By the tiswa editorial team, from the official sources listed at the end of this guide. How we write our guides
Most satisfaction surveys in universities are designed once a year. They run long enough that students give up before the last page, and then the results go into a report that nobody revisits. That is a design problem rather than bad intent. This guide is for quality and student affairs staff who want a survey that answers one real question and leads to a change students can see during the same term.
What decision should the survey inform?
Before writing any question, name the decision the survey will change. Do you need to know whether the registration window suits students? Whether the library is open when they need it? Whether advising reaches the students who need it most? Every question that does not serve a decision becomes noise, and it lowers the quality of the answers around it.
Accreditation sets the same frame. The National Commission for Academic Accreditation and Assessment expects institutions to measure student satisfaction with student services and to use the results for improvement.1 The survey is a tool for deciding, not a goal in its own right.
Is Google Forms acceptable for a satisfaction survey?
It can suit a small internal poll. A satisfaction survey that collects data about students usually needs a system your data protection office has approved. The Saudi Personal Data Protection Law asks that data be collected for a documented purpose, and that the purpose be explained to the person before they answer.2 Ask legal or IT which tool is approved, and if your university already has a survey system, use it before you set up anything new.
Students notice who will see their answers. A plain line at the top saying who reads the responses, and whether they are anonymous, builds more trust than any design touch.
What do university students usually answer honestly?
Specific questions about a recent step, such as a registration form or a meeting with an adviser. Students answer general questions with a feeling they formed long ago, and they skip questions that look like a test of their loyalty.
How should you write the questions?
Ask about a specific, recent experience. "Did you find your adviser appointment time this term?" works better than "What do you think of advising in general?", because the second asks a student to report an impression formed two years ago. Use one rating scale across the whole survey. Mixing a five point scale with a ten point one makes comparison impossible.
| Question type | Example | When it helps |
|---|---|---|
| Rating after using a service | How quickly was your request handled at the admissions and registration office? | After every service, on one scale |
| Did you get what you needed | Did you find the answer you needed on the college website? | To see the gap between service and need |
| One open question | What is the one thing we should change first? | At the end of the survey only |
| Return question | Would you use this service again? | To track trust over time |
When should you send it, and to whom?
- Send it a day or two after the service
A student who finished a process last week remembers the details loosely. One who finished it this morning can tell you which step was the problem.
- Put the link in the notification itself
Do not make students search a second portal. One link inside the message is enough.
- Do not survey everyone for every service
If every student receives a survey about every service, they stop answering. Pick the service you want to improve this term.
- Close the survey after a week
Answers go stale, and the next service starts blending into the last one. Give it a clear closing date.
Do not send any survey during exam periods. A student who has just left an exam answers from their state of mind that day, not from the service.
What do you do once the results arrive?
Take a case where a survey on advising shows that most students do not know their adviser appointment times. The number alone does not say why. The times may not be published, the adviser may not answer email, or students may not read email at all. Before changing the service, talk to a small group of students, add the appointment times to the timetable in the portal, and measure again next term. That costs less than a new campaign, and the answer is more reliable.
Tell students what happened, even if it was small. A line such as "we noticed appointment times were unclear, so we added them to your timetable" is said once, and it keeps the next response rate higher.
Which mistakes spoil a survey?
- Asking many questions because each office added its own, until the survey turns into a list of requests.
- Reading only the average, which can look healthy while complaints cluster in one college or intake.
- Treating general goodwill toward the campus as proof that one service works well. A student can like the place and dislike the process.
- Publishing the overall score without saying what will change, so students stop answering surveys altogether.
Anonymous is not always the right setting. If you need to follow up on a particular complaint, a named survey with the student consent suits better, and a team separate from the service should handle it.
If you want students to know about discounts outside campus, the student deals page gathers them in one place, and a line about it can sit in the survey invitation without asking students for anything.
How does satisfaction connect to retention and activities?
A survey gets stronger when you read it with other data: grades, absences and re-registration for the next term. If satisfaction with advising rises while retention stays flat, the problem is probably somewhere else. The student retention guide lists the signals worth tracking next to a survey.
If the survey covers activities, ask about participation itself, not only the event you ran. The student engagement guide explains how to read those answers.
For wider campus life, such as the library, housing and transport, the questions in the university quality of life guide can save you writing new ones each term.
Should the same questions run every term?
Yes, and that is the point. The value of a survey shows when you compare one service with its own past, not when you compare advising with the library. Keep a few questions fixed every term, and change the rest according to what you need to know.
In colleges where numbers move a lot from term to term, do not read a small change as a finding. Wait until the same result shows up twice before you announce a big decision on it.
How long should the survey be?
Short enough that a student finishes it in a minute or two while standing in a queue. Some teams keep a fixed core and add one or two questions that change each term. The middle ground is a short survey that repeats, rather than a long one that appears once.
