Designing Effective Surveys for Postgraduate Research in Australia

Dr Michael Torres · 2024-06-20 · 8 min read · Research Methods

Best practices for survey design, distribution, and analysis in the context of Australian postgraduate research.

Introduction

Surveys are one of the most common data collection methods in Australian postgraduate research. A well-designed survey produces reliable, valid data; a poorly designed one wastes time and produces meaningless results. This guide covers best practices for survey design.

Planning Your Survey

Define Your Research Questions

Your survey should be directly linked to your research questions. For each survey item, you should be able to explain which research question it addresses.

Choose Your Mode

  • **Online (Qualtrics, REDCap, SurveyMonkey, Google Forms):** Most common; cost-effective; easy distribution; can have low response rates
  • **Paper-based:** Useful for populations with limited internet access
  • **Phone:** Higher response rates but expensive and time-consuming
  • **Mixed-mode:** Combining online and paper to maximise reach

    • Most Australian universities provide free access to Qualtrics or REDCap. Check with your IT services.

    Designing Questions

    Question Types

    **Closed-ended (quantitative):**

  • Likert scales (e.g., 1 = Strongly Disagree to 5 = Strongly Agree)
  • Multiple choice
  • Ranking questions
  • Yes/No questions
  • Demographic questions

    • **Open-ended (qualitative):**

  • Free-text responses
  • Use sparingly — they're harder to analyse
  • Useful for capturing unexpected insights

    • Writing Good Questions

    **Do:**

  • Use clear, simple language
  • Ask one thing per question (no "double-barrelled" questions)
  • Provide balanced response options
  • Include "Not applicable" or "Prefer not to say" options where appropriate
  • Use Australian English and terminology

    • **Don't:**

  • Use leading or loaded questions ("Don't you agree that...")
  • Use jargon or abbreviations without explanation
  • Assume knowledge the respondent may not have
  • Use double negatives
  • Make questions unnecessarily long

    • Common Mistakes

    **Double-barrelled question:** "How satisfied are you with the quality and timeliness of feedback?" (These are two different things)

    **Leading question:** "Most experts agree that X is important. How important do you think X is?" (Biasing the response)

    **Ambiguous question:** "Do you regularly attend lectures?" ("Regularly" means different things to different people)

    Validated Instruments

    Before creating your own survey, check if a validated instrument already exists:

  • Search the literature for previously used and validated scales
  • Using validated instruments strengthens your methodology
  • If you modify a validated instrument, document and justify the changes
  • Report reliability statistics (Cronbach's alpha) for your sample

    • Survey Structure

    Logical Flow

    1. **Introduction:** Purpose, estimated time, confidentiality statement, consent

    2. **Screening questions:** Ensure respondents meet inclusion criteria

    3. **Easy, engaging questions first:** Build rapport

    4. **Sensitive questions in the middle:** After trust is established

    5. **Demographics last:** These are easy but feel intrusive if asked first

    6. **Thank you page:** Express gratitude and provide contact information

    Length

  • **Aim for 10–15 minutes** completion time
  • Longer surveys have higher dropout rates
  • Every question should be justified — remove anything that doesn't address a research question

    • Pilot Testing

    **Always pilot test your survey** before distributing it widely:

  • Test with 5–10 people similar to your target population
  • Ask about clarity, length, and any confusing questions
  • Check the time to complete
  • Review the data produced — can you analyse it meaningfully?
  • Refine based on feedback

    • Maximising Response Rates

    Strategies

  • **Personalise invitations** — Use names where possible
  • **Send reminders** — 2–3 reminders, spaced 1–2 weeks apart
  • **Keep it short** — Respect respondents' time
  • **Offer incentives** — Prize draws are common and usually HREC-approved
  • **Use multiple distribution channels** — Email, social media, organisational networks
  • **Ensure mobile compatibility** — Many respondents will use phones

    • Typical Response Rates

  • Student surveys: 20–40%
  • Professional surveys: 10–30%
  • General population: 5–15%
  • Calculate whether your expected response rate will give you an adequate sample size

    • Ethics Considerations

  • Obtain HREC approval before distributing
  • Include a Participant Information Sheet
  • Ensure informed consent (usually a checkbox before the survey begins)
  • Guarantee anonymity or confidentiality
  • Store data securely on university-approved systems
  • Provide contact details for complaints

    • Conclusion

    A well-designed survey is an investment in the quality of your research. Take the time to plan, design, pilot, and refine before distributing. DissertlyPro can assist with survey design, statistical analysis of survey data, and writing up your methodology and results.