As all data contains variability, it is important to use probabilistic thinking and language when making claims from data. This requires paying attention not only to patterns and comparisons within and across variables but also such things as expected and prior values, sample sizes, and significance.

Formulating a question or identifying a problem that can be addressed with data affects the opportunities for interpretation and results from the data. The ability to make and justify strong claims relies on identifying questions that are testable and can be answered with data. Additionally, identifying the uncertainty or limitations within the problem space is an important component of formulating conclusions.

Though there is often an instinct to use data to make large generalized claims, the applicability of inferences and claims that are made are constrained by the sample, population, and context of the data.