Find Survey Participants: 5 Channels and IRB Review

The questionnaire is done, the link works, and three days later you have 14 responses. Finding survey participants is where most surveys stall, and the questionnaire is rarely the reason.

Participants come from five places: the people you already know, topic groups on social media, survey exchange platforms, an institution such as a university, a club or an employer, and a QR code posted where your target group already stands. Speed does not decide which of the five fits. Your target group does. In the United States one step comes before all of them, because research covered by the Common Rule needs review by an Institutional Review Board before recruitment starts. By the end you will know which route suits your study, what each one costs you in data quality, and how to write an invitation people act on.

📌 The key points at a glance

• Common Rule research needs IRB review before you recruit.
• Your own circle answers fast, but rarely matches your target group.
• Exchange platforms bring point collectors, not interested respondents.
• AAPOR asks you to disclose probability versus non-probability sampling.
• A prize drawing by email ends your survey's anonymity.

How to find survey participants: which route fits your study

The right route depends on a question you have to settle before you start looking: are you testing a relationship between two variables, or are you trying to say what share of a group thinks something? If you are testing a relationship, a skewed sample is a nuisance. If you are reporting a share, it is the whole ballgame.

A sample of volunteers who signed themselves up is a non-probability sample. Nobody drew those people from a known frame with a known chance of selection, so the numbers describe the people who answered and nobody else. Eight hundred responses do not change that. For hypothesis testing such a sample is often good enough, and in a thesis it is the normal case.

The second question is more mundane, and it gets skipped just as often: where does your target group already spend its time? Everything else follows from the answer.

⚠️ Clear the IRB before you recruit

Research with human subjects that falls under the Common Rule has to go through an Institutional Review Board before recruitment starts. Posting the link first and asking afterwards can cost you the whole data set. Your institution's IRB or your faculty advisor, not you, decides whether the rule reaches your study.

How large your sample needs to be depends on the analysis, not on the channel. The article on the representative survey works through that calculation.

Your own circle: fast, but one-sided

Friends, family and classmates are the fastest channel, because you can reach them directly and nobody needs convincing that the request is real.

For the first twenty responses that is the right move, and for the pretest it is the only sensible one. The catch is that your own circle looks a lot like you: similar age, similar region, similar education, often the same major. Those are exactly the traits your analysis needs variation in. A survey that is three quarters your own cohort measures your own cohort.

Treat your own circle as a deliberate test run. Ask three people to fill in the questionnaire and to tell you afterwards where they hesitated, and message them one by one rather than in a group chat, because in a group everyone waits for somebody else to go first. That turns up more errors than rereading the draft yourself ever will. How to spread the link in messenger groups after that is covered in the article on the WhatsApp survey.

Social media groups: where your target group already is

Social media delivers participants when you post in groups built around your topic rather than groups built around surveys. That one difference decides whether the data is worth anything later.

A post on your own timeline mostly reaches people you have already asked. The places worth your time are the ones where your target group has already sorted itself by subject: the Facebook group for hobby beekeepers, the forum for physical therapists, the subreddit for your campus or your topic, the LinkedIn group for procurement, the Slack or Discord community for a professional niche. Those people have a stake in your question, and they answer differently from someone who is only collecting points.

What makes the difference in practice

• Read the group rules before posting, since many boards ban surveys outright.
• Name the target group in the first line, so the wrong people never click.
• Post to working professionals on weekdays, not on weekends.
• Answer questions in the comments, which keeps the post visible.
• Send one reminder after five days, and only one.

Naming the target group in the first line feels like giving away reach. In practice the opposite happens: people who do not belong never click, never quit halfway, and never dilute your data. Responses that start and stop halfway are what make every analysis painful later.

Survey exchange platforms: what you trade away

Survey exchange platforms run on a simple trade: you answer other people's questionnaires, collect points for it, and spend those points on responses to your own survey. Give a lot, get a lot, and the mechanism works surprisingly reliably.

The limit is built into the same trade. On the other side of it there is almost always somebody writing a thesis who needs points. Your sample then consists of students who care about the exchange rate rather than about your topic. In methodological terms that is a self-selected volunteer sample, which puts it squarely in the non-probability camp.

American survey research asks you to say so out loud. The Code of Professional Ethics and Practices (2020) of the American Association for Public Opinion Research requires researchers to “Explicitly state whether the sample comes from a frame selected using a probability-based methodology (meaning selecting potential participants with a known non-zero probability from a known frame) or if the sample was selected using non-probability methods (potential participants from opt-in, volunteer, or other sources).” One honest sentence in your method section settles it.

