Detecting Climate Change in Binary and Continuous Data

Abstract

In this paper, we investigate whether and why people perceive greater change in binary climate data (e.g., lake freeze) compared to continuous climate data (e.g., winter temperature). We conduct a behavioral experiment and find that participants who view binary climate data observed greater change than participants who view continuous climate data. We explore emotional valence and changepoint saliency as potential explanations for this behavior through cognitive experiments and Bayesian changepoint detection.