Common sense and other not-so-obvious checks
When conducting or critiquing a secondary analysis of polling data, the most obvious technique the analyst must apply is common sense.
- Always look carefully at the question that was asked. If the question is confusing to you, it probably was to at least some of the respondents.
- When the results appear surprising, don’t just report them-question them! Investigate why, check out the methodology, the sampling, the question wordings, and compare the results to other surveys on the topic.
- Determine if there is something missing. Has the topic been adequately explored in the surveys being used in the analysis? Is there a different angle that would enhance the study? If so, seek other sources.
- Make sure you are considering the most appropriate sample for investigation of your topic. If you are looking at attitudes about public schools, for instance, you may need to consider opinions of both the general public and parents of school-aged children.
- Do your homework! Always check out published works on the topic. Public Opinion Quarterly and Public Perspective Online are excellent sources of sound, digested analyses on a multitude of topics.
- In reports and releases, watch out for reports of partial interview dates. If the survey isn’t completed, it’s likely that the requirements of sampling frame have not been met and generalizations may not be possible.
For further information please contact The Roper Center at 607.255.8129 or firstname.lastname@example.org.