Faculty of Public Health - Andalas University - OCS, 13th IEA SEA Meeting and ICPH - SDev

Font Size: 
Randomized Response Technique Vs. Direct Interview: Alcohol Use Among Adolescents
Ashish Datt Upadhyay

Last modified: 2018-08-07


Introduction: The survey using direct interview method on sensitive topics (e.g., alcohol use) generally involves huge non-response, or, non-true responses. In India, alcohol use among adolescents is always under reported. To cope up with this, Warner (1965) suggested a randomized response technique for estimating burden on such topics more accurately. Our aim is to compare the data collection approaches (i.e., randomized response technique (RRT) vs. direct interview method) among adolescents on ever alcohol use.

Methods The two groups randomized cross-sectional survey was conducted among 796 (401 under RRT and 396 under Direct Interview) students of age 14 to 19 years. Survey on ever alcohol use was conducted on students of class 9th to 12th at three schools of National Capital Territory, Delhi, India.  Difference in proportion of ever alcohol use was assessed by proportion test and related 95% confidence interval.

Result Out of 401 surveyed adolescents under RRT, 22.8% (95%C.I:17.08-28.51) were ever alcohol users. On the other-hand, among 396 students surveyed under direct interview, 9.1% (95%C.I:6.27-11.95) were ever alcohol users. Study shows that RRT could detect 13.6 % (95% CI: 8.34-19.03) more hidden ever alcohol users than direct interview method.

Conclusion Due to availability of more privacy to respondents on sensitive topics under RRT, chance of true response increases.