Surveys
From GSLISWiki
This page is an attempt at creating a resource built from the Survey Seminar Series by the Survey Research Laboratory to help people with doing survey research. For more information about the Lectures or about the SRL see:
Seminars:
http://www.srl.uic.edu/seminars.htm
Slides:
http://www.srl.uic.edu/seminars/semnotes.htm
This page is not completed
Contents |
[edit] Why Use Surveys?
See also: What's the point of using statistics?
Surveys are useful for obtaining information that would not otherwise be obvious. While it is difficult if not impossible to capture rich detail with surveys, surveys are an excellent tool for capturing population trends, knowing how widespread a variable of interest is, and for exposing unexpected discrepancies between populations.
Surveys can be extremely useful when used in combination with qualitative methods that provide rich details about the population, such as Interviews and Ethnography. For example: by doing a case study first, one can identify interesting variables, and then devise proper survey questions to see what the distribution of these variables are over the entire population.
[edit] Deciding on Proper Survey Design
After you decide what variables you want to research, you have to translate this research question into a survey design. This is probably the most difficult part of doing a survey, and is vital to getting reliable results. The reason that Quantitative Methods and Social Science research in general have such a bad reputation, is because many researchers do not learn how to do the design phase of their research methodologies properly. And then their results are meaningless, because you never know if the results they got can be trusted or can be generalized in the way which they suggest in their write-up.
[edit] Temporal Structure of Survey
- Cross-Sectional Surveys
- Longitudinal Surveys
- Trend
- Cohort
- Panel
[edit] Method of Survey Administration
- Personal / Face-to-Face
- Telephone
- Web / Internet
- Combination
