Measuring Social Smoking: Challenges for Analysis and Implications for Surveys and Intervention

• Activity Number: 520
• Type: Contributed
• Date/Time: Wednesday, August 1, 2012 : 10:30 AM to 12:20 PM
• Sponsor: Section on Statistics in Epidemiology
• Author(s): Kevin Delucchi*+ and Danielle Ramo
• Companies: University of California at San Francisco and University of California at San Francisco
• Keywords: measurement ; survey ; smoking ; young adults

Abstract

One of the current challenges for the survey and analysis of smoking behaviors among young adults, the highest risk group for smoking, is the definition and measurement of 'social smoking.' That is, a non-daily pattern where smoking occurs primarily in social settings-a behavior more common among young adults. Such intermittent behavior, coupled with an uncertain definition of 'social smoking,' may impact the analysis of data on this behavior as well as efforts to increase smoking quit rates in this population. Using data from an internet-based survey of 3740 adults between 18 and 25 we examined the effects of differing definitions of social smoking. Results indicate strong but not complete agreement among answers to questions regarding this behavior. Of those who said they smoke 'only when others are present,' for example, 6 percent also said they smoke 'mainly alone.' Logistic regression models find a significant relationship of both self-described smoking behavior and identification as a social smoker with intention to quit smoking. This work may help with targeting cessation efforts and explain differences among survey findings in recent tobacco use trends.