Survey guides, grouped by the decision they settle
Sixteen guides in three groups, and 109 ready-made surveys. Each card names the decision the guide settles, how long it takes to read and when it last changed, so you can take one answer and go back to your survey.
Choose the questions
Four formats cover everything in the SuperSurvey template library, and two of them cover 81.6 per cent of it, so the two scale guides come first. The counts are below the cards.
Score a Likert battery
Read this when a block of Likert scale statements has to add up to one score: how many items, which end labels, and what you may average.
Choose scale labels
Read this when you need the exact wording for a single 1-to-5 or 0-10 rating scale question, and the label for every point on it.
Multiple choice questions
Read this before you write the answer options. How many, in what order, and what to do about Other.
Open-ended vs closed questions
Read this when you are about to add another text box.
Plan the collection
Six guides on the work either side of the questionnaire, in the order the decisions come up: what the study is for, how it is structured, how the questions are worded, then the two things that go wrong once real people start answering.
Run a survey end to end
Start here if you have never sent one. It is the only guide that covers every stage.
Survey research design
Read this when the method itself will be questioned, not just the result.
Questionnaire design principles
Read this once the questions exist and the argument moves to their order and length.
How to write survey questions
Read this when you have a blank page and want wording you can lift, rather than theory.
Demographic questions
Read this when you want to cut the results by age or role and are unsure what you are allowed to ask.
Response bias
Read this when the answers look too good.
Analyse the results
Six guides. The first two belong before you send, while the invite list can still change; read them with the plan, not after the answers arrive. The other four settle what a result is allowed to mean. Two of the six run past half an hour, so use their contents lists.
How many responses you need
Read this before you send, while you can still widen the invite list. It opens with a calculator.
Sampling methods
Read this when someone asks who you surveyed and who you left out. A reference, not a sitting.
Statistical significance
Read this when two groups differ and you have to say whether the gap is real or noise.
Define quantitative data and analyse it
Read this when the answers are in a spreadsheet and you are not sure which columns are numbers and what you may average.
Correlation between survey items
Read this when two answers move together and someone in the room starts saying one caused the other.
Regression analysis steps
Read this when one outcome has several plausible drivers and you need to weigh them against each other.
Which guides to read first
Most people arrive with one decision to make and leave once it is made. Starting from nothing is the exception, and these three are the order for it: 84 minutes between them.
-
Get the shape of the whole thing
Run a survey end to end takes 12 minutes and is the only guide here that covers every stage, from writing down the decision you are trying to make through to reading the answers back. Read it before you write a question.
-
Fix the wording before the format
Wording moves answers further than any format choice does, so how to write survey questions earns its 35 minutes even though it is the longest of the three. Start at the examples.
-
Work out how many replies you need
This one cannot be repaired afterwards. How many responses you need opens with a calculator, so the number itself takes about a minute to get, and the rest of its 37 minutes is about defending that number to whoever asks for it.
Two of the sixteen are references rather than reads. Sampling runs 47 minutes and the questionnaire guide 33. Open either at the section you need and close it again.
Do not leave the four analysis guides until the answers are in. Decide the variables and the comparisons before you field the survey, because a comparison you did not plan for is one the questionnaire probably did not collect. Ten minutes with the first page of each is enough to know what to build in.
The survey checklist
Eleven checks, in the order they come up, on one sheet. Copy it into the brief or print it; the page prints without its navigation.
Survey checklist
Write the decision the results will change in one sentence. If no decision changes, do not send the survey.
List every comparison you will make (which groups, on which answers) and the variable each one needs. Add the question that collects it now.
One idea per question. Split anything joined with "and".
Answer options cover every case, do not overlap, and offer "Other" or "Prefer not to say" where an honest answer might not fit.
One scale direction and one set of labels for the whole block. Label every point, the middle ones included.
Demographic questions last, and only the ones a planned comparison uses.
Work out how many responses each comparison needs before you send, and check the invite list can supply them.
Send the draft to a few colleagues, watch where they stall, and cut every question that does not feed a comparison.
Send from a named person with a deadline, and plan one reminder to the people who have not answered.
Before comparing groups, compare response rates by group. A group that answered less is a different group.
Print the count behind every percentage, and report a difference with its interval or test, never the point estimate alone.
Why the scale guides lead
What the 1,422 questions are
Multi-select is the interesting absence: 41 questions out of 1,422 let people tick more than one box. Counts from a multi-select do not add up to the number of respondents, so every chart drawn from one is harder to read, and the library reflects that.
Sample surveys by category
The guides tell you how to build a survey. The library holds 109 that are already built, each one editable and sendable as it stands. The median one asks 13 questions.
| Category | Templates | Median length |
|---|---|---|
| employee survey templates | 33 | 13 |
| satisfaction survey templates | 22 | 12 |
| business survey templates | 20 | 14 |
| marketing survey templates | 13 | 13 |
| technology survey templates | 11 | 14 |
| training survey templates | 10 | 11.5 |
| All six categories | 109 | 13 |
Source: SuperSurvey Question Corpus, 109 live templates and 1,422 questions, read 11 September 2026. Median length is the median number of questions in a template in that category, printed exactly: 11.5 means the two middle Training templates have 11 and 12 questions.
Employee surveys are 33 of the 109. If your subject is not one of the six, take the closest category anyway, because the wording inside a template carries a great deal further than the label on the front of it does.
How we counted
Read times are each guide's own content word count at 240 words a minute, rounded to the minute. The count takes the page's article body only, with scripts, styles and chart graphics removed, and never the site navigation. For the guides rebuilt on 11 September 2026 it is taken from the page as built; for the others, from the live page as read on 11 September 2026. The times describe reading a page front to back, which is not how anyone uses the two long references. Update dates are each guide's own.
Template and question counts come from the live library, parsed template by template on 11 September 2026. Draft rows and redirect rows were excluded, leaving 109 templates and 1,422 questions. Those figures describe what one survey tool publishes as good practice. They are not a claim about surveys in general.
One person writes and edits every guide listed here, and that is Michael Hodge. There is no second reader. Where a guide leans on a source, the source is named in that guide's own reference list so you can check the claim instead of taking it on trust.
References
- SuperSurvey Question Corpus, 109 live templates and 1,422 questions, read 11 September 2026.
- Read times: content word counts of the sixteen guide pages at 240 words a minute, rebuilt pages as built on 11 September 2026, the rest as read live on 11 September 2026.
Michael Hodge: Survey methodologist and editor, SuperSurvey. Bachelor of Science (Psychology), University of Wollongong, with coursework in psychometrics and research methods. Designing surveys since 2003. About the author and how these guides are reviewed
Stop reading and send one
The quickest way to find out whether a question works is to put it in front of ten people and watch where they stall.
Create your survey with SuperSurveyOr open one of the 109 ready-made surveys and change the questions that do not fit.