Return to search

Sex, motivation, extraversion and neuroticism as multiple predictors and moderators with predicting two grade nine attainment criteria.

Although the moderator variable concept has acquired popularity among researchers concerned for the low predictive validity of psychological tests, no conclusive evidence has been presented for the greater effectiveness of moderated prediction techniques over the traditional multiple regression model. It was the principal aim of this research to investigate whether use of sex and five measures of personality and motivation as moderators would increase the size of correlations of a mental ability predictor with criteria of academic success. As behaviour may be sensitive to the conditions in which it occurs, the research also sought to determine whether use of school grades and standardized achievement tests as criteria might obtain inconsistent results. The Ss in this investigation were 246 males and females enrolled in grade nine at a public high school. They were, initially administered a battery of tests designed to measure mental ability, extraversion, neuroticism, and three aspects of academic motivation. Six weeks later, standardized reading and mathematics achievement tests were administered to the same Ss. End-of-year grades in english and mathematics also were obtained. The sets of achievement test scores and grades were combined to yield two overall criterion measures. A statistically significant improvement was demonstrated (P < .05, one-tailed) in support of using the extraversion variable as a moderator when predicting achievement test performance. The more accurate prediction was given for a subgroup with scores in the low one-third of the extraversion distribution. Joint use of the sex, personality and motivation variables did not enhance this moderator effect. In no case did the moderator analysis provide a significant improvement over the traditional multiple regression technique when predicting school grades. These research findings were interpreted and discussed in terms of their implications for psychological theory and research.

Identiferoai:union.ndltd.org:uottawa.ca/oai:ruor.uottawa.ca:10393/10500
Date January 1976
CreatorsPerrault, Yvonne L.
PublisherUniversity of Ottawa (Canada)
Source SetsUniversité d’Ottawa
Detected LanguageEnglish
TypeThesis
Format166 p.

Page generated in 0.0017 seconds