Font Size:
Quality Monitoring Strategy of 2013 China Mental Health Survey
Gina Cheung, Yan Sun
Building: Holme Building
Room: Holme Room
Date: 2014-12-08 01:30 PM – 03:00 PM
Last modified: 2014-10-31
Abstract
In this paper we will introduce the quality monitoring strategy of 2013 China Mental health Survey (CMHS).
Three approaches have been used for interview quality monitoring on CMHS: data assessment, audio record evaluation and call back verification.
Not only survey data, but also paradata, including response time to each question, adjusted interview length, and contact records, was used for data assessment. Suspicious samples were flagged after data assessment.
Except for the suspicious samples, a certain percent of random samples were also selected for interview quality evaluating.
All selected samples were put into evaluation or verification which aim at examine the unregulated interviewer behaviors. Both major error such as falsification and minor error such as unsatisfactory probing were monitored. When one approach failed, the other approach needs to be done to examine the interview quality.
The results of evaluation and/or verification were used to evaluate the interviewers’ overall performance. Those results were real time shown up in QC system and daily/weekly QC report were provided both on sample level and interviewer level.
Three approaches have been used for interview quality monitoring on CMHS: data assessment, audio record evaluation and call back verification.
Not only survey data, but also paradata, including response time to each question, adjusted interview length, and contact records, was used for data assessment. Suspicious samples were flagged after data assessment.
Except for the suspicious samples, a certain percent of random samples were also selected for interview quality evaluating.
All selected samples were put into evaluation or verification which aim at examine the unregulated interviewer behaviors. Both major error such as falsification and minor error such as unsatisfactory probing were monitored. When one approach failed, the other approach needs to be done to examine the interview quality.
The results of evaluation and/or verification were used to evaluate the interviewers’ overall performance. Those results were real time shown up in QC system and daily/weekly QC report were provided both on sample level and interviewer level.