New Experimental Modules for HRS 2008 Invited

November 2006

The HRS investigators welcome suggestions for experimental modules from the research community. Ideas for modules should be sent to the committee on experimental modules at the address below. It is not necessary to provide specific survey questions; just the basic analytic idea, with some indications of the variables that the module should try to measure, will be sufficient. It is also helpful to include references to recent publications on the topic, other studies that have treated the topic, names of persons familiar with the topic, or specific survey instruments. Most modules are written by HRS investigators, often in collaboration with outside consultants. However, modules designed entirely by outside researchers will be considered. All modules must conform to the same overall design limitations, and the academic community, as well as the public, will have access to the data at the same time as the authors and investigators. The HRS investigators will make the final decision about which modules to include in each wave of interviews.

Experimental modules are administered to randomly-selected sub-samples of HRS self-respondents, after the main interview is completed, with a strict limit of two to three minutes in average length. Each respondent is randomly assigned to one and only one module. Respondents can refuse to participate in any module before assignment, and can refuse to answer questions in the module to which they are randomly assigned, but interviewers cannot offer nor can respondents choose to respond to a module other than the one selected at random. The sample size for each module depends on the number of modules included. For example, in 2006 there were 10 experimental modules to be divided among an expected 17,500 self-respondents, yielding an average of about 1750 persons asked to do each module. Response rates to modules have averaged about 80% in prior waves, which would leave about 1400 responses in each module.

In order to preserve the random selection of participants in all modules, it is essential that each module be designed for the entire HRS sample. Modules that target specific groups by age, sex, health status, or any other variable do not fit well within this structure. If one module were asked of all 51-56 year olds, for example, then the other modules would not represent that age group, with negative consequences for all of them. It is possible, however, to construct different sequences of questions for different groups within a module. For example, a module about cancer knowledge could ask women about breast cancer and men about prostate cancer. Also note that because individual respondents are assigned randomly, it is not possible to require husbands and wives to answer the same module.

The timetable for module development for the 2008 interview wave is as follows:

  • January 29, 2007: Proposals due.
  • February 23, 2007: Final selection by HRS co-investigators.
  • June 4, 2007: Final question text and specifications due.
  • May, 2009: Preliminary release of data from HRS 2008.

Suggestions for modules should be sent via e-mail to: Please include the word "module" as the first word in your subject heading.

If you prefer to submit written suggestions, they may be sent to:

     HRS Committee on Experimental Modules
     Institute for Social Research
     University of Michigan
     426 Thompson St., Rm. 3050
     Ann Arbor, MI 48106-1284

If you have any questions, please contact the Help Desk, or send e-mail to