., Eisenberg et al., 2008;Leonard, Camarata, Brown, & Camarata, 2004). Other LSA measures, such as mean length of utterance and number of different words, broadly characterize a child's linguistic performance for the purpose of identifying a language impairment (LI) or measuring change over time (e.g., Rice et al., 2010;Scott & Windsor, 2000). These are referred to as general language performance measures. The current article looks at one general language performance measure -percent grammatical utterances-that measures grammaticality (i.e., the extent to which a child's language production conforms to the conventions of English grammar and usage) using picture-elicited language samples.One frequently given reason for not using language sample analysis in clinical work is the amount of time it takes for eliciting, transcribing, and analyzing the sample (Hux, Morris-Friehe, & Sanger, 1993;Kemp & Klee, 1997). This time could be reduced by using shorter samples. However, it needs to be established whether those shorter samples will yield the same results as longer samples. In the current article, we focused on evaluating whether percent grammatical utterances in 3-year-old children varies with the number of pictures that are used to elicit language samples. Why Grammaticality Is ImportantOne aim of the common core educational standard for language is for students to demonstrate command of the conventions of standard English grammar and usage when speaking or writing (e.g., State of New Jersey Department of Education, 2010). This standard is particularly relevant for children with LI because the presence of grammatical deficits is a key characteristic of these children ...

show abstract

supporting

confidence: 60%

supporting

confidence: 60%