Perceptual and Motor Skills, 1979, 48, 1011-1021. @ Perceptual and Motor Skills 1979
LONGITUDINAL STUDY OF COGNITIVE ABILITIES AND ACADEMIC ACHIEVEMENT VIRGINIA L. BRUININKS University of Minnesota
AND
JEWEL H. MAYER St. Paul, Minnesota
Summary.-Assessment was made of the effectiveness of a battery of cognitive tests administered in kindergarten in the prediction of sixth grade achievement in reading comprehension, spelling, language usage, and arithmetic. Cognitive abilities included several factors of intelligence, visual percep tion and visual sequential memory, visual-motor integration, and auditory perception and auditory sequential memory. Measures of prior learning were also included. Subjects were 58 children in a suburban public school district. A criterion for predictive utility for correlation coefficients was established, and simple correlation coefficients for various kindergarten measures and sixth grade achievement ranged from the criterion of .35 to .69. In partial correlations with the effects of ability to understand ideas expressed in words removed, correlation coefficients for various cognitive measures and achievement tests ranged from the criterion of .35 to .63. Combinations of kindergarten measures having optimal multiple correlations with later school achievement generally approached or exceeded .70. These findings are discussed and suggestions are made for further research.
Preschool and kindergarten programs in the public schools are presently receiving increased attention and federal support, particularly early education programs for handicapped children. Better understanding is needed of the cognitive abilities that contribute to academic achievement so that more appropriate screening devices can be developed to help identify young children who may need special help in school and more effective early intervention programs can be provided. A plethora of correlational studies of cognitive abilities and reading achievement exists, leading in the aggregate to inconclusive results. Less informarion is available about correlates of arithmetic achievement, and few studies have investigated the relationships of cognitive abilities and other important areas of achievement. Relatively few studies have been longitudinal investigations, and most were concerned with correlates of kindergarten measures and achievemenr in the primary grades. The purposes of the present study were ( a ) to identify kindergarten abilities related to sixth grade achievement in reading comprehension, spelling, language usage, and various aspects of arithmetic achievement, and ( b ) to identify combinations of kindergarten measures having optimal multiple correlations with later school achievement. The kindergarten abilities investigated included several factors of intelligence, visual perception and visual sequential
1012
V. L BRUININKS & J. H.MAYER
memory, visual-motor integration, and auditory perception and auditory sequential memory. In addition, knowledge of letter names was included because this has been regarded as one of the best predictors of reading achievement (Barrett, 1965; Bond & Dykstra, 1967). METHOD Subjects Initially a group of 112 prereading boys and girls was randomly selected from a pool of 300 kindergarten children in a suburban school district. These subjects were selected to participate in a 3-yr. study of kindergarten predictors of word attack skills, word recognition skills, and reading comprehension (Clark, Bruininks, & Glaman, 1978). Mean chronological age of the group was 5 yr., 7 mo. (range = 5-0 to 6-2), and none of the subjects had poor visual or auditory acuity. The group was divided among four public elementary schools and two parochial schools for first grade. Subjects were 58 of these children ( 2 6 boys and 32 girls) who were currently in the sixth grade in the public schools. Average IQ for the present sample was 105 (SD = 10.13) as measured by the Primary Mental Abilities Test when they were in kindergarten. All were in the appropriate age/grade placement. Attrition from the original sample was due to the loss of subjects who attended parochial schools, to the usual mobility of students during a 6-yr. period, and perhaps to the retention of some subjects in the first five grades. The present sample is viewed as representative of a "normal population" of sixth grade students in a suburban public school district. Instruments and Procedures In kindergarten, each subject took a battery of tests designed to measure various abilities. The tests selected for use in the present study and their sources are listed in Table 1. All tests were administered individually to subjects by experienced examiners in a series of testing sessions lasting approximately 20 min. each. The Primary Mental Abilities Test (PMA) was administered first. As indicated in Table 1, the Spatial Relations test was divided into two parts with separate scores recorded for each. All other tests were administered according to test directions. In sixth grade, each subject took the Iowa Test of Basic Skills, Form 15, Level 12 (Lindquist & Hieronymus, 1971), which included the following subtests of interest to this study: Reading Comprehension, Spelling, Language Usage, Map Reading, Reading Graphs and Tables, Math Concepts, and Math Problem Solving. Scores on the Map Reading and Reading Graphs and Tables tests were combined. All tests were administered according to directions in a group-testing situation as part of the school district's regular testing program.
