PHYSICAL FORMAT OF TEST INSTRUMENTS AND FIELD DEPENDENCE-INDEPENDENCE AS FACTORS IN ACHIEVEMENT TEST SCORES
The present study examined the effect of variation in the physical format of a standardized arithmetic test on the performance levels of elementary school students. The relationship of pupil sex and level of field dependence-independence (FD-FI) to the variations in test format was also assessed.^ Subjects were 77 fifth-grade students enrolled in regular classes in two suburban elementary schools. The sample consisted of 57 Black, 4 Hispanic and 16 white children.^ Two forms of the Wide Range Achievement Test (WRAT) were administered to all subjects in a test-retest experimental design with a 10-day interval between testings. Order of administration of the standard form (WRAT I) and the experimental form (WRAT II) was counterbalanced to correct for practice effects. WRAT II differed from WRAT I only in the spacing of the arithmetic problems on the page. The Children's Embedded-Figures Test, administered individually, was used as a measure of field dependence-independence. Standard scores for Total Math and Total Reading from the Metropolitan Achievement Tests were obtained for each subject.^ Preliminary analysis indicated the need for using Total Math scores as a covariate in the final analysis of the data. A 2 x 2 analysis of covariance was computed in a repeated measures design with sex of subject and level of FD-FI as the independent variables and WRAT scores as the dependent variables.^ Results indicated that the difference in physical format between WRAT I and WRAT II had produced significant differences in the scoring levels attained by this sample. Adjusted cell means indicated that all four groups (male, female, FD, and FI) had achieved higher scores on the experimental form of the WRAT. No significant main effects for sex of subject or level of FD-FI were observed, and there were no significant interactions. FD-FI was significantly correlated with both WRAT I and WRAT II before covariance of Total Math scores. There was also a significant main effect for FD-FI in relation to Total Math and Total Reading scores from the Metropolitan Achievement Tests in preliminary analysis of the data. The variable of sex of subject had been significantly correlated with FD-FI, but not with test format, in early data analysis. This relationship was maintained after computation of partial correlations controlling for Total Math, but was not strong enough to retain significance in the final analysis of the data.^ A major premise of this study, that changes in the physical format of a test instrument could produce significant changes in scoring levels, was clearly supported by the data. This is consistent with earlier research demonstrating such effects, and it extends the effect to a stadardized test used with normal populations. Since the present study involved only performance in arithmetic, there is a need for additional research testing the generalizability of the effect to other tests and other grade levels.^ A second premise of this study, that the response to changes in test format would be related to levels of FD-FI, was not clearly supported. Although FD-FI was significantly correlated with both WRAT forms in the preliminary analysis of the data, the significance was not maintained in the final analysis.^ A third premise of this study, that CEFT performance as well as performance on WRAT I and WRAT II would vary with sex of subject, was not supported. However, research idicates that student sex should be considered in studies using other tests of FD-FI, or in samples which contain substantial numbers of Black females. ^
Educational tests & measurements
GULOTTA, JANICE MARY, "PHYSICAL FORMAT OF TEST INSTRUMENTS AND FIELD DEPENDENCE-INDEPENDENCE AS FACTORS IN ACHIEVEMENT TEST SCORES" (1980). ETD Collection for Fordham University. AAI8020987.