An analysis of the test of language development-primary for item bias

Carol Scheffner Hammer, Maria Pennock-Roman, Sarah Elizabeth Zappe, J. Bruce Tomblin

Research output: Contribution to journalArticle

19 Scopus citations

Abstract

The purpose of this research was to examine the Test of Language Development-P:2 (TOLD-P:2; Newcomer & Hammill, 1991) for item bias. The TOLD-P:2 was administered to 235 African American and 1,481 White kindergarten children living in the Midwest. Test items were examined for evidence of differential item functioning (DIF) using inferential and descriptive methods. Sixteen percent of all items of the TOLD-P:2 were found to have DIF. Of these items, 75% were found to be harder for the African American group. The percent-ages of items on the five core subtests identified as having DIF were as follows: Picture Vocabulary, 17%; Oral Vocabulary, 17%; Grammatic Understanding, 12%; Sentence Imitation, 20%; and Grammatic Completion, 13%. The implications of these findings are discussed in relation to the TOLD-P:3.

Original languageEnglish (US)
Pages (from-to)274-284
Number of pages11
JournalAmerican journal of speech-language pathology
Volume11
Issue number3
DOIs
StatePublished - Jan 1 2002

All Science Journal Classification (ASJC) codes

  • Otorhinolaryngology
  • Developmental and Educational Psychology
  • Linguistics and Language
  • Speech and Hearing

Fingerprint Dive into the research topics of 'An analysis of the test of language development-primary for item bias'. Together they form a unique fingerprint.

  • Cite this