Alice E. Klein

The Validity of the Beery Test of Visual-Motor Integration in Predicting Achievement in Kindergarten, First, and Second Grades

The ability of the Visual-Motor Integration Test (VMI) to predict academic achievement was assessed by using three samples of pre-kindergarten pupils from three successive years in a large midwest school district. At the beginning of the kindergarten year, the Screening Test of Academic Readiness (STAR) was administered to all three samples. Obtained VMI-STAR correlations ranged from .102 to .595. At the beginning of the first grade, the Stanford Early School Achievement Test (SESAT), Primary Battery II, was administered to two samples of pupils. The VMI-SESAT correlations ranged from .271 to .498. At the beginning of the second grade, the Stanford Achievement Test (SAT), Primary Battery I, was administered to one sample of pupils. The VMI-SAT correlations ranged from .251 to .402. All of the 61 correlations were statistically significant. The VMI appeared to be most predictive of STAR Total and Numbers scores. SESAT Mathematics and Word Reading scores, and SAT Total Mathematics scores. Test-retest realiability, obtained on one sample over a time period of seven months, was .630.

The Validity of the Screening Test of Academic Readiness in Predicting Achievement in First and Second Grades.

The ability of the Screening Test of Academic Readiness (STAR) to predict academic achievement was assessed through using two samples of beginning kindergarten pupils from two successive years in a large midwest suburban school district. At the beginning of the first grade, the Stanford Early School Achievement Test (SESAT), Primary Battery II, was administered to both samples of pupils. Obtained STAR-SESAT correlations ranged from .087 to .700. In addition, the Stanford Achievement Test (SAT), Primary Battery I, was administered to one sample two years after the STAR administration, when the pupils were in the second grade. The STAR-SAT correlations ranged from .059 to .699. All but one of the 216 correlation coefficients were statistically significant. The total score on the STAR and the STAR Numbers subtest appeared to be the best predictors of academic achievement in both first and second grades.

A Cross-Validated Factor Analysis of the Nowicki-Strickland Locus of Control Scale for Children

The Nowicki-Strickland Locus of Control Scale for Children (N- SLOCSC) is a test of perception of internal vs. external control of reinforcement that was developed for use with children and youth. As part of a larger study, the scale was factor analyzed and then cross-validated with two similar samples of high school students. Al though the authors assume their scale to be unidimensional, no one strong dimension emerged from these analyses. Only 8 of the 21 items used in this study were correlated similarly on the two factors for both samples. The two resultant dimensions appeared to mea sure Social Control (six items) and Self Control (two items).

Further Evidence on the Redundancy of the Stanford Achievement Test

The latest revision of the Stanford Achievement Test (SAT), Primary Battery I, consists of seven subtests combined to yield three subtotal scores and one Total Battery score. As part of the ongoing evaluation of an early childhood preschool program, the SAT was administered to second-grade children. It was noted that the results appeared to be similar regardless of which SAT scores were used. After it had been determined that an older version of the SAT, Advanced, had been factor analyzed and found to be highly redundant, a factor analysis was performed on the SAT, I scores of 927 beginning second grade pupils. Results of the analysis showed one common factor with no evidence for the differential utility of the seven subtest or three subtotal scores generated by the SAT.

Redundancy in the Iowa Tests of Basic Skills

The latest revision of the Iowa Tests of Basic Skills (ITBS), Forms 7 and 8, consists of 11 subtests, nine of which are combined to yield three subtotal scores and one composite score. As part of one school districts testing program, the ITBS was administered to one group of beginning fourth grade students and one group of beginning eighth grade students. A factor analysis was performed for each of the two groups on Levels 10 and 14 of the ITBS. The intercorrelations provided by the 1977 fourth and eighth grade national standardization sample for these same levels were also factor analyzed. Results of all the analyses showed that there was essentially one common factor with little evidence that the user would gain much non-redundant information from the 15 scores yielded by the ITBS.

Redundancy in the Comprehensive Tests of Basic Skills

The Comprehensive Tests of Basic Skills (CTBS) Form S consists of 11 subtests, eight of which are combined to yield three subtotal scores and one Total Battery score. As part of the long term evalua tion of an early childhood preschool program, the CTBS was ad ministered to one sample of beginning fourth grade pupils and one sample of beginning sixth grade pupils. A factor analysis was per formed for the fourth grade sample on Level 1 and for the sixth grade sample on Level 2 of the CTBS. Results of both analyses showed but one common factor, with scant evidence for the useful ness of the eight different subtest or three subtotal scores produced by the CTBS.

Measures of Anomie and Locus-of-Control for Adolescents: Evidence of Factorial Validity.

Because of the lack of clarity of anomie and locus-of-control that has resulted from use of the same terminology in discussion of the two constructs, items from the Crumbaugh Purpose-in-Life Test (anomie) and the Nowicki-Strickland Locus-of-Control Scale for Children were submitted together for factor analyses. The first sample included 349 adolescents; the cross-validation sample contained 404 adolescents from the same population. Similar factor structures were found in both samples when principal axes were rotated to varimax solutions and to oblique solutions. No item from one instrument loaded above .30 on any factor characterized by items from the other instrument. It was concluded that the instruments measure different constructs and are factorially valid when considered together.

Test-Retest Reliability and Predictive Validity of the Northwestern Syntax Screening Test.

The Northwestern Syntax Screening Test (NSST), a measure that has been a source of considerable controversy since its publication, has also been one of the most widely used clinical screening tools of syntactic development. As part of the evaluation of a preschool in tervention program, the NSST was administered to approximately 700 pupils at the beginning and end of the program year. The NSST was found to have moderate test-retest reliability (.542 for the recep tive portion and .697 for the expressive portion). It was also found to be moderately accurate in predicting general academic achievement test scores of pupils in kindergarten and first grade.

The Reliability and Predictive Validity of the Slosson Intelligence Test for Pre-Kindergarten Pupils.

The ability of the Slosson Intelligence Test (SIT) to predict academic achievement was assessed by using three samples of pre-kindergarten pupils from three successive years in a large midwest school district. At the beginning of the kindergarten year, the Screening Test of Academic Readiness (STAR) was administered to all three samples. Obtained SIT-STAR correlations ranged from .184 to .689. At the beginning of the first grade, the Stanford Early School Achievement Test (SESAT), Primary Battery II, was administered to two samples of pupils. The SIT-SESAT correlations ranged from .296 to .653. At the beginning of the second grade, the Stanford Achievement Test (SAT), Primary Battery I, was administered to one sample of pupils. The SIT-SAT correlations ranged from .392 to .586. All of the 102 correlations were statistically significant. The SIT appeared to be most predictive of STAR Total and Numbers scores; of SESAT Mathematics, Word Reading, Aural Comprehension, and Environment scores; and of SAT Total Battery, Total Auditory, Vocabulary, and Mathematics Concepts scores. Test-retest reliabilities, obtained on all three samples over a time period of seven months, ranged from .689 to .778.

Some Evidence Concerning the Factorial Validity of the Screening Test of Academic Readiness

The Screening Test of Academic Readiness (STAR) consists of eight subtests combined to yield one Total score. As part of the ongoing evaluation of an early childhood education program, the STAR was administered to three samples of kindergarten pupils. It had been previously determined that the STAR was predictive of later academic achievement. A series of factor analyses was performed to determine whether the STAR could be used for individualized diagnosis and remediation. Results showed that two oblique factors emerged which had limited meaning. These results raised questions about which STAR scores might prove most useful to the potential consumer.