Chapter V. A comparison of CATI and non-CATI questionnaires

Area of Science:

• Survey methodology
• Health services research
• Data collection techniques

Background:

• The increasing use of telephones for surveys necessitates evaluating new data collection technologies.
• Computer-Assisted Telephone Interviewing (CATI) systems are emerging as a significant mode for future surveys.
• Understanding the implications of transitioning to CATI is crucial for survey design and implementation.

Purpose of the Study:

• To benchmark the transition to CATI by comparing it with traditional non-CATI telephone interviewing.
• To identify potential problems and benefits associated with mode changes in telephone surveys.
• To assess differences in key survey criteria between CATI and non-CATI methods.

Main Methods:

• A comparative experiment was conducted using both CATI and non-CATI interviewing modes.
• Data collection involved assessing response rates, interviewer/respondent reactions, and health statistics.
• Statistical analysis focused on identifying significant differences between the two interviewing methods.

Main Results:

• Most criteria, including response rates and health statistics, showed minor differences between CATI and non-CATI interviews.
• CATI interviews resulted in a longer average interview duration compared to non-CATI.
• CATI interviewing demonstrated lower skip error rates and potentially reduced interviewer variability.

Conclusions:

• While CATI interviews are longer, they offer significant advantages in reducing survey errors like skip patterns.
• The findings provide a benchmark for the adoption of CATI systems, highlighting areas for optimization in software and hardware.
• CATI systems present a promising future for telephone surveys, enhancing data quality and accuracy.