LONGITUDINAL STUDY : COGNITIVE ABILITY/ACHIEVEMENT
1013
TABLE 1
TESTSADMINISTEREDIN KINDERGARTBN Test or Subtest Verbal Meaning Perceptual Speed Number Facility Spatial Relations Part I-items involve selecting from a series of four options the drawing to complete the stimulus drawi ng Part II-items require the subject to draw missing parts on a figure to make the figure match a stimulus drawing Beginning Reading-Letter Recognition Embedded Figures Visual Closure Visual Sequential Memory Visual-motor Integration
Source Primary Mental Abilities Test, Grades K-1 (Thurstone & Thurstone, 1963)
Beginning Reading Test (Woodcock & Pfost.. 1967) - . Children's Embedded Figures Test (Karp & Konstadt, 1963) Visual Automatic Test (Kass, 1962) Illinois Test of Psycholinguistic Abilities (Kirk, McCarthy, & Kirk, 1968) Developmental Test of Visual-motor Integration (Beery & Buktenica,
1967) Auditory Discrimination-Quiet Auditory Discriminarion-Noise Auditory Attention Span for Related Svllables
Goldman-Fristoe-Woodcock Test of Auditory Discrimination (Goldman. Fristoe, & Woodcock, 1970)
Detroit Tests of Learnin Aptitude (Baker 8r Leland. 19f7)
RESULTS Relacion of Cognitive Abilities to Later Achievement Product-moment correlations were computed for kindergarten measures of cognitive abilities and sixth grade achievement measures. A criterion for predictive usefulness was needed, and .35 was used as the cut-off point between coefficients viewed as having practical predictive utility and those without. This criterion was used by Hammill and Larsen (1974) and Larsen and Hammill (1975) in reviews of studies using correlational scacistics. It represents a compromise between ( a ) the position taken by Guilford (1956) who suggested the predictive usefulness of coefficients ranging from .3 to .8, and ( b ) the position taken by Garrett (1954) who suggested that only coefficients of .4 or above were useful. Table 2 shows coefficients reaching cricerion of .35 for kindergarten measures of cognitive abilities and sixch grade achievement measures. To identify more precisely the cognitive abilities related to later academic
TABLE 2 CORRELATIONS A N D PARTIALCORRELATIONS OF KINDERGARTEN COGNI~VB TESTSWITHSIXTH GRADEACHIEVEMENT TESTS Kindergarten Measures of Cognitive Abilities
Reading Comprehension
r
~ab.c
Spelling-
r
~.b.e
Language Map Reading Usage and Reading nG&@@!&s tab.^
PblA Total 1Q .48 * .59 * .50 * Verbal Meaning .43 * .38 * .38 * Perceptual Speed Number Facility .55 .43 .54 t .54 .44 Spatial Relations Total .41 t Part I Parr I1 .40 t .35 t Beginning Reading-Letter Recognition .41 t .61 .52 .46 .36 Embedded Figures Visual Closure .41 t Visual Sequential Memory .35 t Visual-motor Integration Auditory Discrimination-Quiet .58 .46 .41 t Auditory Discrimination-Noise .35 t Auditory Attention Span for Related Syllables .52 .39 .50 .38 .44 t *Not included in the partial correlation analysis. t A relationship that failed to meet criterion of .35 when the effect of performance on Verbal
Mathematics ConProblem Solving C ~ P ~ S T
.52 .43
*
"
.59 .36
.58 .42
.49
.41 .47
.35 .37
.69 .49 .41 .41 .41
.40
t
.43 .40 .60
t
Jnb.~
r
Tnb.~
c
r
3
.44
* *
C
.63 .38
.55 .41
.48
cn
3
i
r
t
.37 .45
t t
.38
F
.45
.38
.46
.39
.35 .36 .40
.37
t t
t
.36
.36 .43
.51
.42
t
t
t
.39
Meaning was removed (Below Criterion).
R)
I
LONGITUDINAL STUDY : COGNITIVE ABILITY/ACHIEVEMENT
1015
success, parcial correlations were computed for relationships reaching the criterion of 3 5 . The effect of performance on Verbal Meaning was removed because vocabulary measures are thought to be highly related to general intelligence and achievement at all ages, and it was felt that a clearer relationship could be shown at kindergarten between a particular cognitive ability measure and the various sixth grade achievement tests when shared variance with VerbaI Meaning- was removed. Table 2 also shows partial correlations. Simple correlation coefficients for the various kindergarten measures and sixch grade achievement tests ranged from the criterion of .35 to .69. Kindergarten skills measured by che Perceptual Speed and Embedded Figures tests were not sufficiently related to sixch grade achievement for predictive utility. All other kindergarten tests showed predictive utility for at least one sixth grade achievement area. Kindergarten skills measured by the PMA Number Facility and Beginning Reading-Letter Recognition tescs showed a moderately high relationship to all sixth grade achievemenc tests. As expected, kindergarten Primary Mental Abilities IQ was sufficiently related to performance on all sixth grade achievement measures for useful prediction. With the effect of performance on Verbal Meaning removed, correlation coefficients for the various cognitive measures and achievemenc tests ranged from the criterion of .35 to .63. Performance on the first part of the PMA Spatial Relations Test and performance on the Visual Sequential Memory Test were no longer sufficiently related to sixth grade achievement for predictive utiliry, although both measures produced simple correlations of .35 or above. Kindergarten performance on the PMA Number Facility Test was still sufficiently related to later school achievement for predictive utiliry in all areas except Spelling. The relationships between kindergarten performance on Beginning Reading-Letter Recognition and the various achievement tests remained high enough for predictive utility in Spelling, Language Usage, Map Reading and Reading Graphs and Tables, and Math Problem Solving. Table 3 summarizes the results of the simple correlations and partial cotrelations. Beside each tesc of cognitive abilities administered in kindergarten appears the name of the sixth grade achievement area for which it appears to have some predictive usefulness. Five kindergarten tests were sufficiently related to sixth grade achiwement for predictive utility in Reading Comprehension, Spelling, and Language Usage. Four of these same kindergarten tests and five additional ones were sufficiently related co sixch grade achiwement in Map Reading and Reading Graphs and Tables, Math Concepts, and Math Problem Solving. Reading Comprehension in sixth grade was rehted to kindergarten performance on Verbal Meaning, Number Faciliry, Auditory Discrimination-Quiet, and Audirory Attention Span for Related Syllables. Spelling in sixth grade was related to kindergarten performance on Verbal Meaning,
V. L. BRUININKS & J. H. MAYER TABLE 3 SUMMARY OF KINDERGARTEN COGMTIVBTESTSAND THEIRRELATION FOR PREDICTIVE PURPOSESTO SIXTH GRADEACHIEVEMENT Kindergarten Measure of Cognitive Ability M Verbal Meaning
Perceptual Speed Number Facility
Spatial Relations Total Score Spatial Relations, I Spatial Relations, I1 Beginning Reading, Letter Recognition Embedded Figures Visual Closure Visual Sequential Memory Visual Motor Integration Auditory Discrimination, Quiet Auditory Discrimination, Noise Auditory Attention Span for Related Syllables
SD
Sixth Grade Achievement Measure
M
SD
Reading Comprehension Spelling Language Usage Map Reading and Reading Graphs and Tables Math Concepts None Reading Comprehension Language Usage Map Reading and Reading Graphs and Tables Math Concepts Math Problem Solving Math Concepts None Map Reading and Reading Graphs and Tables Spelling Language Usage Map Reading and Reading Graphs and Tables Math Problems Solving None Math Concepts Math Problem Solving None Math Concepts Reading Comprehension Math Concepts Math Problem Solving Reading Comprehension Spelling Map Reading and Reading Graphs and Tables
Beginning Reading-Letter Recognition, and Auditory Attention Span for Related Syllables. Language Usage in sixth grade was related to kinderganen performance on Verbal Meaning, Number Facility, and Beginning Reading-Letter Recognition. Map Reading and Reading Graphs and Tdbles in sixth grade was related to kindergarten performance on Verbal Meaning, Number Facility, Spatial Relations Part 11, Beginning Reading-Letter Recognition, and Auditory Attention Span for Related Syllables. Math Concepts in sixth grade was related to kindergarten performance on Verbal Meaning, Number Facility, Spatial Relations Total Score, Spatial Relations Part 11, Visual Closure, Visual Motor Integration, and Auditory Discrimination-Noise.
