10. SPSS Classroom - How to Analyze, Interpret, Report One Way ANOVA in SPSS

One-way ANOVA (one-way analysis of variance) is the go-to method when a researcher needs to compare a continuous outcome across three or more independent groups—something a two-group independent-samples t test can’t handle. The core idea is straightforward: compute the mean of the dependent variable within each group, then use an F statistic to test whether those group means differ more than would be expected by chance. If the F test is large enough to reject the null hypothesis, the analysis moves beyond “there is a difference somewhere” to “which groups differ from which.”

The transcript frames one-way ANOVA as a between-groups procedure built around a single categorical factor (the grouping variable) with three or more levels. Examples include comparing lecturer income across cities (London, Paris, Islamabad), comparing optimism scores across age levels, or comparing staff satisfaction across employment types (permanent, part-time, casual). In each case, the grouping variable is categorical, while the outcome being compared—income, optimism, satisfaction—is treated as a dependent variable measured on an interval or ratio scale.

At the statistical heart of the method is the F statistic. It contrasts variability between group means against variability within groups. A higher F ratio indicates stronger evidence that group means are not equal; an F ratio near or below 1 suggests no meaningful differences. When the null hypothesis (that all group means are equal) is rejected, the analysis still doesn’t identify the specific pairs responsible for the difference. That job falls to post hoc multiple comparisons, which systematically test pairwise contrasts such as Group A vs. Group B, Group A vs. Group C, and Group B vs. Group C.

The transcript distinguishes overall effects from individual comparisons. The “main effect” from ANOVA reflects whether any differences exist across the factor levels in general. Post hoc tests then determine which levels differ. Common post hoc options mentioned include LSD and Tukey HSD, with the practical workflow in SPSS emphasizing the need to choose the correct post hoc approach based on variance assumptions.

A major practical section lays out assumptions needed for one-way ANOVA: the dependent variable should be interval/ratio and approximately continuous (often created by combining multiple Likert items into a single scale), the independent variable must have three or more categorical groups, observations must be independent, outliers should be limited, and each group’s dependent variable should be approximately normally distributed (checked via skewness and kurtosis). Homogeneity of variance is also required; the transcript highlights Levene’s test for equality of variances. If homogeneity fails, the analysis should switch from standard one-way ANOVA to Welch ANOVA (a more robust alternative) and use an appropriate post hoc procedure.

The SPSS walkthrough uses organizational commitment as the dependent variable and job rank (junior, middle, senior) as the factor. It demonstrates how to check outliers (via Explore), assess normality by splitting files by rank and reviewing skewness/kurtosis, run one-way ANOVA with options for homogeneity and Welch, and then interpret results. When Levene’s test indicates unequal variances (p < .05), the transcript instructs using Welch-based significance and then reading post hoc pairwise comparisons to conclude, for example, that junior vs. middle may be non-significant while junior vs. senior is significant. Finally, it outlines how to report results using the appropriate test statistic (Welch) and pairwise comparison p-values, plus a summary table of group means and standard deviations.