ABSTRACT
This short note emphasizes the need for multivariate analysis when multiple correlated dependent variables are used in a study. The use of multivariate analysis and the consequences of not using it are illustrated in relation to a previously published study that used self-concept subscales as dependent variables.
Subject(s)
Child, Gifted/psychology , Personality Development , Self Concept , Child , Humans , Intellectual Disability/psychology , Multivariate Analysis , Research DesignABSTRACT
The present study investigated the congruence between the factor structures of the learning environment and cognitive variables for mathematics and English courses in Grade 10. Three common factors were obtained in each course. Procedural characteristics and cognitive factors in mathematics were highly similar to the corresponding factors in English. However, the formative characteristics factor identified in mathematics was not congruent to the corresponding factor in English. Learning environment variables produced two factors. These factors resembled the congruent factor properties for the two courses when the learning environment and cognitive variables were factor analyzed together. Implications for learning and instructional research are mentioned.
Subject(s)
Achievement , Learning , Psychology, Adolescent , Social Environment , Adolescent , Female , Humans , Male , Mathematics , Personality Inventory , ReadingABSTRACT
Differences in social-psychological environments between parochial and public schools, and between mathematics and English courses, were investigated for 317 boys and girls in Grade 10. Multivariate and univariate analyses of variance produced significant main effects and an interaction of school x subject. Differential validities of environmental variables in predicting achievement using multiple regression suggested that perception of the learning environment for course content substantially predicted achievement in that course content. Incremental validity of the learning environment with achievement when IQ was partialled out was also substantial. Implications for instructional effectiveness and school organization are discussed.
Subject(s)
Achievement , Cognition , Learning , Social Environment , Adolescent , Educational Measurement , Female , Humans , Intelligence Tests , MaleABSTRACT
EEG correlates of intellectual functioning were investigated in 35 academically talented and 20 academically handicapped adolescents. EEG data from Ertl's Brainwave Analyzer were factor-analyzed and factor scores entered into regression equations predicting four primary mental abilities, verbal, numerical, reasoning, spatial, and two sequential tests of educational progress, reading and mathematics. Analysis indicated that (1) frequency ratio, with relatively greater left-hemisphere activity was the most important EEG variable related to mental functioning in the talented group; (2) while EEG period and alpha ratio were the two most important EEG variables related to intellectual functioning in the mentally handicapped group; (3) hemispheric symmetry was of some importance for both groups, with greater symmetry being associated with superior performance. Contrary to Ertl's neural efficiency theory, (1) neural information transfer rate was significantly positively (not negatively as predicted) related to intellectual performance (for the handicapped group only), and (2) alpha was significantly related to intellectual functioning (for the handicapped group only). Further research is needed to extend and confirm the above findings dealing with persons scoring in the normal IQ range and during mentally active as well as resting conditions.
Subject(s)
Child, Gifted/psychology , Electroencephalography , Intelligence , Learning Disabilities/psychology , Adolescent , Alpha Rhythm , Cerebral Cortex/physiology , Child , Dominance, Cerebral/physiology , Evoked Potentials , Female , Humans , MaleABSTRACT
64 children aged 8 yr. were trained on a tactile simultaneous discrimination task. Selective attention was measured in terms of percentage contact time per trial to the relevant dimension. Inter- and intra-couplings per trial were also recorded. Multivariate analyses were carried out to examine the role of component factor scores, obtained from a component curve analysis of the percentage touching time per trial, and selected cognitive variables in differentiating between the fast and slow learner groups. Percentage touching time factor scores and a memory factor were significant, but there was no significant difference between the groups in the number of couplings made. As learning progressed the number of inter- and intra-couplings decreased.
Subject(s)
Attention , Learning Disabilities/etiology , Aptitude , Child , Discrimination, Psychological , Form Perception , Humans , Intelligence , Learning Disabilities/diagnosis , Mental Recall , Practice, Psychological , Stereognosis , Time Factors , TouchSubject(s)
Child Development , Memory , Verbal Learning , Visual Perception , Age Factors , Child , Humans , Information Theory , Verbal BehaviorABSTRACT
Effect of concept grouping (A), dominance level (B), and sex (C) on immediate venbal recall was investigated using a univariate and a multivariate analysis of covariance (ANCOV and MANCOV) respectively, with digit span as a concomitant variable. 138 adult Ss, 56 males and 82 females, participated in this experiment. The MANCOV in 23 factorial design showed A, C, and A x B as significant beyond the .05 level. The covariate effect was also significant. The confounding of the variance due to the neglect of sex effecbs and the lack of appropriate analytic technique in t.he previous studies using concept forma4ion materials are proposed to account for the differences in results.