1017
LONGlTUDlNAL STUDY: COGNITIVE ABILlTY/ACHIEVEMENT TABLE 4 CUMULATIVE PREDICTION OF SIXTH GRADEACHIEVEMENT BY KINDERTARTEN MEASURES OF COGN~TIVE ABILITIES Achievement Measure Reading Comprehension (Approximately 49% of 2 )
Spelling (Approximately 45% of 2 )
Language Usage (Approximately 36% of 2 ) Map Reading and Reading GraphdTables (Approximately 55% of 2) Math Concepts (Approximately 55% of 2 )
Math Problem Solving (Approximately 42% of d)
Kinderearten Measure of Cognitive Abilitv
Cum. R
Auditory Discrimination-Quiet Measures the ability to distinguish the fine differences among various speech sounds in words of the English language by requiring the subject to select the appropriate response to an auditory stimulus from a series of pictures Number Facility Involves the ability to work with numbers, to handle simple quantitative problems rapidly and accurately. and to understand and recognize quantitative differences (uses an auditory stimulus and pictures in the response format) Visual Motor Integration Measures the ability to reproduce geometric forms which are presented visually Primary Mental Abilities Total IQ Includes measures on Verbal Meaning, Perceptual Speed, Number Facility, and Spatial Relations Total Score Auditory Attention Span for Related Syllables Involves short-term memory for sentences Auditory Discrimination-Noise Measures the ability to distinguish the fine differences among various speech sounds in words of the English language by requiring the subject to select the appropriate response to an auditory stimulus, which is presented against a background of noise (cafeteria sounds), from a series of pictures Number Facility* Auditory Attention Span for Related Syllables* Visual Closure Requires the subject to identify objects from incomplete drawings (visual stimulus and oral response) Auditory Attention Span for Related Syllables* Number Facility* Auditory Discrimination-Noise* Spatial Relations Part I1 Involves identifying missing parts on a figure and drawing them in so that the figure looks like a stimulus figure Number Facility* Auditory Discrimination-Noise' Spatial Relations Total Score Involves the ability to visualize incomplete objects and figures-in Part I selecting from a series of four options the drawing to complete the stimulus drawingin Part 11 identifying missing parts on a figure and drawing them in so that the figure looks like a stimulus figure Number Facility* Auditory Discrimination-Noise* Visual Closure*
.5 8
.68
.70 .59
.64 .67
.54 .58
.GO .60 .68
.72 .74
.69
.72 .74
.55 .62 .65
1018
V. L. BRUININKS & J. H. MAYER
Math Problem Solving in sixth grade was related to kindergarten performance on Number Facility, Beginning Reading-Letter Recognition, Visual Closure, and Auditory Discrimination-Noise.
Selection of Best Predictors Stepwise regression analyses were run to identify combinations of kindergarten cognitive measures having optimal multiple corre'lations with school achievement in sixth grade. Table 4 shows the results of these analyses. For each achievement area, kindergarten measures are listed in the order of their selection, and only those measures are shown which made a statistically significant independent contribution to the cumularivc relationship between kindergarten measures and sixth grade achievement. In general, tests that produced moderately high simple correlations with achievement were also represented in predictive clusters produced in the regression analyses. The kindergarten test of Visual Motor Integration, however, appears in Table 4 as useful for prediction of sixth grade Reading Comprehension when it failed to reach the .35 criterion for sufficient correlation with the achievement measure. In combination with other tests, it added significant independent variance for prediction in the regression analysis.
