Correlation Analysis with SPSS: Bivariate and Multiple Pearson Correlation, Assumptions and APA Reporting

Abstract

Description: This presentation is the second of three workshop sessions on quantitative data analysis with SPSS, delivered at the College of Humanities and Social Sciences, Kampala International University. The session covers Pearson product moment correlation in both its bivariate form and as a full correlation matrix. It defines correlation in terms of direction, strength and significance, sets out conventional guidelines for interpreting the magnitude of r, and explains the coefficient of determination as a measure of shared variance. The seven assumptions of Pearson correlation are presented alongside the specific SPSS procedures that test each one, with scatterplot figures contrasting acceptable patterns against curvilinear and heteroscedastic ones. Alternatives for violated assumptions are covered, including Spearman rank order correlation, Kendall tau b, phi and the point biserial coefficient. Methodological grounding is provided on the quantitative non-experimental method, correlational design in its relationship, prediction and explanatory variants, probability and non-probability sampling techniques, and sample size in relation to statistical power. The session closes with the full analytical procedure through Analyze, Correlate, Bivariate, annotated mock SPSS output for both a single pair and a four-variable matrix, worked APA in-text sentences for significant and non-significant results, and a correctly formatted APA correlation table. Keywords: SPSS; Pearson correlation; correlation matrix; bivariate correlation; statistical assumptions; correlational design; sampling techniques; scatterplot; APA reporting; coefficient of determination; quantitative research methods