Appendix B: Detailed Methology AARP Multicultural Survey
C. Response and Refusal Rates for AARP Boomer Sandwich Survey
The response and refusal rates calculated below are based on the Standard Definitions of the American Association for Public Opinion Research (AAPOR) for the final dispositions of cases codes and outcome rates for RDD telephone surveys. The Standard Definitions are taken from AAPOR’s website (www.aapor.org) as of May 3, 2001. We also take into account the recommendation from the AAPOR Council to provide information about what happened to every element in the sample to permit a knowledgeable consumer to calculate any of the different rates described at the end of the Standard Definitions document.
While AAPOR provides standard definitions for disposition codes for sample units, it does not provide a singular standard response or refusal rate formula. AAPOR suggests a number of outcome dispositions consisting of six formulae for response rates, three formulae for refusal rates, four formulae for cooperation rates, and three formulae for contact rates. We provide the information so that any of the formulae can be used. We also calculate response and refusal rates for each of the national sample, and each of the over samples based on one of the formulae offered by AAPOR for determining response and refusal rates.
While the over samples were targeted at specific race or ethnic groups (African-American, Hispanic, and Asian-American), respondents who met the age and minority status required to be eligible for the survey were interviewed regardless of the sample used to contact them. Therefore, the response and refusal rates apply to the sample frames and not to the ethnic groups that the frames targeted.
Sample Unit Final Dispositions
The accompanying table displays the final disposition outcomes and codes for the sample units (telephone numbers) used for each of the samples in the study. The nature of each of the samples is slightly different (see section on Sample Design) so rates can only be calculated for each of the samples. Therefore, there is no overall response or refusal rate for the study. The codes are those designated by AAPOR and will be used in the formulae for determining response and refusal rates.
The final dispositions are fairly clear. The sampling units produced completed interviews, refusals, non-contacts (eligible persons unavailable to be interviewed), other cases where there was an inability to interview eligible persons (usually due to a language barrier), numbers with unknown statuses (no answer or busy on every contact attempt for up to ten attempts in some cases), and ineligible cases. This information can be used to calculate any of the AAPOR rates for outcome distributions.