DISCUSSION Performance on measures which appear to directly involve prior learning (achievement) was sufficiently related to sixth grade achievement for predictive utility in a number of areas. The ability to work with numbers, to handle simple quantitative problems rapidly and accurately, and to understand and recognize quantitative differences in kindergarten (PMA Number Facility) was related to performance on all sixth grade achievement tests. The ability to understand ideas expressed in words in kindergarten (PMA Verbal Meaning) was related to performance in all sixth grade achievement areas except solving math problems. In addition, knowledge of letter names in kindergarten (as measured by Beginning Reading-Letter Recognition) was related to performance in all sixth grade achievement areas. The finding that past achievement is a strong predictor of fumre achievement is not a novel finding. For example, it was recently reported by Stevenson, Parker, Wildinson, Hegion, and Fish (1976) in their longitudinal study of individual differences in cognitive development and scholastic achievement which followed children from prekindergarten through third grade. Another means of interpreting the results of the present study is to examine the extent to which the various kindergarten measures stressed visual or auditory skills. Performance on tests of visual perception, visual sequential memory, and visual-motor integration was not generally sufficiently related to later school achievement for predictive utility. This is consistent with conclusions reached by Larsen and Hammill (1975) after review of a number of correla-
tional studies involving visual-perceptual abilities and school learning. There were, however, several notable exceptions. Visual closure on several tasks in kindergarten was related to sixth grade performance on both the Math Concepts and Math Problem Solving tests, and visual-mo:or incegration in kindergarten was related to performance on the Math Concepts test. As in all correlational relationships, however, there is the possibility that in each of these cases the abilities being measured were strongly related to a third unidentified factor or ability. Several kindergarten tests of auditory skills were sufficiently related to sixth grade achievement for predictive utility. Auditory sequential memory was related to sixth grade achievemenc on Reading Comprehension, Spelling, and Map Reading and Reading Graphs and Tables tests. The ability to distinguish the fine differences among speech sounds in English words in kindergarten was also related to Reading Comprehension in sixth grade. Finally, performance on the Auditory Discrimination-Noise test (which involves the ability to distinguish fine differences among speech sounds in English words when the stimulus is presented against a background of noise) was related to sixch grade performance on both the Math Concepts and Math Problem Solving tests. It is not clear if Auditory Discrimination-Noise measures auditory figure-ground discrimination, attention, or some other factor. Further research on this issue is important, since this test also added significant independent variance to multiple correlations for prediction of sixch grade performance on the Spelling, Map Reading and Reading Graphs and Tables, Math Concepts, and Math Problem Solving tests. It should be noted that many of the predictive measures involved integration of perceptual as well as perceptual-motor abilities. Auditory Discrimination-Noise, Auditory Discrimination-Quiet, and several other kindergarten measures which were generally related to sixth grade achievement could be viewed as visual/auditory-motor tasks. This means that a visual stimulus was presented to the subject and verbal task directions given (auditory stimulus), to which the subject then responded by marking [he visual stimulus material (motor response). For example, in the Number Facility tesc, subjects were provided visual stimulus material and told, "Slide your marker down ( t o the row of pictures for the item). Mark the largest dog, or mark five cups-five." For the Verbal Meaning test, subjects were provided visual stimulus material and told, "Slide your marker down under the next row of pictures. Mark rhe cat, or mark the row of evergreen trees." The integration of visual-auditory stimuli involved in performance of these tasks is a factor that warrants further research. It appears to have been predictive of achievemenc in a number of areas in the present study, and although the measurement devices were different, Stevenson, Parker, Wilkinson, Hegion, and Fish ( 1776) found visual-auditory
1020
V. L. BRUININKS & J. H. MAYER
association related to reading and arithmetic achievement in their longitudinal study. It was possible in the present study to identify several sets of tests of cognitive abilities which appear to have some promise for prediction of academic achievement in reading comprehension, spelling, language usage, and various aspects of arithmetic achievement. The variance explained in sixth grade achievement by some of the kindergarten predictors was rather high (often approximating 40 to 50%), especially in light of the length of time between administration of tests. Multiple correlations of early predictors with achievement generally approached or exceeded .70. It is very important, however, that these findings be cross-validated with a large independent sample of children because the multiple regression procedure used to identify predictors capitalizes on all chance relationships. With a larger sample size, the predictive accuracy of the measures could also be evaluated to see how well they would have identified children who actually did have difficulty later in school. In addition, for each academic area, there is need to determine just how the kindergarten measures contribute to the prediction of later achievement. For example, it might be that a high score on one predictor would offset a low score on another predictor, or perhaps some minimal level of performance is required on all predictors for achievement. Further work, however, will perhaps lead to the development of appropriate screening devices for children who may need special help in school and to more effective early intervention programs for them. '
REFERENCES BAKER,H. J., & LELAND,B. Detroit Tests of Learning Aptitude; examine7's handbook. Indianapolis, Ind.: Bobbs-Merrill, 1967. BARRE~T,T. Predicting reading achievement through readiness tests. In J. A. Figure1 (Ed.), Reading and inquiry, proceedings of the International Reading As~ociation, 1965, 10, 26-28. BEERY,K. E., & B U K ~ N I C A N., A. Developmental Test of Visual-motor Integration. Chicago: Follett, 1967. BOND, G. L.. & DYKSTRA. R. The cooperative research program in first-grade reading instruction. Reading Research Quarterly, 1967, 2, 5-142. CLARK,C. R., BRUININKS, R. H., & GLAMAN,G. V. Kindergarten predictors of three aspects of reading achievement. Perceptual and Motor Skills, 1978, 46, 411-419. GARRETT,H. E Statistics in psychology and education. New York: Longmans Green, 1954. GOLDMAN,R., FRISTOE, M., & WOODCOCK. R. W. The Goldman-Fristoe-Woodcock Test of Auditory Discrimination. Circle Pines, Minn.: American Guidance Service, 1970. GUILPORD, J. P. Fundamental statistics in psychology and education. New York: McGraw-Hill, 1956. HAMMILL, D. D., & LARSEN, S. C. The relationship of selected auditory perceptual skills and reading ability. ]ournal o f Learning Disabilities, 1974, 7, 429-436. KARP, S. A., & KONSTADT. N . W. Manual for the Children's Embedded Figures Test. New York: Cognitive Tests, 1963.
LONGITUDINAL STUDY: COGNITIVE ABILITY/ACHIEVEMENT
1021
KASS, C. E. Some psychological correlates of severe reading disability. Unpublished doctoral dissertation, Univer. of Illinois, 1962. KIRK, S. A.. McCARTHY, J. J., & KIRK, W. Examiner's manual: Illinois Tert of Psycholinguistic Abilities. (Rev. ed.) Urbana, IL: Univer. of Illinois Press, 1968. LARSEN,S. C., & HAMMILL,D. D. The relationship of selected visual-perceptual abilities to school learning. Journal o f Special Education, 1975, 9. 281-291. LINDQUIST,E. F., & HIERONYMUS,A. N. Iowa Test of Basic Skills. (Rev. ed.) New York: Houghton Mifflin, 1971. STEVENSON, H. W., PARKER,T., WILKINSON,A., HEGION,A., & FISH, E. Longitudinal study of individual differences in co nitive development and scholastic achievement. Journal of Educdional Psychofogy, 1976, 68, 377-400. THURSTONE,L. L., & THURSTONE,T. G. Examiner's manual: Primary Mental Abilities for Grades K - 1 . Chicago: Science Research Assoc., 1963. WOODCOCK,R. W., & PFOST, P. The Beginning Reading Tert. ( E x p . edition) Circle Pines, Minn.: American Guidance Service, 1967. Accepted April 17, 1979.