The decision rule is short:

✓ Your target group is students or a broad general public
✕ Your target group is a specific profession, an industry or an age group past 40
→ In the second case the exchange platform costs more time than it returns

If you do use an exchange platform, go through the data afterwards. Someone clicking for points clicks fast: one answer repeated down a whole matrix, completion times far below everyone else's, no open text answers at all. Drop those cases, and say in the write-up why you dropped them.

Institutions and IRB review: the route almost everyone forgets

The route through an institution brings you exactly the target group you need, because somebody there has already done the sorting for you. Student government knows everyone in the major, a club knows its members, an HR department knows the staff.

What that takes is a request that creates no work on the other end: a finished text ready to forward, the link, the length, the name of your faculty advisor and the date the survey closes. An email that only needs forwarding gets forwarded. An email that asks somebody to write something first does not. Department listservs work the same way, and whoever runs the list will ask for your exact wording anyway.

Print is the second route. A QR code on a flyer, a bulletin board, a lecture slide or the counter at a club event catches people in the place your topic belongs to. At empirio.ai, an online survey tool from Germany, you can create a QR code without needing any extra software for it (as of August 2026). What matters when printing and scanning is covered in the article on the survey with a QR code.

IRB review comes before recruitment, not after

Institutional Review Boards are not a formality you can catch up on later. Under 45 CFR 46.109, IRB review of research, “An IRB shall review and have authority to approve, require modifications in (to secure approval), or disapprove all research activities covered by this policy”. The approval criteria are written down as well. In 45 CFR 46.111, criteria for IRB approval of research, paragraph (a)(3) states that “Selection of subjects is equitable.” Recruiting only the people who happen to be easiest to reach is therefore more than a statistical weakness.

Be precise about the scope, because the rule does not cover every questionnaire ever written. The regulation applies to research covered by the policy, and for that research the review comes first. Many institutions go further and apply the policy to every project involving human subjects, funded or not. Whether your survey falls under it is not a decision you make yourself: your institution's IRB or your faculty advisor decides. Ask before you write the invitation, not after the first responses arrive.

⚠️ Caution

Surveys that run through an employer, a school or a public agency follow that organization's own rules on top of any IRB decision. The normal case is a written sign-off before you send anything. The organization's privacy or legal team decides, and at a university your faculty advisor does. What else applies to data protection in surveys is covered in a separate article.

Nothing in this article counts as legal advice. The binding answer for your own study comes from your IRB, your faculty advisor or your organization's counsel, and it is worth having in writing before the invitation goes out.

Incentives and prize drawings: where most guides get it wrong

The raffled gift card in nearly every sample invitation online is the weakest incentive survey research knows. A prize drawing is not cash, and you only get it if you take part. Both of those weaken it.

American research practice adds a second point: whatever you hand out, you have to report it. The AAPOR Code of Professional Ethics and Practices asks researchers to “Describe any compensation/incentives provided to research subjects and the method of delivery (debit card, gift card, cash).” Write that sentence while you are planning the incentive, not while you are assembling the appendix.

How much the incentive type actually matters

Incentive research sorts the forms along two axes: cash or goods, and paid up front to everyone or afterwards only to participants. Klaus Pforr summarizes the evidence in the GESIS Survey Guidelines on incentives (2015), a German-language source, and the finding is consistent: cash and unconditional incentives produce higher response rates than non-monetary and conditional ones. A prize drawing sits on the weaker side of both axes.

Pforr cites several international meta-analyses for this. Church (1993) evaluated 38 mail studies and found that unconditional cash incentives raise the response rate by 19 percentage points, while gifts raise it by 8 percentage points. Edwards and colleagues (2002) conclude across 292 studies that unconditional incentives increase the odds of participation by 71 percent compared with conditional ones. For web surveys the effect is smaller: Göritz (2006) finds a gain of roughly 20 percent there.

The numbers come mostly from mail surveys and do not carry over one to one to an online survey. The ranking of the forms holds all the same. One sentence from the same source matters more than any response rate: an incentive helps data quality only if it lifts response without making nonresponse more selective. If your gift card mainly attracts people with little money and plenty of time, you end up with more answers and a more lopsided sample.

