Comparing the Utility of Predictor Variables at University Level

Matthew Rudd

doi:10.7176/ikm/9-3-07

Abstract

This quantitative research study investigated the predictive utility of six separate predictor variables; self-efficacy scales, teacher ratings, prior performances, students’ study habits, perceived attitudes, as well as the reliability of self-reporting instruments. Anticipated English language outcomes across all measures were compared with end-of-term scores using significance testing (0.05) and Pearson’s Correlation Coefficient. The sample population consisted of 84 third year university students (females = 51; males = 33) studying at a private university located on the outskirts of Bangkok, and were from two separate faculties; business administration (n = 58) and engineering (n = 26). Overall, the mean level of English language performance was 59.94%, and comparative to which, the most reliable predictor of academic performance were self-efficacy scales (S.E: 60.07%; not sig, p <0.05), which also manifested a relatively close correlation (r = 0.5598). Likewise, teacher ratings (57.98, not sig, p <0.05) were almost as precise, and highly correlative (r = 0.7068), but considered to be less pragmatic owing to potential subjectivity in non-experimental conditions. However, contrary to previous research, students’ prior performances (54.36) were significantly inferior to end-of term attainments, despite the inference of a strong correlation (r = 0.842). Moreover, study habits (42.14) and attitudes towards English (76.31) were grossly inaccurate with minimal correlative value. Separately, results generated from self-reported instruments closely resembled the students’ official grades (prior: 54.36 vs. reported: 55.37%; r = 0.9393); notwithstanding this level of accuracy, the efficiency of this instrument may be put into question when researching debuting students with little academic history. Thus, in terms of overall dependability and practicality, self-efficacy judgments were classified as the most efficacious means of predicting academic performance. Key Words: Self-efficacy, predictor, attainments, performance, variable DOI : 10.7176/IKM/9-3-07 Publication date :March 31 st 2019

Highlights

The most well researched conceptually-rich predictor variable discussed in this paper is that of self-efficacy beliefs, which forms the theoretical basis of this paper, and is benchmarked against a range of conventional predictor variables, which include; teacher ratings, prior performances, students’ study habits, perceived attitudes, as well as the reliability of self-reporting instruments
According to Bandura (1986) peoples’ behaviour can be better predicted by the beliefs they hold with respect to their capabilities; a claim this paper aimed to thoroughly test alongside a host of predictor variables that have been the focus of educational research, and are discussed at length in the literature review
This paper examines the accuracy of selfreporting instruments as a means of base-lining students’ academic performance for research purposes; very little research has investigated this issue

Summary

Introduction

The most well researched conceptually-rich predictor variable discussed in this paper is that of self-efficacy beliefs, which forms the theoretical basis of this paper, and is benchmarked against a range of conventional predictor variables, which include; teacher ratings, prior performances, students’ study habits, perceived attitudes, as well as the reliability of self-reporting instruments. Self-efficacy beliefs are summarised as people’s beliefs about their capabilities to produce designated levels of performance that exercise influence over events that affect their lives ... (and) determine how people think, feel, motivate themselves and behave (Bandura 1994). According to Bandura (1986) peoples’ behaviour can be better predicted by the beliefs they hold with respect to their capabilities; a claim this paper aimed to thoroughly test alongside a host of predictor variables that have been the focus of educational research, and are discussed at length in the literature review

Objectives

Methods

Findings

Conclusion