An experiential cookie-based activity to increase engagement in introductory business statistics

This paper describes the Cookie Experiment, a low-cost experiential activity conducted in a single class session early in the term. Students collect data on chocolate chip cookies and pool their observations in a shared spreadsheet. One familiar object yields four variables spanning three data types: weight and diameter (continuous), visible chip count (discrete), and appearance rating (ordinal). Because all four come from the same object, data-type distinctions are drawn by contrast within one context rather than across unrelated datasets. The pooled data support instruction in data cleaning, measurement error, graphical display, and measures of center, spread, and association. Exploratory evidence from eleven course sections is reported.