Response Rates and Results of the Advance Letter Experiment 2004 RRFSS Workshop Toronto, June 23, 2004 David A. Northrup, Renée Elsbett-Koeppen and Andrea Noack ISR, York University
Outline general comments on response rates how response rates are calculated a very brief history of response rates what strategies have/are being put in place to deal with declining response rates
Outline (continued) response rates and RRFSS what did it take to get the 62% rate for 2003 RRFSS number of calls refusal conversions results of the advance letter experiment
Calculating Response Rates Completions / estimate of number of eligible households (HH) eligible HHs include completions, refusals, callbacks, and a % of the “never answered” ISR method same as BRFSS, aka “CASRO 3” RRFSS 2003 = 62%, exclude callbacks = 71%
Response Rates for American Election Study
Response Rates for BRFSS
Strategies for Improving Response Rates interviewer training increase call attempts “convert” refusals use advance letters payments (as a lottery, to completers, to the whole sample)
Data Collection at ISR for RRFSS Response Rates minimum number of 14 calls (more when there is reason to think extra calls might obtain a completion) limitation of one month sample release costs about 3 to 7 points on response rate at least one attempt to convert almost all refusals
390,106
RRFSS: Fun with Numbers 1 (2003 Data) number of calls: 390,106 percent of interviews completed first call: 21 number of interviews completed on the 10 th or subsequent calls: 3,158 number of interviews completed after a refusal: 2,678
RRFSS: Fun with Numbers 2 (2003 Data) average number of calls per completed interview: 4.65 most calls made for a single completion: 33 (for two (different) interviews) response rate if 10 plus calls and refusal conversions are dropped: 48.2% number of complaints about interviewer calling registered at ISR: 13
Characteristics of Refusers: 2003 RRFSS Data variablestandardconverted mean age education: > than high school (%) university (%) % employed % saying health fair or poor % doctor told high blood pressure % smoke 100 cigarettes # of cases: standard = 24,700, converted = 2,640 all differences significant
Characteristics of Easy and Hard to Reach: 2003 RRFSS Data variableeasy to reach hard to reach mean age education: > than high school (%) university (%) % employed % saying health fair or poor % doctor told high blood pressure % smoke 100 cigarettes # of cases: easy = 17,000, hard = 3,150 all differences significant
Letter Experiment: 1 six Health Units participated (Durham, London, Grey Bruce, Halton, Waterloo, Sudbury) test two versions of letter: ISR and HU needed to work with our monthly target and wanted to acknowledge random variation in response rates per HU per month used sample “replicates” to implement experiment
Letter Experiment: 2 Month one: replicate 1, ISR letter; replicate 2, HU letter; replicate 3, and 4 (when used), control group changed presentation in months 2 and 3 copy of letter at the end of this set of handouts exactly the same text, different letterhead, signature & envelope Except Halton
Letter Experiment: 3 survey introduction exactly the same except one additional sentence “Recently, we sent a letter to your household about an important research project.” questions about the letter the same except Durham
Why the Letter Might Improve Response Rates to RDD Surveys reduces the possibility that the telephone call catches people by surprise increases legitimacy of research project in the eye of the potential respondents demonstrates social value improves the confidence of the interviewer
Why Advance Letters Might Not Improve Response Rates to RDD Surveys letter does not reach, or is not read by, respondent ceiling effects survey topic & subpopulations they give “timid” participants a chance to prepare to say “no”
Response Rates for Months 1 & 2 of the Experiment p value =.035 (for letter (1,200) versus no letter (1,345)) p value =.025 (ISR (600) versus HU (600))
Response Rates Months 1 & 2: All Six Health Units See next slide for numbers & p values
RR by HU and Treatment treatment, RR (%) P value HU none ISR HU none/ ISR none/ HU ISR/ HU Durham London Grey Halton Waterloo Sudbury Number of cases per HU: ISR = 50, HU = 50, none = 100
Mean Calls per Completion mean # of calls P value None ISR HU none/ ISR none/ HU ISR/ HU Durham London Grey Halton Waterloo Sudbury Number of cases per HU: ISR = 50, HU = 50, none = 100
Mean Calls per Completion by Letter Status Letter/no letter p =.890, ISR/HU p =.230, not see/saw p =.001
% of First Call Attempts Leading to Completions & Refusals letter/no letter p =.204, ISR/HU p =.008
At the Start of the Interview
Awareness of Letter Variable (based on 602 cases) total R indicated saw letter at intro30 R indicated letter came to house21 total respondents aware of letter51 personally read the letter40 got more info (web site, 1-800)1 letter made a lot of difference to decision to participate 26
Data Characteristics VariableISR (n=300) HU (n=300) none (n=620) P year of birth male (%) employed (%) health excellent smoked at least 100 cigarettes (%)
Costs: Month One cost of materials: $314; staff cost: $1,919 total: $2,233 per case cost: $3.62 buys 72 interviews or 12 per HU need to estimate savings from making fewer calls, and making fewer refusal conversion calls
Conclusions HU letter (seems to) increase response and warrants consideration as a tool to improve RRFSS response rates affect on variable distributions minimal, but small sample size limits scope of examination social-political distance between respondent and sender probably matters letters may have value other than just increasing response rates
Questions