A new procedure for determining the optimal number of interpretable factors to extract from a correlation matrix is introduced and compared to more conventional procedures. The new method evaluates the magnitude of the Very Simple Structure index of goodness of fit for factor solutions of increasing rank. The number of factors which maximizes the VSS criterion is taken as being the optimal number of factors to extract. Thirty-two artificial and two real data sets are used in order to compare this procedure with such methods as maximum likelihood, the eigenvalue greater than 1.0 rule, and comparison of the observed eigenvalues with those expected from random data.
The authors ofthe Eysenck Personality Inventory (EPI) and the Eysenck Personality Questionnaire (EPQ) have claimed that the extraversion scales contained in the two tests are equivalent. Although scores on the two scales are moderately highly correlated, supplementary analyses suggest that they differ in at least one important respect. While the EPI scale measures extraversion as a reasonable mix of impulsivity and sociability, the EPQ's scale is almost purely a measure ofsociability. Recent experimental evidence demonstrates that impulsivity is responsible for several findings previously attributed to extraversion. This evidence raises serious doubt about the usefulness ofthe EPQ extraversion scale in experimental research on extraversion.After a measurement device has been in use for some time, the original investigator will often want to revise it on the basis ofgrowth in the theory that spawned the scale. Alternatively, psychometric analysis may suggest that the reliability and/or validity of the scale would be enhanced by the deletion of existing items or the addition of new items. In either of these cases, ifthe revised scale can be shown to be simply an alternate form of the original, we feel justified in applying the validational data accumulated with the original scale to the revision. Three criteria should be met in order to demonstrate that the two scales are in fact equivalent:(I) Scores on the two scales should correlate highly. This is the traditional and minimal requirement.(2) The two scales should have the same pattern of relationships to theoretically important constructs. One way in which this is demonstrated is by the pattern of correlations between the two putatively parallel scales and other personality scales. Ideally, the new scale should demonstrate equivalent or better ability to predict relevant non-test behaviours.(3) The two scales should have the same internal structure, especially if there is reason t o believe that the structure is not clearly unidimensional.Eysenck & Eysenck (1975) introduced the Eysenck Personality Questionnaire (EPQ) as a major revision of the Eysenck Personality Inventory (EPI; Eysenck & Eysenck, 1968). The EPQ is a response both to theoretical development (by adding a psychoticism scale to the previously introduced extraversion and neuroticism scales) and psychometric criticisms.Regarding the extraversion and neuroticism scales, the authors claim that 'the E and N scales of the present questionnaire are so similar to the corresponding scales of the other questionnaires that whatever has been discovered about correlates of E and N with the use of the older scales must be assumed to apply with equal force to the new scales ' (Eysenck & Eysenck, 1975, p. 3). The fact that only 25 out of 57 items from the EPI appear on the EPQ, and that of these 25, eight have slightly altered wording (Helmes, 1980), suggests the need for empirical investigation of this claim. The trait of extraversion in particular has generated and continues to generate a great deal of research...
To facilitate their initial acquisition of knowledge from text material, we paired students with same-sex partners with whom they read and studied two passages describing technical equipment. We tested three different scripts for cooperative interactions. In one group (n = 26), partners each read only one passage, then taught each other the information they had read. In the second group (n = 20), both partners read both passages, stopping periodically to summarize the material to each other. In the third group (n = 25), each partner read alternate pages of both passages, stopping to teach each other the material they had read. Free-and cued-recall tests revealed that participants using the cooperative teaching script significantly outperformed participants in the other groups. Further analyses indicated that after playing a teaching role, students recalled significantly more material for the passage they taught. After playing a learning role (i.e., for the passage they did not read but were taught by their partners), they did not recall significantly less than those who read both passages. Although we discuss several possible explanations for these effects, further experimentation is needed to determine their validity.Our purpose in this research was to explore ways in which learners could interact during acquisition of new knowledge that would lead to effective understanding and recall of written information. Past research has established the value of cooperative learning in classroom settings (e.g.to enhance academic performance but also to increase self-esteem and encourage positive social relations among classmates.Beneficial effects of cooperative learning have also been found in more controlled laboratory studies (Dansereau, 1985;Dansereau et al., 1979;McDonald, Larson, Dansereau, & Spurlin, 1985). In these studies, a general script was developed for cooperative learners working as partners. The script involves two roles: recaller and listener-facilitator. First, both partners read a section of text material silently. Then the partner acting as recaller gives a verbal summary of all he or she can remember from the text. The listening partner's role is to detect any errors or missing information in the summary. Both partners then discuss ways they will try to remember the information, and then continue to read the next portion of text. It has been shown that cooperative learners' performance on recall tests is significantly higher than the recall of individuals studying alone. When further research was conducted to
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.