Estimating Mode Effects Using Data from CATI and CAWI
Abdullah Alhokail (Saudi Center for Opinion Polling) - Saudi Arabia
Shahad Al Mashan (Saudi Center for Opinion Polling) - Saudi Arabia
Keywords: mixed modes, propensity weighting, Saudi Arabia
Abstract
This study investigates the differences in survey results obtained through two distinct survey designs: Computer-Assisted Telephone Interviewing (CATI) and Computer-Assisted Web Interviewing (CAWI). The comparative analysis is conducted in three stages:
1. Unweighted Analysis: Initial variations in response patterns between the two methodologies are examined without any adjustments, providing a baseline for comparison.
2. Demographic Weighting: Results are weighted using demographic benchmarks provided by GaStat to enhance representativeness and align with the target population distribution.
3. Propensity Weighting: Adjustments are applied using propensity weighting techniques to mitigate selection bias and mode effects, ensuring greater comparability between CATI and CAWI results.
This structured approach aims to identify and control for methodological disparities, enabling a robust evaluation of the impact of survey modes on data quality and accuracy.