LONGITUDINAL STUDY OF COGNITIVE ABILITIES AND ACADEMIC ACHIEVEMENT VIRGINIA L. BRUININKS University of Minnesota
AND
JEWEL H. MAYER St. Paul, Minnesota
Summary.-Assessment was made of the effectiveness of a battery of cognitive tests administered in kindergarten in the prediction of sixth grade achievement in reading comprehension, spelling, language usage, and arithmetic. Cognitive abilities included several factors of intelligence, visual percep tion and visual sequential memory, visual-motor integration, and auditory perception and auditory sequential memory. Measures of prior learning were also included. Subjects were 58 children in a suburban public school district. A criterion for predictive utility for correlation coefficients was established, and simple correlation coefficients for various kindergarten measures and sixth grade achievement ranged from the criterion of .35 to .69. In partial correlations with the effects of ability to understand ideas expressed in words removed, correlation coefficients for various cognitive measures and achievement tests ranged from the criterion of .35 to .63. Combinations of kindergarten measures having optimal multiple correlations with later school achievement generally approached or exceeded .70. These findings are discussed and suggestions are made for further research.
Preschool and kindergarten programs in the public schools are presently receiving increased attention and federal support, particularly early education programs for handicapped children. Better understanding is needed of the cognitive abilities that contribute to academic achievement so that more appropriate screening devices can be developed to help identify young children who may need special help in school and more effective early intervention programs can be provided. A plethora of correlational studies of cognitive abilities and reading achievement exists, leading in the aggregate to inconclusive results. Less informarion is available about correlates of arithmetic achievement, and few studies have investigated the relationships of cognitive abilities and other important areas of achievement. Relatively few studies have been longitudinal investigations, and most were concerned with correlates of kindergarten measures and achievemenr in the primary grades. The purposes of the present study were ( a ) to identify kindergarten abilities related to sixth grade achievement in reading comprehension, spelling, language usage, and various aspects of arithmetic achievement, and ( b ) to identify combinations of kindergarten measures having optimal multiple correlations with later school achievement. The kindergarten abilities investigated included several factors of intelligence, visual perception and visual sequential
1012
V. L BRUININKS & J. H.MAYER
memory, visual-motor integration, and auditory perception and auditory sequential memory. In addition, knowledge of letter names was included because this has been regarded as one of the best predictors of reading achievement (Barrett, 1965; Bond & Dykstra, 1967). METHOD Subjects Initially a group of 112 prereading boys and girls was randomly selected from a pool of 300 kindergarten children in a suburban school district. These subjects were selected to participate in a 3-yr. study of kindergarten predictors of word attack skills, word recognition skills, and reading comprehension (Clark, Bruininks, & Glaman, 1978). Mean chronological age of the group was 5 yr., 7 mo. (range = 5-0 to 6-2), and none of the subjects had poor visual or auditory acuity. The group was divided among four public elementary schools and two parochial schools for first grade. Subjects were 58 of these children ( 2 6 boys and 32 girls) who were currently in the sixth grade in the public schools. Average IQ for the present sample was 105 (SD = 10.13) as measured by the Primary Mental Abilities Test when they were in kindergarten. All were in the appropriate age/grade placement. Attrition from the original sample was due to the loss of subjects who attended parochial schools, to the usual mobility of students during a 6-yr. period, and perhaps to the retention of some subjects in the first five grades. The present sample is viewed as representative of a "normal population" of sixth grade students in a suburban public school district. Instruments and Procedures In kindergarten, each subject took a battery of tests designed to measure various abilities. The tests selected for use in the present study and their sources are listed in Table 1. All tests were administered individually to subjects by experienced examiners in a series of testing sessions lasting approximately 20 min. each. The Primary Mental Abilities Test (PMA) was administered first. As indicated in Table 1, the Spatial Relations test was divided into two parts with separate scores recorded for each. All other tests were administered according to test directions. In sixth grade, each subject took the Iowa Test of Basic Skills, Form 15, Level 12 (Lindquist & Hieronymus, 1971), which included the following subtests of interest to this study: Reading Comprehension, Spelling, Language Usage, Map Reading, Reading Graphs and Tables, Math Concepts, and Math Problem Solving. Scores on the Map Reading and Reading Graphs and Tables tests were combined. All tests were administered according to directions in a group-testing situation as part of the school district's regular testing program.
