Analyzing Marginal Response Rates in the CAHPS Medicare Fee-for-Service Survey
Campbell, L. N., Brown, G. G., Carpenter, L., & Dimitropoulos, L. L. (2005, May). Analyzing Marginal Response Rates in the CAHPS Medicare Fee-for-Service Survey. Presented at American Association for Public Opinion Research 60th Annual Conference, Miami Beach, FL.
This paper is an effort to extend recent work that evaluated the effect of increasing the level-of-effort by raising the maximum number of call attempts in a telephone follow-up of the CAHPS Medicare Fee-for-Service Survey. In that work, we used survival analysis to analyze marginal response rates among non-respondents to the mail survey who participated in the telephone follow-up to determine the appropriate number of attempts to make to each sample member. The goal was to optimize the response rate and marginal cost of a completed survey. Our key findings were:
- With the increased level-of-effort, we did not penetrate an under-represented segment of the population with greater number of call attempts;
- With the maximum number of calls set at twelve, we observed neither significantly diminishing returns with greater calls, or increased marginal cost per completed survey; and
- The CAHPS measures themselves were unaffected by increased call attempts.
This paper extends this work by further developing the statistical methods used to analyze the response data, including statistical tests for checking the significance of our previous and new findings. Discussion will focus on further analysis of the marginal effects on response rate and cost for increased call attempts, further development of the survival modeling, and of the balancing of the effects of bias and precision in the estimates.