In this episode of The Method to Our Madness, Stacey and Bill unpack the different types of statistics commonly used in program evaluation—and why “simple” doesn’t mean unsophisticated. They explore how evaluators balance accessibility with rigor, discussing when descriptive statistics, t-tests, and ANOVA are the right tools, and when more advanced methods like hierarchical linear modeling or propensity score matching make sense. Along the way, they dig into the hidden complexity behind clean-looking results, the limits of common data visualization tools, and why effect sizes often matter more than p-values. Drawing on real research experiences, disciplinary perspectives from sociology and psychology, and lessons learned from both practice and academia, this conversation highlights what it really takes to produce statistical findings that are both credible and useful. The episode wraps up with a reflection on communicating results clearly—and an open invitation for listeners to suggest future topics they’d like to explore in greater depth.

 

Podcast music written and produced by Stacey Merola. Podcast artwork by Zoe Targino.