Response Rate

A study's response rate is an important gauge for the quality of data collection. The response rate, in its most basic form, refers to the proportion of people eligible for a study who actually enroll and participate. In fact, despite the name, the response rate is a proportion rather than a rate.

Although the concept is simple, the computation of response rates can be complex, and the use of multiple formulas diminishes the ability to compare studies by degree of nonresponse. Any comparison of response rates between studies requires knowledge of the study designs, sampling frames, modes of study recruitment, and formulas for computing response rates.

The American Association for Public Opinion Research (AAPOR) has attempted to standardize response rates for surveys conducted by mail or random-digit dialing by offering guidance on different computation methods. For example, calculating response rates for cases in a case-control study are reasonably straightforward because a list of cases is likely available (e.g., incident cases of a specific cancer received by a cancer registry). Thus, a simple proportion of the individuals with incident disease who agree to participate in the study can be computed. For controls selected from the general population, response rates need to combine information about who could be contacted, and among who could be contacted, who agrees to participate. The response rate for controls can be computed using one of the AAPOR standard formulas. Cohort studies and randomized trials tend to sample from [Page 918]defined subpopulations with a complete list of eligible participants or clinical settings with methods that allow for straightforward response rate computation.

It is widely recognized that response rates for all study types have decreased. This decrease may correlate with the increase in (and dislike of) telemarketing, overscheduled lifestyles, and lack of trust in government, academia, and medicine to use time efficiently and effectively. The Behavioral Risk Factor Surveillance System is a national random-digitdial (RDD) telephone survey that collects information about health behaviors and health care access and is administered by the Centers for Disease Control and Prevention in conjunction with the states. This survey provides an example of the decrease in response rates over the past two decades. The BRFSS response rate declined from 71% in 1993 to 51% in 2005.