Al-Hikmah University Journal

The purpose of the study was to use multiple linear regression as a model to analyse and interpret results beyond standardized regression coefficients beta weights. Using a driven data, this study used the model to illustrate how predictors could predict dependent variable by estimating the percentages of contribution. Quantitative approach using descriptive survey was espoused for the study. Using G*POWER, a sample size of 248 was selected from a population of 623 basic school teachers in the Cape Coast Metropolis, Ghana. To prove the instrument reliability, an alpha coefficient of .798 and correlation coefficient .897 were obtained. It was established from the results that in reporting MLR data, researchers should report on the change statistics (R2 change). This helps to describe both the unique and shared variance contributions of all independent variables that contribute to the R-squared (R2). Also, to know the contribution of each of the factors, researchers should estimate the R2 Change and convert the values into percentages. This study recommended that in analysing and interpreting MLR, it should focus on and include variance partitioning statistic in addition to beta weights in the presence of correlated predictors. In this way, comparisons can be drawn in text between techniques that partition R 2 if multiple techniques are used