LONGITUDINAL STUDY : COGNITIVE ABILITY/ACHIEVEMENT
1013
TABLE 1
TESTSADMINISTEREDIN KINDERGARTBN Test or Subtest Verbal Meaning Perceptual Speed Number Facility Spatial Relations Part I-items involve selecting from a series of four options the drawing to complete the stimulus drawi ng Part II-items require the subject to draw missing parts on a figure to make the figure match a stimulus drawing Beginning Reading-Letter Recognition Embedded Figures Visual Closure Visual Sequential Memory Visual-motor Integration
Source Primary Mental Abilities Test, Grades K-1 (Thurstone & Thurstone, 1963)
Beginning Reading Test (Woodcock & Pfost.. 1967) - . Children's Embedded Figures Test (Karp & Konstadt, 1963) Visual Automatic Test (Kass, 1962) Illinois Test of Psycholinguistic Abilities (Kirk, McCarthy, & Kirk, 1968) Developmental Test of Visual-motor Integration (Beery & Buktenica,
1967) Auditory Discrimination-Quiet Auditory Discriminarion-Noise Auditory Attention Span for Related Svllables
Goldman-Fristoe-Woodcock Test of Auditory Discrimination (Goldman. Fristoe, & Woodcock, 1970)
Detroit Tests of Learnin Aptitude (Baker 8r Leland. 19f7)
RESULTS Relacion of Cognitive Abilities to Later Achievement Product-moment correlations were computed for kindergarten measures of cognitive abilities and sixth grade achievement measures. A criterion for predictive usefulness was needed, and .35 was used as the cut-off point between coefficients viewed as having practical predictive utility and those without. This criterion was used by Hammill and Larsen (1974) and Larsen and Hammill (1975) in reviews of studies using correlational scacistics. It represents a compromise between ( a ) the position taken by Guilford (1956) who suggested the predictive usefulness of coefficients ranging from .3 to .8, and ( b ) the position taken by Garrett (1954) who suggested that only coefficients of .4 or above were useful. Table 2 shows coefficients reaching cricerion of .35 for kindergarten measures of cognitive abilities and sixch grade achievement measures. To identify more precisely the cognitive abilities related to later academic
TABLE 2 CORRELATIONS A N D PARTIALCORRELATIONS OF KINDERGARTEN COGNI~VB TESTSWITHSIXTH GRADEACHIEVEMENT TESTS Kindergarten Measures of Cognitive Abilities
Reading Comprehension
r
~ab.c
Spelling-
r
~.b.e
Language Map Reading Usage and Reading nG&@@!&s tab.^
PblA Total 1Q .48 * .59 * .50 * Verbal Meaning .43 * .38 * .38 * Perceptual Speed Number Facility .55 .43 .54 t .54 .44 Spatial Relations Total .41 t Part I Parr I1 .40 t .35 t Beginning Reading-Letter Recognition .41 t .61 .52 .46 .36 Embedded Figures Visual Closure .41 t Visual Sequential Memory .35 t Visual-motor Integration Auditory Discrimination-Quiet .58 .46 .41 t Auditory Discrimination-Noise .35 t Auditory Attention Span for Related Syllables .52 .39 .50 .38 .44 t *Not included in the partial correlation analysis. t A relationship that failed to meet criterion of .35 when the effect of performance on Verbal
Mathematics ConProblem Solving C ~ P ~ S T
.52 .43
*
"
.59 .36
.58 .42
.49
.41 .47
.35 .37
.69 .49 .41 .41 .41
.40
t
.43 .40 .60
t
Jnb.~
r
Tnb.~
c
r
3
.44
* *
C
.63 .38
.55 .41
.48
cn
3
i
r
t
.37 .45
t t
.38
F
.45
.38
.46
.39
.35 .36 .40
.37
t t
t
.36
.36 .43
.51
.42
t
t
t
.39
Meaning was removed (Below Criterion).
R)
I
LONGITUDINAL STUDY : COGNITIVE ABILITY/ACHIEVEMENT
1015
success, parcial correlations were computed for relationships reaching the criterion of 3 5 . The effect of performance on Verbal Meaning was removed because vocabulary measures are thought to be highly related to general intelligence and achievement at all ages, and it was felt that a clearer relationship could be shown at kindergarten between a particular cognitive ability measure and the various sixth grade achievement tests when shared variance with VerbaI Meaning- was removed. Table 2 also shows partial correlations. Simple correlation coefficients for the various kindergarten measures and sixch grade achievement tests ranged from the criterion of .35 to .69. Kindergarten skills measured by che Perceptual Speed and Embedded Figures tests were not sufficiently related to sixch grade achievement for predictive utility. All other kindergarten tests showed predictive utility for at least one sixth grade achievement area. Kindergarten skills measured by the PMA Number Facility and Beginning Reading-Letter Recognition tescs showed a moderately high relationship to all sixth grade achievemenc tests. As expected, kindergarten Primary Mental Abilities IQ was sufficiently related to performance on all sixth grade achievement measures for useful prediction. With the effect of performance on Verbal Meaning removed, correlation coefficients for the various cognitive measures and achievemenc tests ranged from the criterion of .35 to .63. Performance on the first part of the PMA Spatial Relations Test and performance on the Visual Sequential Memory Test were no longer sufficiently related to sixth grade achievement for predictive utiliry, although both measures produced simple correlations of .35 or above. Kindergarten performance on the PMA Number Facility Test was still sufficiently related to later school achievement for predictive utiliry in all areas except Spelling. The relationships between kindergarten performance on Beginning Reading-Letter Recognition and the various achievement tests remained high enough for predictive utility in Spelling, Language Usage, Map Reading and Reading Graphs and Tables, and Math Problem Solving. Table 3 summarizes the results of the simple correlations and partial cotrelations. Beside each tesc of cognitive abilities administered in kindergarten appears the name of the sixth grade achievement area for which it appears to have some predictive usefulness. Five kindergarten tests were sufficiently related to sixth grade achiwement for predictive utility in Reading Comprehension, Spelling, and Language Usage. Four of these same kindergarten tests and five additional ones were sufficiently related co sixch grade achiwement in Map Reading and Reading Graphs and Tables, Math Concepts, and Math Problem Solving. Reading Comprehension in sixth grade was rehted to kindergarten performance on Verbal Meaning, Number Faciliry, Auditory Discrimination-Quiet, and Audirory Attention Span for Related Syllables. Spelling in sixth grade was related to kindergarten performance on Verbal Meaning,
V. L. BRUININKS & J. H. MAYER TABLE 3 SUMMARY OF KINDERGARTEN COGMTIVBTESTSAND THEIRRELATION FOR PREDICTIVE PURPOSESTO SIXTH GRADEACHIEVEMENT Kindergarten Measure of Cognitive Ability M Verbal Meaning
Perceptual Speed Number Facility
Spatial Relations Total Score Spatial Relations, I Spatial Relations, I1 Beginning Reading, Letter Recognition Embedded Figures Visual Closure Visual Sequential Memory Visual Motor Integration Auditory Discrimination, Quiet Auditory Discrimination, Noise Auditory Attention Span for Related Syllables
SD
Sixth Grade Achievement Measure
M
SD
Reading Comprehension Spelling Language Usage Map Reading and Reading Graphs and Tables Math Concepts None Reading Comprehension Language Usage Map Reading and Reading Graphs and Tables Math Concepts Math Problem Solving Math Concepts None Map Reading and Reading Graphs and Tables Spelling Language Usage Map Reading and Reading Graphs and Tables Math Problems Solving None Math Concepts Math Problem Solving None Math Concepts Reading Comprehension Math Concepts Math Problem Solving Reading Comprehension Spelling Map Reading and Reading Graphs and Tables
Beginning Reading-Letter Recognition, and Auditory Attention Span for Related Syllables. Language Usage in sixth grade was related to kinderganen performance on Verbal Meaning, Number Facility, and Beginning Reading-Letter Recognition. Map Reading and Reading Graphs and Tdbles in sixth grade was related to kindergarten performance on Verbal Meaning, Number Facility, Spatial Relations Part 11, Beginning Reading-Letter Recognition, and Auditory Attention Span for Related Syllables. Math Concepts in sixth grade was related to kindergarten performance on Verbal Meaning, Number Facility, Spatial Relations Total Score, Spatial Relations Part 11, Visual Closure, Visual Motor Integration, and Auditory Discrimination-Noise.
