Research participant recruitment: advice and information

Information on research, statistics and publications - tips including how to recruit participants, gain funding, understand your results and get them published.
Post Reply
User avatar
Posts: 1439
Joined: Thu Apr 05, 2007 9:41 am
Location: Midlands

Research participant recruitment: advice and information

Post by choirgirl »

Advice and information for research participant recruitment difficulties

One of the common difficulties when designing and running a research project is data collection. This Wiki aims to give some information and handy hints based on other ClinPsy members’ experiences.

• First up, don’t panic! If your response rate is low, or in fact for any issue in your research causing you to struggle, before you do anything else go back to your supervisor with each of your options, ask him or her what they think, and ask for their recommendations on how to proceed, long you should wait before starting to be concerned, etc. Trust your supervisor; they'll have been there before.

• You will still be able to analyse the data even if you don't get a high percentage of responses, and then your discussion will mostly be about the difficulties of recruitment, how generalisable the data are given the small sample, what you could do differently if you were to repeat the process, and how the study can be expanded upon in the future.

• Consider doing some preliminary analysis and working out your effect size and the minimum number of participants you would need to get a more meaningful result. You might find out you only need a few more, or you might find out that you have a small effect size and need lots more participants. Then you can decide whether to wait and see, or whether you need to think about a completely different strategy.

• If the research project is for an undergraduate dissertation, the department and your supervisor will not be expecting a perfectly executed piece of research – they are expecting you to have applied the basic principles, achieved some sort of result and discussed it critically.

• Who are you targeting for your project?
- If students: can you hit another module/department/school/university?
- If a marginalised group or if you have exclusion criteria, what may seem to you like an unreasonably low response rate may not in reality be so bad.

• What advertising methods have you used?
- Have you minimised the amount of work they have to do to participate?
- Are you using more than one method of advertising - posters, handouts, presentations, direct approach...? Can you post information about it on message boards (real or online) or in newsletters which are frequently accessed by your target sample group?

• How long is your survey? How intensive is it? - Long or intensive surveys will always have a lower return rate!

• What method of data gathering are you using? Postal questionnaires rarely have a great return rate in large samples; however, the larger the sample, the more difficult and time-consuming it may be for you to carry out face-to-face interviews, for example. You have to consider which is the most viable method in your present circumstances.

• Can you put your study online? You don't need NHS ethical approval to recruit from a clinical population online - university approval is fine, and this is typically much quicker. One ClinPsy member commented “I went from 60/1000 participants over 6 very slow months, to having over 400 participants within a month of putting it on the internet.” Another member emailed every student at their university a link to their survey and got 2500 responses!

However, remember that that brings its own challenges, as you will have very little control over your sample, but there are papers that suggest it isn't a major problem and indeed that you might get an even more representative sample on the internet than via more targetted recruitment methods. It may make participation easier for certain sample groups (and you can play with the formatting to make it seem shorter!) It all depends on your particular study and what's appropriate. SurveyMonkey is one free survey construction tool which allows you to copy and paste responses into SPSS, and costs only US$10/month (about £5-6) to host, but there are others available.

It’s vital to remember that, if you use online surveys in addition to other data gathering tools, you will need to consider the use of internet research as a different method of data collection in your analysis. This is not hard to do: you would do a MANOVA to compare all your variables across internet and postal samples, then you can run your analysis separately in both samples. If you find differences on the MANOVA (e.g. you might find that because internet respondents have access and the ability to use a computer, they may be more educated/wealthy than those recruited via post), then you can control for this in your separate analyses of internet and postal participant samples. The separate analyses may be underpowered, but they will (hopefully) demonstrate that the pattern of results is similar across the two samples. (If it's not, then you've got even more to write about in your discussion!) Assuming there are no major difficulties, you can put the lot into your appendices and forget about them, other than a few brief paragraphs outlining what you've done and justifying combining the two samples.

• Do you have any means of contacting your potential participants with a reminder letter or a phone call? A little bit of gentle persuasion goes a long way. A phone call is best so you can actually speak to the person, but even an email or letter reminder can help you get a few extra. However, ethics committees may view this as manipulative, so unless you have ethical approval for this, don't do it without!

• What incentives, if any, are you using? If you can't offer research credits, money, or other incentives, can you create interest in the topic by presenting to potential participants? Can you trade off with other researchers, i.e. you’ll do their project if they'll do your survey?

• Don’t forget that you don't need to make a breakthrough with your research - you just need to be able to show that you can conduct good research. Having inconclusive or non-significant results isn’t necessarily a disaster; just do the best you can with what you have.

With many thanks to mazzie4, astra, LaLeonessa, baa, Ruthie, vars and bisty for their contributions.

Note: If you have a suggestion about how to improve or add to this wiki please post it here. If you want to discuss this post please post a new thread in the forum.

Content checked by a Team Member on 29/04/2012
Last modified on 29/04/2012
"Music washes away from the soul the dust of everyday life." - Red Auerbach
Post Reply