Two-Wave Online Survey and Negative Binomial Regression: Using Optimal Methods and Statistics in a Binge-Drinking Study

Abstract

In this case study, I describe the procedure of conducting a binge-drinking study using a two-wave online survey and a negative binomial regression, taking readers to the heart of some methodological and analytical issues that arose during the research process. It sheds light on the particular challenges of correctly applying the theory of planned behavior and using a prospective and more precise measure of the behavioral outcome. In recounting my research process, I have paid particular attention to the use of the two-wave online survey and why I decided it was the optimal method for testing the TPB model, as well as the use of the negative binomial regression and why it is the most appropriate analytical technique for count data. Such practices have made the results of this study more credible and its causal inferences more persuasive. By adopting these practices, I improved the overall quality of the study, and thus its contributions to empirical research on binge-drinking are more significant. This case study will assist student researchers in selecting the most appropriate methods and statistical techniques, dealing with methodological challenges involved in a two-wave survey, and mitigating potential weaknesses, when designing and conducting their own research.

locked icon

Sign in to access this content

Get a 30 day FREE TRIAL

  • Watch videos from a variety of sources bringing classroom topics to life
  • Read modern, diverse business cases
  • Explore hundreds of books and reference titles