The University of Michigan Energy Survey uses a set of 18 questions asked quarterly as a part of the monthly University of Michigan Surveys of Consumers (“SCA”). The energy survey therefore shares the sample design and other methodological attributes of the SCA.
The survey is nationally representative and is carried out by conducting telephone interviews with adult men and women from a monthly sample of approximately 500 households in the coterminous United States (48 states plus the District of Columbia). The SCA uses random-digit dialing to contact respondents and, in order to maximize its ability to observe changes in attitudes over time, involves a rotating panel sample design. For each monthly sample, an independent cross-section representing 200 of the total 500 households is drawn. These 200 respondents are then re-interviewed six months later. Therefore, each month’s sample in is comprised of 60% new respondents and 40% who are being interviewed for a second time, six months after they were first interviewed. Because it is conducted quarterly, the energy survey also shares the rotating panel design. Further details on the method can be found in the SCA Sample Design documentation.
The energy survey questions were developed through a nearly two-year process starting with an in-depth review of existing and prior energy surveys and consultations with energy experts. That preliminary phase was then followed by a formal questionnaire development process that involved focus groups, cognitive interviews and pre-testing. The energy survey was launched as part of the October 2013 SCA and subsequently has been fielded every three months in January, April, July and October.
See the “Energy Survey Drills Deep” article for a summary of the process. Further details can also be found in our initial report and in the peer-reviewed paper, “U.S. consumer attitudes and expectations about energy,” published in the journal Energy Policy.