Skip to main content
Thirty-Third Annual Meeting of the Society for Research in Psychopathology

Full Program »

9. What do we Measure when Smokers Quit? Identifying Smoking Cessation Constructs using an Explicit Measurement Model

Quit smoking clinical trials typically use separate questionnaires to individually capture smoking withdrawal, craving, positive and negative affect, and medication side effects. This results in an implicit measurement model that ignores item and conceptual overlap occurring among these measures. Such an approach may not accurately reflect the actual constructs involved in cessation. To clarify the number and nature of cessation constructs we developed an explicit measurement model and compared it to the implicit measurement model. EFAs and CFAs were conducted on self-report questionnaires from 1246 smokers. A 5-factor model (negative affect, positive affect, craving, somatic symptoms, and sleep problems) fit the data well prior to and during quitting and these factors were generally consistent over time. Withdrawal and side effects did not emerge as separate constructs; items from “withdrawal” and “side effect” measures spread across most factors. Correlational analyses and fit indices indicated that the 5-factor model attenuated construct overlap and better represented the latent constructs compared to the implicit model. Future analyses will examine the relationship between these constructs with treatment condition and cessation status.

Sarah Tonkin
University at Buffalo

Trevor Williams
University at Buffalo

Leonard Simms
University at Buffalo

Stephen Tiffany
University at Buffalo

Larry Hawk
University at Buffalo


Powered by OpenConf®
Copyright ©2002-2018 Zakon Group LLC