Surveys and Polls: which one you need
When you make something new in Septenar, the first choice is between a survey and a poll. In the builder they look alike — you write questions, participants answer them — but they are different instruments. They produce different kinds of data, carry different privacy properties, and suit different work. The type is set when you create the form and doesn't change afterward, so it's worth getting right at the start.
This guide covers what each one is, the single question that usually decides between them, and why Septenar calls one of them a "survey" when a stricter word would be "questionnaire."
The question that usually decides it
Most cases come down to one question: do you need to read individual responses, or only the totals?
- Only the totals — "68% chose the second option," "the average rating was 4.1" — points to a poll.
- Reading what individual people wrote, following up with them, or keeping detailed per-person records — points to a survey.
The rest of the differences follow from that one.
What a poll is
A poll collects countable answers and reports them as aggregates. Participants pick options, rate on a scale, rank choices, or answer yes/no; Septenar counts the responses and shows the tallies. Individual answers are stored so no one votes twice, but no view — not even an administrator's — shows how any single participant answered.
A few things follow from that, worth knowing before you choose:
- The question types are the countable ones — single choice, multiple choice, yes/no, scales, and rankings. There's no free-text box and no file upload; a poll can't collect a written sentence.
- Polls are short — up to ten questions.
- Results are safe to show. A tally reveals nothing about any individual, so you can publish poll results openly, in real time, or after voting — at any compliance level. "142 people chose this option" identifies no one.
- A poll keeps no personal data by default. It verifies an email to confirm a real person and prevent double-voting, then discards it, keeping only a one-way token. Once the poll closes, the key behind that token is destroyed and the responses become permanently unlinkable to anyone.
Polls fit quick community feedback, event and product votes, board and membership decisions, and municipal pulse-checks — wherever the totals are the point.
What a survey is
A survey collects individual-level responses and lets you read them. Alongside the countable question types, it can ask for short and long free text and file uploads, and it presents results as per-participant records you can open, export, and work through one at a time.
A survey is also structurally richer:
- It's organized into pages — a welcome, a consent step, one or more question pages, a thank-you — and it can branch, showing or hiding questions based on earlier answers.
- It can follow up. A survey can retain identity — or use single-use anonymous tokens where identity would put participants at risk — so you can send reminders, run the same survey again over time, and link a person's answers across waves.
- What people can see of the results depends on sensitivity. A free-text answer can identify someone by its content alone: "my supervisor did X on the 15th" names a person whether or not an email is attached. Surveys at Law 25 or HIPAA restrict results to administrators for that reason — the restriction tracks the risk in the data.
Surveys fit research studies, program evaluation, employee engagement, needs assessments, and patient-reported outcomes — wherever you need the detail, the follow-up, or the individual record.
At a glance
| Poll | Survey | |
|---|---|---|
| What you get | Totals and charts | Individual responses you can read |
| Question types | Countable only — choice, yes/no, scale, ranking | The same, plus free text and file uploads |
| Length | Up to 10 questions | Longer; organized into pages, with branching |
| Personal data | None kept by default | Email kept, or anonymous tokens for sensitive work |
| Showing results | Any visibility, any compliance level | Admins only at Law 25 / HIPAA |
| Follow-up | Not possible — responses are unlinkable | Reminders, repeat waves, longitudinal linking |
| Reach for it when | You need the totals | You need to read individuals, or follow up |
Still unsure? Go back to the question at the top: totals, or individuals?
Why "survey" and not "questionnaire"
There's a naming choice worth being open about, because the stricter word is arguably the more correct one.
In research methodology, a questionnaire is the instrument — the set of questions you put to people. A survey is the whole activity around it: designing the study, writing the questionnaire, distributing it, collecting the responses, and analyzing them. By that definition, what you build in Septenar is a questionnaire, and the survey is everything you do with it.
Septenar calls it a survey anyway. Almost everyone — researchers included, in day-to-day use — says "take our survey" and "fill out this survey" when they mean the questionnaire, and the tools people already use follow the same usage. Naming the feature "Questionnaire" would be more precise, and would make you translate a familiar idea into an unfamiliar word for no practical gain. Matching the word you already use is the better trade. The distinction is real; it just isn't one you should have to hold in your head to run a study.
("Poll" carries no such tension — it means what you'd expect: a quick count of opinion.)
Where to go next
- Creating a Survey — pages, consent, branching, results, and publishing, step by step
- Creating a Poll — the poll builder and its privacy model in detail
- Choosing your App's compliance posture — how Law 25 and HIPAA change what a survey can show
- Collecting Sensitive Data Anonymously — when a survey should use anonymous tokens instead of email