Complexity of measuring "cigar use" in adolescents: results from a split sample experiment

Area of Science:

• Public Health
• Survey Methodology
• Tobacco Control

Background:

• Health surveys assessing cigar use can be influenced by the inclusion of brand-specific examples (BE).
• Previous research indicates that BE impacts response rates in surveys.
• The consistency of this impact across different demographic groups and locations requires further investigation.

Purpose of the Study:

• To investigate whether the impact of brand-specific examples (BE) on health survey response rates for cigar use is consistent across race, gender, and geographic locale.
• To analyze the differential effects of including BE versus no brand-specific examples (NBE) on reporting of lifetime and current cigar use.

Main Methods:

• A split-sample experimental design was employed using the 2009 Cuyahoga County Youth Risk Behavior Survey.
• Two survey versions were created: one with NBE and one with BE in items assessing cigar use.
• Surveys were randomly distributed within classrooms in 20 randomly selected high schools.

Main Results:

• Within the City, BE respondents reported higher lifetime and current cigar use prevalence compared to NBE, with significance only among Black respondents (OR = 1.45).
• In the Outer Ring, White BE respondents showed significantly lower odds of reporting lifetime (OR = 0.73) and current cigar use (OR = 0.66) compared to White NBE respondents.
• Findings indicate differential impacts of BE based on race and geographic location.

Conclusions:

• The inclusion of BE in cigar product use measures may enhance estimates within certain at-risk populations.
• However, to achieve better estimation of cigar product use, developing additional items to assess subtypes of cigars is recommended.
• Survey design modifications, including the use of BE, should consider demographic and geographic variations for accurate health data.