Big Data And The Well-Being Nexus: Tracking Google Search Activity By State IQ

Michael A. McDaniel, Virginia Commonwealth University
Bryan J. Pesta, Cleveland State University
Allison S. Gabriel, Virginia Commonwealth University

Abstract

In the era of “big data,” internet search activity can provide interesting insight into human behavior. Here we used the Google Correlate algorithm (a database tracking billions of user searches) to identify search terms that co-varied most strongly with U.S. state-level IQ and wellbeing (see Pesta, McDaniel, & Bertsch, 2010). First, we identified the 100 strongest positive (e.g., crock pot applesauce, custom woodworking) and negative (e.g., ASVAB for Dummies, Hello Kitty) search term covariates for state IQ. We then rationally clustered search terms into composites (e.g., “food,” “job seeking activity”) based on similarity of concept. Thereafter, we correlated the composite scores with other well-being variables (e.g., crime, health). Search-term composite scores correlated stronglywith allwell-being variables.We offer post-hoc explanations for the various composite-score correlations, showing how state differences in internet search activity fit within the “well-being nexus” for the U.S.Moreover, we explore how the use of Google Correlate can inform additional research inquires in this domain.