Measurement and Model Building Introduction

Abstract

The goal of empirical study is to build, test, and evolve models of a discipline. This requires studying the variables of interest in multiple contexts and building a set of models that can be evolved, discredited, or used with confidence. This implies we need to perform multiple studies, both replicating as closely as possible and varying some of the variables to test the robustness of the current model. It involves running many studies in different environments, addressing as many context variables as possible and either building well parameterized models or families of models that are valid under different conditions. This is beyond the scope of an individual research group. Thus it involves two obvious problems: how do we share data and artifacts across multiple research groups and what are good methods for effectively interpreting data, especially across multiple studies.