Unit 6, Modeling Measurements, encourages students to integrate data and chance by constructing and revising models that explain the variability of their measurements or of their productions (e.g., counts of the numberof toothpicks in each package). Students build and test models of the process of measurement that resulted in the arm span (or some other attribute) measurements they previously obtained. Their models of the measurement process estimate both the true measure of the attribute and random errors made as they measured. For example, just by chance, gaps and overlaps in measurement occur due to difficulties in using the short ruler to measure a long distance, such as an arm-span. Students identify sources of error like these and construct chance devices to model each source. Then, they combine the devices that model chance errors with the estimate of the true measure to produce a model of the measurement process. They compare their models’ outputs to their sample data and explore the models’ behavior from sample-to-sample, using the concept of sampling distribution to judge the adequacy of their models. They even build and test a "bad" model. The Modeling Variability (MoV) construct illustrates students’ conceptions of modeling.