Survey Research at ISPR
The Institute for Survey and Policy Research (ISPR) provides consulting on all aspects of survey research to U.W.M. faculty, staff, and students and to local governments, non-profit agencies, and businesses. We offer advice on the type of survey technique to employ (e.g., telephone, mail, or in-person), and on questionnaire design, sampling, statistical analysis, and report preparation.
In addition to providing consulting services, ISPR conducts telephone, mail, and in-person surveys from start to finish -- that is, from the questionnaire design stage to the preparation of a final report.
- The capacity to handle hierarchical, sometimes called rostered, survey instruments in which one or more sets of questions are asked about more than one person or item.
- The ability to compute frequency distributions of survey results almost immediately after the last interview has been completed.
- The ability to automatically deliver the most appropriate prospective respondent to the next available interviewer, by taking into account such criteria as the number of contact attempts and time-zone differences. The resulting increase in efficiency makes it possible to conduct large nation-wide surveys that would otherwise be prohibitively time-consuming and expensive.
For telephone surveys, we use a state-of-the-art Computer-Assisted Telephone Interviewing (CATI) system. The ISPR CATI lab consists of several Pentium-based microcomputers, each of which is linked to a central server. Each CATI station has a dedicated telephone line.
When using a CATI system, telephone interviewers read the survey questions from a computer screen rather than from a paper questionnaire, and enter the responses directly into the computer.
Among the advantages of the CATI system used by ISPR are the following:
- A high degree of accuracy because responses are entered directly into a computer file rather than being entered as a second step, and because the system won't allow an invalid code to be entered.
- The ability to utilize extremely complex contingency patterns of questions.