ERIC Number: ED052866
Record Type: Non-Journal
Publication Date: 1971-Feb
Reference Count: N/A
Analysis of Intelligence Scores.
Hines, Brainard W.
In September 1969, a random sample of 160 3- and 4-year-olds was selected from a larger group of individuals in 3 Early Childhood Education Program treatment groups (television only; television + home visitor; and television + home visitor + mobile classroom), along with a no-treatment comparison group, for the purpose of analyzing intelligence scores and congruence with program objectives. These children were tested in June and September 1970 by individuals trained by the Appalachia Educational Laboratory but not otherwise involved in the program. In this way, it was hoped that any examiner bias would be minimized and that which remained would be a constant factor throughout all treatment groups. Data from the 4 groups were analyzed by a 3-way analysis of variance (ANOVA) and an analysis of covariance (ANCOVA) procedure. The ANOVA procedure involved 4 treatment levels, 2 age levels, and 2 levels of sex; the ANCOVA used the "Peabody Picture Vocabulary Test" raw score and chronological age as covariates. The BMDX64 general linear hypothesis program for unbalanced design was used to run the ANOVA and ANCOVA for each variable. There was no need to test the homogeneity of variances since the program adjusts for any lack of homogeneity. In the document, graphs of significant interaction effects are presented for each subtest where such effects occurred, along with results of Scheffe post hoc comparisons. In the analysis of variance table for each subtest, the sum of squares column is replaced by a list of eta squared calculations. (LS)
Publication Type: N/A
Education Level: N/A
Authoring Institution: Appalachia Educational Lab., Charleston, WV.
Identifiers - Location: West Virginia