• Presentation

Interviewer Characteristics Related to Field Interviewer Falsification

Citation

Griffin, A. B., Pratt, J. G., Shuangquan, S., Coffey, L. E., Terrey, S. A., & Clark, C. W. (2011, May). Interviewer Characteristics Related to Field Interviewer Falsification. Presented at AAPOR 2011, Phoenix, AZ.

Abstract

Falsification of data is detrimental to the integrity of any research. When a field interviewer falsifies data costs increase when investigating possible falsification, reworking fraudulent cases, and hiring replacement staff. Given the negative impact of falsification on research integrity and budgets, survey data collection organizations and their sponsors have an interest in finding means of detecting and preventing falsification.  Can we identify any interviewer characteristics that are directly associated with falsification? Do factors such as interviewer age, experience level, or other typically measured performance ratings such as response rate or cost per interview correlate with falsification?  Are there differences between interviewers whose work is questionable but not falsified? Using the National Survey on Drug Use and Health (NSDUH) as an example, these questions will be addressed.


The National Survey on Drug Use and Health (NSDUH) is an annual face-to-face, household survey. The NSDUH is sponsored by the Substance Abuse and Mental Health Services Administration. First conducted in 1971, this study provides national, state and substate data on substance use and mental health in the civilian, noninstitutionalized population age 12 and older. Approximately 140,000 household screenings and 67,500 NSDUH interviews are completed annually and data are collected in all 50 states plus the District of Columbia by approximately 700 interviewers.
 

The NSDUH implements a complex process to verify the quality and accuracy of each interviewer’s work. In some instances, an interviewer’s work is field verified to determine whether or not he or she made proper contact. Using 2005 to 2010 non-identifiable archived data, this presentation will examine the correlation of detected falsification with factors such as interviewer experience on the NSDUH (considering both time and workload), age, and response rate history. Trends in the number of interviewers found falsifying and their characteristics will also be investigated as well as any differences between interviewers with questionable work requiring field verification, interviewers who falsified, and those who did not falsify.