The purpose of this study was to evaluate the Metropolitan Achievement Test, Primary I Battery (MAT-PI) for its early predictive validity in the later identification of mildly mentally handicapped students as well as its predictive validity in Identifying children who will be referred for an individualized psycho-educational assessment In the course of their elementary school career. All selected subjects (N=225) were students In the Gary Community School Corporation in Gary, Indiana, who had taken the MAT-PI during April of first grade, between the years 1981-83. Seventy-five subjects were randomly chosen from three categories; students who were subsequently designated as mildly mentally handicapped (MIMH, N= 429), students who were subsequently referred and evaluated as part of psycho-educational evaluation and who were not classified as mildly mentally handicapped (Referral Group, 1Y 974) and students who were not subsequently referred or evaluated as part of a psycho-educational evaluation nor classified under any special education classification (Normal, & -6,336). An additional 27 students from each group (MIMH, Referral and Normal) were retained to perform a cross validation. Discriminant Analysis was used with the selected sample (N=225) to predict group membership (MIMH, Referral and Normal) using standard scores from four subtests (Word Knowledge, Word Analysis, Reading and Math) of the MAT-PI as independent variables. A step-wise discriminant procedure was utilized with a minimum significance level of A <.05 for the two possible discriminant functions. A series of one-way analyses of variance examined the mean differences between the three groups (MIMH, Referral and Normal). A post-hoc analysis of mean differences between groups was done using the Scheffe test of significance (R <.05).Findings1. The results of the analysis of variance (one-way ANOVA's) indicated significant (p. <.01) mean differences between the three groups (MIMH, Referral and Normal) on each of the four predictor variables (MAT-PI subtests: Word Knowledge, Word Analysis, Reading and Math).2. The results of the discriminant analysis for both a three-group (MIMH, Referral and Normal) and a two-group (Referred and NonReferred) classification system, produced statistically significant (p <_.01) discriminant functions.3. The Word Knowledge, Word Analysis and Math subtests of the MAT-PI were the best predictors of group membership in both the three-group and the two-group classification systems.Conclusions1. The null hypothesis that the MAT-PI will not significantly discriminate students who will be identified as in need of special education services for the mildly mentally handicapped was rejected. 2. The null hypothesis that the MAT-PI will not significantly discriminate students who will subsequently be referred for a psycho-educatioanl evaluation was rejected.3. A comparison of these results with related studies substantiates that the MAT-PI has comparable predictive ability for early identification of "high risk" children to individually administered preschool screening tests, the use of biographical data, parent questionnaires or teacher ratings.4. The major problem with the MAT-PI, as an economical screening instrument, is its tendency to over-predict (false positives).
| Identifer | oai:union.ndltd.org:BSU/oai:cardinalscholar.bsu.edu:handle/180557 |
| Date | January 1986 |
| Creators | Schwartz, Joel B. |
| Contributors | Dean, Raymond S. |
| Source Sets | Ball State University |
| Detected Language | English |
| Format | 85 leaves ; 28 cm. |
| Source | Virtual Press |
Page generated in 0.0026 seconds