Likely yield for pairs experiment

I am planning an experiment using Prolific’s ability to recruit pairs of participants in the United States. To do so, I will use the screeners Participating together with a friend on Prolific and Participating together with a romantic partner on Prolific.

There are currently 4,205 people with friends on Prolific, and 1,907 with a romantic partner in the USA active in the past 90 days. I have read Prolific’s help page on studies with couples, and I am wondering how many pairs of participants I can reasonably expect to recruit.

Can anyone share their experience trying to recruit friends/couples on Prolific, and their sample sizes? This info would be valuable for study design and power calculations.

Some relevant info:

  • Will only have two surveys (initial screener + main study) and little delay between them
  • Short, likely less than 5 minutes total (screener + main study) for median participant
  • Well compensated: will target $15-20 / hour plus a small bonus
  • I’m willing to leave the study running for an extended period (4–6 weeks, maybe even longer)
  • Asynchronous. No need for participants to meet live or even record anything
  • No other screeners besides USA and participating with another person
  • Initial screener won’t eliminate anyone (just preliminary info gathering)
  • Will be asking about political views on a hot-button issue (might turn off some people?)
  • No deception

How many pairs can I get?

Welcome back Chris

I am afraid I have never done an experiment involving pairs.

I would like groups of six. Thinking about this, you may find
that if you did not get enough pairs using the participating
together with a friend or romantic partner on prolific you
might find that a prescreener of your own for more people
that know another prolific participant may yeild some more.

But basically I have no experience, though as a participant
I have been asked to participate with my wife in the past so
there are people who are using the feature.


1 Like