NotesFAQContact Us
Search Tips
ERIC Number: ED262952
Record Type: Non-Journal
Publication Date: 1985-Nov
Pages: 19
Abstractor: N/A
Reference Count: N/A
An Examination of Possible Item Bias in the Responses of Spanish-Speaking and English-Speaking Children on the Cooperative Preschool Inventory Using Logit Regression.
Powers, Stephen; And Others
A total of 102 Spanish-speaking and 104 English-speaking preschool 4-year-old pupils were individually administered the Spanish and English versions of the Cooperative Preschool Inventory (CPI) to examine item responses for possible bias. Miminum logit chi-square regression was applied to each of the 30 items of the CPI verbal scale. Each of the verbal items was cross-classified by response (correct, incorrect), verbal achievement (low, middle, high), and language proficiency (monolingual Spanish or monolingual English). A total of four items exhibited a significant chi-square for language proficiency in the model. In each case, the English speaking students' probability of a correct response was higher than the Spanish-speaking students. The only item displaying an achievement by language proficiency interaction effect was the question, "If you wanted to find a lion, where would you look?" Spanish-speaking children of low achievement had a higher probability (24%) of answering this item correctly than low achieving English-speaking children (4%). To know if these differences are, indeed, more than just potential bias would require follow-up interviews and further examination of the items. Future research may help to verify if there is bias in the items of the Cooperative Preschool Inventory. Tables and figures are included. (NEC)
Publication Type: Reports - Research; Speeches/Meeting Papers
Education Level: N/A
Audience: Researchers
Language: English
Sponsor: N/A
Authoring Institution: N/A