A prize drawing and anonymity do not fit in the same form

Running a prize drawing requires a way to contact the winner, usually an email address. Collecting that address in the same questionnaire ends the anonymity, whatever the invitation promised, because every response is then attached to a person. That exact combination shows up in countless sample invitations, often in the same paragraph as the promise of anonymity.

Separating the two fixes it. Put a second link on the thank you page that opens its own short form, and collect the address for the drawing only there. The two data sets have no connection to each other, and the study stays an anonymous survey.

United States law has no single federal privacy statute for surveys comparable to European data protection law, so do not expect one rule to settle this. The one verifiable anchor is the IRB criterion in 45 CFR 46.111(a)(7): “When appropriate, there are adequate provisions to protect the privacy of subjects and to maintain the confidentiality of data.” Your IRB or faculty advisor decides in your case, and inside a company the legal or privacy team does.

What works better than a prize drawing and costs nothing

An incentive never replaces a reason. More effective than five raffled gift cards: a questionnaire under five minutes, a named person instead of an anonymous link, one sentence about what the results will be used for, and an offer to send the summary once the study is finished. That last offer costs one email and often wins over a professional group faster than money.

The survey invitation: five facts, or nobody clicks

An invitation to a survey needs five facts: who it is for, what it is about, how long it takes, what happens to the data, and by when you need answers. Anything more is decoration, anything less costs you responses.

Length is the one people guess at instead of measuring. “Only takes 5 minutes” appears in almost every call for participation, and when it turns out to be twelve, people quit halfway and leave you with unusable responses. Run through your own questionnaire once with a stopwatch, then state the number you measured plus a small buffer. Add the number of questions, which reassures people more than any minute count.

An opening that works

Looking for people who bought a bicycle in the last twelve months. 14 questions, about four minutes, anonymous, for my thesis at Ohio State University. Open through October 15.

The first sentence sorts, the second removes every excuse. People who do not belong stop reading, and that is exactly right. How to build the matching questions behind it is covered in the guide to creating a questionnaire.

Common mistakes when recruiting participants

Anyone who gets too few responses usually goes looking for more channels. The problem is often not reach at all, but one of the following four things.

All four show up only once the survey is live, and by then the repair costs far more than a check beforehand would have. Half an hour with this list before you write the invitation is the cheapest insurance in the project.

You look for participants although the questionnaire is the problem

A steep drop-off is a questionnaire problem, not a reach problem. Plenty of people open the link and few arrive at the end, and no extra channel fixes that. The questionnaire is losing them, usually at a required question, at a long matrix or on the third page. Look at which question the drop-offs happen on. At empirio.ai the drop-off rate shows exactly that, from the Lite plan upwards (as of August 2026).

You post everywhere at once and cannot tell afterwards what worked

Spreading the link across eight groups in one afternoon makes it impossible to say later where the responses came from. That is a real loss for the analysis, because the source often explains the differences in the data. Add one short question instead: “How did you hear about this survey?” Ten seconds of work saves you a paragraph in the method section.

You never remind people, or you remind them constantly

One single reminder after about five days almost always produces a second wave, because many people saw the link, meant to take part and then forgot. Two reminders feel pushy, three are annoying. Keep it short, skip the guilt trip, and repeat the closing date.

You do not write down how you found the participants

Recruitment belongs in the method section, not in a footnote. If you recruited through friends, groups and exchange platforms, you collected a convenience sample, and the thesis or capstone project should use exactly that term. AAPOR asks for the same disclosure in professional practice: probability or non-probability, and what the incentive was. Add a sentence or two in the limitations about what that means for how far your results reach.

Conclusion

Too few participants is usually a target group problem, not a reach problem. The fastest channels deliver the most lopsided samples, and the slowest route through an institution often delivers exactly the people the study is about. Work out where your target group already is, then pick the channel. In the United States, settle the IRB question before both.

One limit belongs in the same breath: empirio.ai does not supply participants. There is no panel and no community to pull responses from. The tool covers everything before and after, from the questionnaire to the analysis, and the search for people stays your job.

Where to go next

• Wondering how many responses are enough? When a survey counts as representative
• Need the right selection method? Drawing and describing a sample
• Questionnaire not written yet? Creating a questionnaire: the guide
• Want the study to stay genuinely anonymous? Creating an anonymous survey

Ready to send out your survey?

Link, QR code and website embedding are part of the free plan at empirio.ai, and so are the first 50 participants per month.

Create a survey for free