A Heuristic Method for Estimating the Relative Weight of Predictor Variables in Multiple Regression (original) (raw)

The paper presents a heuristic method for estimating the relative weight of predictor variables in multiple regression analysis, addressing the common need for clearer measures of predictor importance. Traditional methods like standardized regression coefficients and zero-order correlations often overlook the combined effects of predictors, leading to inadequate assessments. The proposed method aims to refine these evaluations by better capturing both individual and joint contributions of variables, enhancing the reliability of insights generated from regression analyses.