Research Support
Survey Design
Bad questions produce clean-looking data that means nothing.
A survey is an instrument, and an untested one produces numbers that look fine and measure nothing. Half the questionnaires we see ask two things in one item, or lead the respondent to an answer, or run so long that the last section is answered at random. We build items from your variables, pilot them, and fix what the pilot exposes.
What you get
- Items written from your variables, not from instinct
- Scales that are consistent and actually tested
- Leading and double-barrelled questions removed
- Length kept where response rates survive
- A pilot round before you send it to anyone
Related reading
Referencing
APA 7th Edition
The details that cost marks: capitals, italics, ampersands and DOIs.
Read more →
Thesis & Dissertation
Discussion Chapter
Where your results stop being numbers and become findings.
Read more →