On parameter estimation in the physics lab based on inverting a slope regression coefficient

Abstract

Measurement uncertainty is a non trivial aspect of the laboratory component of most undergraduate physics courses. Confusion about the application of statistical tools calls for the elaboration of guidelines and the elimination of inconsistencies were possible. Linear regression is one of the fundamental statistical tool often used in a first year physics laboratory setting. In what follows we present an argument that leads to an unambiguous choice of (a) variable(s) to be used as predictor(s) and variable to be predicted.