Gary C. Ramseyer

Testing the Difference between Dependent Correlations Using the Fisher Z.

Several proposed statistics for testing the significance of the difference in two correlated r’s were first reviewed. A simple alternate procedure based on the familiar Fisher z was then suggested. This procedure, unlike its predecessors, is applicable to a wide range of problems involving tests between dependent correlations. It also has documented mathematical support when its power curves are examined. Since the procedure is asymptotic, a large sample size is required. The only other assumption is that the observations associated with the sample are drawn from a joint multivariate normal distribution.

The Robustness of the Studentized Range Statistic to Violations of the Normality and Homogeneity of Variance Assumptions

Multiple comparison procedures in recent years have earned a prominent role in the analysis and interpretation of experimental research in the behavioral sciences. Most of these procedures are designed either to test individual contrasts between means after the null hypothesis of no treatment differences in ANOVA has been rejected or to test a selected set of mean contrasts which are of a priori interest to an investigator in an experiment. Three popular techniques which have primarily been employed for the first purpose are the Tukey WSD method (1953), the Newman-Keuls test (Keuls,

USE OF ANALYSIS OF VARIANCE TO ASSESS INDIVIDUAL DIFFERENCES IN LEARNING

Historically statistical tests of significance have not permitted determination of whether the trials effect (learning) differs reliably across learners. A common statistical test, analysis of variance, can ordinarily only be used to assess whether the trials effect varies across experimental treatments. However, the learning situation can be so rearranged that reliable individual differences in learning can be statistically evaluated. The proposed arrangement is such that learning materials, rather than learner, is the randomization unit. An illustration of the use of analysis of variance to evaluate individual differences in learning, when the learning setting is rearranged, is given.

Effects of Temporal Sequencing of Student Study and Teacher Presentation on Examination Performance

Whether students should read class materials prior to or after teacher presentation was investigated. One hundred twenty-eight General Psychology students were assigned to the two conditions, 64 per condition. Students in the prior-reading condition read two reprints, 10 minutes per reprint. The post-reading group was then assembled with the prior-reading group for a teacher presentation of reprint materials. Following teacher presentation, the post-reading group was retained and read the reprints. One day later both groups were given multiple-choice tests on the reprints. No difference in test performance was found for the two groups.

Remarks on an Avoidable Error in a Posteriori Testing

Abstract This article discusses a widespread error that occurs with a posteriori testing in factorial analyses of variance. Frequently, an incorrect sample size is used under the radical in the formula for computing a critical difference such as the Fisher least significance difference. This usually leads to an inappropriately large critical difference; thus, few if any significant differences are uncovered between the means. An interesting example of this error, from a journal article by Van Berschot (1979), is examined. Researchers are urged always to use under the radical the number of observations on which each mean being compared is based.

Some Remarks on the Expected Mean Squares in Analysis of Variance in the Assessment of Individual Differences in Learning

Leicht, et al. (1978) proposed that learning units, e.g., lists, passages, be randomization units, rather than learners (subjects) as is typical, so that the significance of trial-learner interactions can be tested with analysis of variance. In an illustrative application of the suggested rearrangement, learners were arbitrarily selected (fixed-effects factor). If the investigator wishes greater generalizability of individual differences in learning, learners must be randomly selected from a specified population (random-effects factor). The appropriate expected mean squares and tests of significance are presented for both models and subsequent comments are made.

NEEDS AND PRESS OF TEACHER-EDUCATION STUDENTS, A 10-YEAR STUDY

This article describes a 10-yr. comparison study of needs and press scores for those students enrolled in a teacher-education program at one university. Responses on the Activities Index and College Characteristics Index were contrasted for the years 1961 and 1971. Current prospective teachers exhibited less intellectual and dependency needs but greater needs for impulse expression than those of 10 yr. ago. Correspondingly, current students perceived their college environment to exert less press for intellectualism and dependency. However, in apparent conflict with their needs, present day teacher-trainees perceived their environment as demanding less impulse expression. Some comparisons indicated that teacher-education students have become somewhat more like the general student population in terms of needs and press.