A factor analytic investigation of the Kindergarten Diagnostic Instrument
School psychology has been influenced by forces acting on the educational setting. Included are assumptions from developmental theory and the medical model of disease, need for accountability and measurement, efforts to identify problems early and intervene to prevent difficulties, and efforts to determine the readiness of individual students in a cost effective manner through use of screening devices.
This investigation sought findings relevant to the above by examining the construct validity of the Kindergarten Diagnostic Instrument (KDI). If a screening device can be proven valid, then definitions of readiness may be clarified, cost effective methods of providing accountability for early intervention may be established and theoretical questions relating to prekindergarten development may be approached.
Hundreds of published preschool screening devices are used in this country and other nations. Research indicates little consistency of practice in screening preschoolers or agreement regarding definitions of readiness. Age of school entrance varies widely despite its agree upon significance to readiness. Gender differences are inconsistently reported across researchers and instruments. No preschool screening instrument of proven reliability and validity with representative national norms currently exists.
The KDI was administered to a sample of 3,184 Ss from 4 to 6 years of age living in two Northeastern States. Individual administration occurred in one State and team administration in the other. Ss were assigned to 1 of 3 age groups and State samples were kept separate for analyses. Comparisons were made across location, gender and age. Differences between means were examined by independent samples t tests and ANOVAs with Welch and Brown-Forsythe corrections for unequal variances where appropriate. Principal components analyses with Varimax rotation were performed in exploratory and confirmatory fashion on each State sample, two gender subgroups and each of the 6 age subgroups.
Results indicated strong correspondence between test scores and age. Small but statistically significant differences between means were found on some subtests across location and gender with advantages to individual administration and females. A two factor solution was derived which accounted for 50% to 53% of the variance and proved robust across location, gender and age. Factors were designated Verbal and Visual and deemed similar to constructs extant in research on older children. Development of factor scores for the KDI was recommended given the robustness of the solution. A three factor solution was also derived consisting of Verbal, Visual-Motor and Visual-Cognitive constructs. Developmental implications were drawn from its appearance in older subgroups but not younger and in males but not females. Questions relating to the robustness of this solution across location, gender and age make it of great research interest but premature for clinical application.
It was concluded that the construct validity of the KDI, as measuring Verbal and Visual abilities, was supported by these findings. Further research is needed to confirm these findings with other samples, to explore predictive validity and to examine various implications of the 3 factor solution.