1017
LONGlTUDlNAL STUDY: COGNITIVE ABILlTY/ACHIEVEMENT TABLE 4 CUMULATIVE PREDICTION OF SIXTH GRADEACHIEVEMENT BY KINDERTARTEN MEASURES OF COGN~TIVE ABILITIES Achievement Measure Reading Comprehension (Approximately 49% of 2 )
Spelling (Approximately 45% of 2 )
Language Usage (Approximately 36% of 2 ) Map Reading and Reading GraphdTables (Approximately 55% of 2) Math Concepts (Approximately 55% of 2 )
Math Problem Solving (Approximately 42% of d)
Kinderearten Measure of Cognitive Abilitv
Cum. R
Auditory Discrimination-Quiet Measures the ability to distinguish the fine differences among various speech sounds in words of the English language by requiring the subject to select the appropriate response to an auditory stimulus from a series of pictures Number Facility Involves the ability to work with numbers, to handle simple quantitative problems rapidly and accurately. and to understand and recognize quantitative differences (uses an auditory stimulus and pictures in the response format) Visual Motor Integration Measures the ability to reproduce geometric forms which are presented visually Primary Mental Abilities Total IQ Includes measures on Verbal Meaning, Perceptual Speed, Number Facility, and Spatial Relations Total Score Auditory Attention Span for Related Syllables Involves short-term memory for sentences Auditory Discrimination-Noise Measures the ability to distinguish the fine differences among various speech sounds in words of the English language by requiring the subject to select the appropriate response to an auditory stimulus, which is presented against a background of noise (cafeteria sounds), from a series of pictures Number Facility* Auditory Attention Span for Related Syllables* Visual Closure Requires the subject to identify objects from incomplete drawings (visual stimulus and oral response) Auditory Attention Span for Related Syllables* Number Facility* Auditory Discrimination-Noise* Spatial Relations Part I1 Involves identifying missing parts on a figure and drawing them in so that the figure looks like a stimulus figure Number Facility* Auditory Discrimination-Noise' Spatial Relations Total Score Involves the ability to visualize incomplete objects and figures-in Part I selecting from a series of four options the drawing to complete the stimulus drawingin Part 11 identifying missing parts on a figure and drawing them in so that the figure looks like a stimulus figure Number Facility* Auditory Discrimination-Noise* Visual Closure*
.5 8
.68
.70 .59
.64 .67
.54 .58
.GO .60 .68
.72 .74
.69
.72 .74
.55 .62 .65
1018
V. L. BRUININKS & J. H. MAYER
Math Problem Solving in sixth grade was related to kindergarten performance on Number Facility, Beginning Reading-Letter Recognition, Visual Closure, and Auditory Discrimination-Noise.
Selection of Best Predictors Stepwise regression analyses were run to identify combinations of kindergarten cognitive measures having optimal multiple corre'lations with school achievement in sixth grade. Table 4 shows the results of these analyses. For each achievement area, kindergarten measures are listed in the order of their selection, and only those measures are shown which made a statistically significant independent contribution to the cumularivc relationship between kindergarten measures and sixth grade achievement. In general, tests that produced moderately high simple correlations with achievement were also represented in predictive clusters produced in the regression analyses. The kindergarten test of Visual Motor Integration, however, appears in Table 4 as useful for prediction of sixth grade Reading Comprehension when it failed to reach the .35 criterion for sufficient correlation with the achievement measure. In combination with other tests, it added significant independent variance for prediction in the regression analysis.
