substitute, sample, stagger, share
Establish Research Questions
- Research questions guide and center your research.
- Are you assessing an intervention, gauging interest in a program?
- Does each question in your survey serve a purpose for your research?
- What broader campus goals or initiative does this research support?
- Keep in mind your analysis plan when designing research project and project scope.
Choose the right methodology
- Do the people you plan to survey have the information needed/is their opinion informed?
- Are there other data sources that could be used (analysis of existing data, literature reviews, analysis of administrative data, Cal Answers)
- Will you need to disaggregate data for small sub-populations?
- Are other methodologies better suited for your research questions such as focus groups or interviews?
- Consider using a sample rather than the population.
- Surveys with an expected response rate below 15% are typically not encouraged.
- Consider requesting adding questions to UCUES, SONS, or other Tier 1 surveys in lieu of conducting a one-off survey.
- OPA will work with you to provide a non-overlapping sample, when possible.
- Tier 2 and 3 surveys may have samples that overlap with Tier 1 surveys.
- Check the Tier 1 survey schedule to ensure that your survey is not in conflict with other key campus surveys.
- If a Tier 2 or 3 survey conflicts with a Tier 1 survey we can work with you on scheduling around planned survey collection dates to minimize overlap.
- Tier 1 Surveys will receive highest scheduling priority.
- Check with OPA for other survey activities that might conflict with your planned survey.
- Test, test, test
- Pilot with a small group of actual respondents, if possible.
- Consult with Committee for Protection of Human Subjects to determine if Institutional Review Board approval is necessary.
- Use multiple e-mail reminders with varied signatories and content.
- Send e-mail reminders only to non-responders.
- Consider use of social media (e.g., departmental Facebook pages)
- Use incentives, if possible (especially with student populations!); for example, Amazon gift cards, Cal1 Card credit.
- When possible, use campus-approved survey software for data collection (e.g. CASES, Qualtrics, Google Forms).
- If possible, use CalNet authentication for survey access. Be sure to provide a means to logout of CalNet.
- Lock access to completed surveys to prevent unauthorized re-entry.
- Store data securely.
- Implement a strategic email reminder plan
- Monitor response rate (possibly send targeted emails to group(s) with lower response rates, provide additional incentives, leave survey open for longer or consider increasing sample size)
- Clean data (e.g., handle missing data, straight-lining, partial completions).
- Prepare a codebook.
- If appropriate, share aggregated results.
- Suppress small cell sizes in any public report.