Inside Out: Externalizing Assumptions in Data Analysis as Validation Checks

TL;DR

This paper formalizes the concept of analysis validation checks to externalize and evaluate informal assumptions in data analysis, aiding in diagnosing unexpected results through a systematic procedure.

Contribution

It introduces a method to identify key validation checks that predict unexpected outcomes, enhancing transparency and diagnostic capabilities in data analysis.

Findings

Checks can accurately predict unexpected results

Selected checks are often independent and informative

Method applied successfully to real-world data examples

Abstract

In data analysis, unexpected results often prompt researchers to revisit their procedures to identify potential issues. While some researchers may struggle to identify the root causes, experienced researchers can often quickly diagnose problems by checking a few key assumptions. These checked assumptions, or expectations, are typically informal, difficult to trace, and rarely discussed in publications. In this paper, we introduce the term *analysis validation checks* to formalize and externalize these informal assumptions. We then introduce a procedure to identify a subset of checks that best predict the occurrence of unexpected outcomes, based on simulations of the original data. The checks are evaluated in terms of accuracy, determined by binary classification metrics, and independence, which measures the shared information among checks. We demonstrate this approach with a toy example using step count data and a generalized linear model example examining the effect of particulate matter air pollution on daily mortality.