DISCUSSION Performance on measures which appear to directly involve prior learning (achievement) was sufficiently related to sixth grade achievement for predictive utility in a number of areas. The ability to work with numbers, to handle simple quantitative problems rapidly and accurately, and to understand and recognize quantitative differences in kindergarten (PMA Number Facility) was related to performance on all sixth grade achievement tests. The ability to understand ideas expressed in words in kindergarten (PMA Verbal Meaning) was related to performance in all sixth grade achievement areas except solving math problems. In addition, knowledge of letter names in kindergarten (as measured by Beginning Reading-Letter Recognition) was related to performance in all sixth grade achievement areas. The finding that past achievement is a strong predictor of fumre achievement is not a novel finding. For example, it was recently reported by Stevenson, Parker, Wildinson, Hegion, and Fish (1976) in their longitudinal study of individual differences in cognitive development and scholastic achievement which followed children from prekindergarten through third grade. Another means of interpreting the results of the present study is to examine the extent to which the various kindergarten measures stressed visual or auditory skills. Performance on tests of visual perception, visual sequential memory, and visual-motor integration was not generally sufficiently related to later school achievement for predictive utility. This is consistent with conclusions reached by Larsen and Hammill (1975) after review of a number of correla-
tional studies involving visual-perceptual abilities and school learning. There were, however, several notable exceptions. Visual closure on several tasks in kindergarten was related to sixth grade performance on both the Math Concepts and Math Problem Solving tests, and visual-mo:or incegration in kindergarten was related to performance on the Math Concepts test. As in all correlational relationships, however, there is the possibility that in each of these cases the abilities being measured were strongly related to a third unidentified factor or ability. Several kindergarten tests of auditory skills were sufficiently related to sixth grade achievement for predictive utility. Auditory sequential memory was related to sixth grade achievemenc on Reading Comprehension, Spelling, and Map Reading and Reading Graphs and Tables tests. The ability to distinguish the fine differences among speech sounds in English words in kindergarten was also related to Reading Comprehension in sixth grade. Finally, performance on the Auditory Discrimination-Noise test (which involves the ability to distinguish fine differences among speech sounds in English words when the stimulus is presented against a background of noise) was related to sixch grade performance on both the Math Concepts and Math Problem Solving tests. It is not clear if Auditory Discrimination-Noise measures auditory figure-ground discrimination, attention, or some other factor. Further research on this issue is important, since this test also added significant independent variance to multiple correlations for prediction of sixch grade performance on the Spelling, Map Reading and Reading Graphs and Tables, Math Concepts, and Math Problem Solving tests. It should be noted that many of the predictive measures involved integration of perceptual as well as perceptual-motor abilities. Auditory Discrimination-Noise, Auditory Discrimination-Quiet, and several other kindergarten measures which were generally related to sixth grade achievement could be viewed as visual/auditory-motor tasks. This means that a visual stimulus was presented to the subject and verbal task directions given (auditory stimulus), to which the subject then responded by marking [he visual stimulus material (motor response). For example, in the Number Facility tesc, subjects were provided visual stimulus material and told, "Slide your marker down ( t o the row of pictures for the item). Mark the largest dog, or mark five cups-five." For the Verbal Meaning test, subjects were provided visual stimulus material and told, "Slide your marker down under the next row of pictures. Mark rhe cat, or mark the row of evergreen trees." The integration of visual-auditory stimuli involved in performance of these tasks is a factor that warrants further research. It appears to have been predictive of achievemenc in a number of areas in the present study, and although the measurement devices were different, Stevenson, Parker, Wilkinson, Hegion, and Fish ( 1776) found visual-auditory
1020
V. L. BRUININKS & J. H. MAYER
association related to reading and arithmetic achievement in their longitudinal study. It was possible in the present study to identify several sets of tests of cognitive abilities which appear to have some promise for prediction of academic achievement in reading comprehension, spelling, language usage, and various aspects of arithmetic achievement. The variance explained in sixth grade achievement by some of the kindergarten predictors was rather high (often approximating 40 to 50%), especially in light of the length of time between administration of tests. Multiple correlations of early predictors with achievement generally approached or exceeded .70. It is very important, however, that these findings be cross-validated with a large independent sample of children because the multiple regression procedure used to identify predictors capitalizes on all chance relationships. With a larger sample size, the predictive accuracy of the measures could also be evaluated to see how well they would have identified children who actually did have difficulty later in school. In addition, for each academic area, there is need to determine just how the kindergarten measures contribute to the prediction of later achievement. For example, it might be that a high score on one predictor would offset a low score on another predictor, or perhaps some minimal level of performance is required on all predictors for achievement. Further work, however, will perhaps lead to the development of appropriate screening devices for children who may need special help in school and to more effective early intervention programs for them. '
REFERENCES BAKER,H. J., & LELAND,B. Detroit Tests of Learning Aptitude; examine7's handbook. Indianapolis, Ind.: Bobbs-Merrill, 1967. BARRE~T,T. Predicting reading achievement through readiness tests. In J. A. Figure1 (Ed.), Reading and inquiry, proceedings of the International Reading As~ociation, 1965, 10, 26-28. BEERY,K. E., & B U K ~ N I C A N., A. Developmental Test of Visual-motor Integration. Chicago: Follett, 1967. BOND, G. L.. & DYKSTRA. R. The cooperative research program in first-grade reading instruction. Reading Research Quarterly, 1967, 2, 5-142. CLARK,C. R., BRUININKS, R. H., & GLAMAN,G. V. Kindergarten predictors of three aspects of reading achievement. Perceptual and Motor Skills, 1978, 46, 411-419. GARRETT,H. E Statistics in psychology and education. New York: Longmans Green, 1954. GOLDMAN,R., FRISTOE, M., & WOODCOCK. R. W. The Goldman-Fristoe-Woodcock Test of Auditory Discrimination. Circle Pines, Minn.: American Guidance Service, 1970. GUILPORD, J. P. Fundamental statistics in psychology and education. New York: McGraw-Hill, 1956. HAMMILL, D. D., & LARSEN, S. C. The relationship of selected auditory perceptual skills and reading ability. ]ournal o f Learning Disabilities, 1974, 7, 429-436. KARP, S. A., & KONSTADT. N . W. Manual for the Children's Embedded Figures Test. New York: Cognitive Tests, 1963.
LONGITUDINAL STUDY: COGNITIVE ABILITY/ACHIEVEMENT
1021
KASS, C. E. Some psychological correlates of severe reading disability. Unpublished doctoral dissertation, Univer. of Illinois, 1962. KIRK, S. A.. McCARTHY, J. J., & KIRK, W. Examiner's manual: Illinois Tert of Psycholinguistic Abilities. (Rev. ed.) Urbana, IL: Univer. of Illinois Press, 1968. LARSEN,S. C., & HAMMILL,D. D. The relationship of selected visual-perceptual abilities to school learning. Journal o f Special Education, 1975, 9. 281-291. LINDQUIST,E. F., & HIERONYMUS,A. N. Iowa Test of Basic Skills. (Rev. ed.) New York: Houghton Mifflin, 1971. STEVENSON, H. W., PARKER,T., WILKINSON,A., HEGION,A., & FISH, E. Longitudinal study of individual differences in co nitive development and scholastic achievement. Journal of Educdional Psychofogy, 1976, 68, 377-400. THURSTONE,L. L., & THURSTONE,T. G. Examiner's manual: Primary Mental Abilities for Grades K - 1 . Chicago: Science Research Assoc., 1963. WOODCOCK,R. W., & PFOST, P. The Beginning Reading Tert. ( E x p . edition) Circle Pines, Minn.: American Guidance Service, 1967. Accepted April 17, 1979.
