A STUDY OF MULTIPLE REGRESSION ANALYSIS- A REVIEW

Abstract

Overall this study by Uyanik and Guler (2013), published in the journal of Procedia-Social and Behavioral Sciences is a relatively good one, with some weaknesses in the areas of data analyses and methods. However, the article fulfills the following research prerequisites 1) identified a research problem; 2) stated an objective-the purpose of the study was to determine whether or not the five independent variables are significant predictors of KPSS-scores, 3) employed a method, 4) tested a hypothesis, 5) and generated discussion and conclusion. The authors succinctly present the use of multiple regression analysis by way of using 240 undergraduate students pursuing Psychological Counselling and Guidance at Sakarya University, for the academic year 2011-2012. The study relates to an employee selection examination that was conducted in 2012. The introduction of the paper is on substantiate issue of regression, highlighting the types (i.e., univariate and multivariate regressions), rationale for their usage (i.e. univariate, one dependent variable; multivariate, two or more dependent, and independent
variables), testing of assumptions of linearity (i.e. normality and linearity) and presents a purpose for the usage multiple regression analysis.  
The weaknesses of the study are almost entirely in the method section of the paper. The researchers did not provide detailed information on the research design, operationalization, reliability and validity of variables, issues relating to ethics, and issues relating to conflict of interest. The findings are detailed, with some omissions (descriptive statistics-means, standard deviation, and confidence interval) and discussion was rich with appropriate conclusions that are generated from the research findings.  
The findings section begins with a table that presents frequency for each of the variables, independent (i.e., measurement and evaluation; educational psychology; curriculum development; guidance and teaching methods) and dependent KPSS-scores, and missing values (Table 1). It can be deduced from this Table (Table 1) that the response rate is 100%.

The weaknesses of the study are almost entirely in the method section of the paper. The researchers did not provide detailed information on the research design, operationalization, reliability and validity of variables, issues relating to ethics, and issues relating to conflict of interest. The findings are detailed, with some omissions (descriptive statistics – means, standard deviation, and confidence interval) and discussion was rich with appropriate conclusions that are generated from the research findings.