Improving tobacco use assessment among youth via telephone
Initial Award Abstract
In order to monitor the cigarette smoking epidemic, as well as to support the planning, implementation, and evaluation of tobacco use prevention initiatives, it is crucial that valid and reliable population-based data on tobacco use and related attitudes and behaviors among youth are available. The limitations of many school-based surveys (i.e., exclusion of school dropouts and frequent absentees, poor overall response rates, and inconsistent survey administration procedures) underscore the need for alternative methods to monitor patterns of tobacco use. Unfortunately, household and telephone surveys which are usually interviewer-administered yield substantially lower estimates of smoking prevalence than school surveys (which use self-administered questionnaires) despite better coverage of the youth population. Prior research suggests that this is because respondents are more likely to under-report sensitive behaviors in interviewer-administered surveys (as compared to self-administered surveys) due to respondents' concerns about privacy.
Telephone interviewing has become popular because it is economical and efficient, and it is the easier method to maintain quality control on household interviewing. However, telephone surveys which are administered by interviewers exacerbate response biases that self-administered surveys help avoid. The recent availability of interactive voice response (IVR) technology has made it feasible to conduct self- administered interviews over the telephone. An innovative survey methodology has been developed based upon this technology: "Telephone Computer-Assisted Self-lnterviewing" (T-CASI). With T-CASI the respondent listens to pre-recorded, computer-controlled questions and responds by pressing the keypad on a touch-tone telephone. Preliminary research suggests that this new survey method is promising for collecting sensitive information over the telephone. We are aware of no research using this methodology to interview youth or to survey tobacco use.
We propose to evaluate the efficacy of T-CASI in a cross-sectional, household survey of tobacco use among adolescents 12-17 years of age in California. We plan to conduct a randomized study (N=2,400) in which T-CASI is contrasted with conventional interviewer-administered, Computer-Assisted Telephone Interviewing (CATI). Our primary aims are to determine the extent to which T-CASI enhances self-reports of current smoking (as compared to CATI), and to determine the extent to which T-CASl's enhancement of smoking self-reports is a function of the respondent’s age (12-13, 14-15, 16-17), gender (male, female), and race (White, Black). The study also has the following secondary aims: (1) examine whether T-CASI enhances self-reports of smokeless tobacco use and other tobacco-related behaviors; (2) examine whether T-CASI reduces socially-desirable responding on attitudes toward tobacco use; and (3) develop two new methods to assess the likelihood that respondents under-report smoking in telephone surveys: one method based upon subjects' responses to a "lie scale” and the other method based upon subjects' response latencies in answering survey questions. |
|Assessment of Cigarette Smoking and Smoking Susceptibility among Youth (2004)
|Periodical: Public Opinion Quarterly
|Authors: Joel M Moskowitz