Recruiting methods
You now have to persuade people to do your questionnaire or test. Some suggested methods are described in the following sections.
Email should certainly be considered, but perhaps only if you are fairly confident that the groups you have email access to are indeed groups that you are happy to test. There may be groups of people that do not use the Internet regularly whom you do not wish to exclude.
Email can be used to invite people to participate in surveys via a link or may directly include or attach the survey/test/questionnaire.
There have been isolated reports of better response rates to Internet surveys compared to mail surveys - these have been with small, specialised samples (Barrios, Villarroya, Borrego, & Olle, 2011). The largest study has been by Kaplowitz, Hadlock, and Levine (2004) who integrated the results of five large surveys (totalling nearly 20,000 students) comparing the response rates to hard-mail and Internet surveys. Response rates to hard-mail requests were superior (approximately 30%) to response rates to electronic surveys (about 20%) (http://fluidsurveys.com/university/ response-rate-statistics-online-surveys-aiming/ offer a similar figure for online email surveys). Kaplowitz et al. suggested that the superior response rates to hard-mail surveys - about 10% better than to electronic surveys -might not justify the extra cost. They estimated Sil per respondent for hard mail, and $2 for electronic mail.
But some groups of people may seldom or never use the Internet and if they should not be excluded you should consider using hard mail, or a mixture of both methods. And of course, for an email survey you need access to suitable lists of email addresses. Saleh and Bista (2017) offer an interesting study of different factors influencing response rates to email surveys, such as being offered a reward, or being sent a reminder.
Social media
Social media can be considered as a way of accessing potential participants. Obviously this is going to give you a biased sample: Facebook or Twitter users only, for example.
For instance, https://www.facebook.com/simple.surveys is a source of instructions for doing simple surveys via Facebook. Or use SurveyMonkey to help you use Facebook for your sample (https://www.surveymonkey. com /mp/facebook/).
Again the sample will be a biased one, but you can sometimes examine some bias by collecting information on gender, age, occupation and other factors.
There is currently no social-scientific work on response rates to social media surveys. The kinds of surveys that are feasible in the social media may differ from the kind of test or questionnaire you are considering — perhaps shorter, simpler and possibly on more “fun” topics. But these may be worth a try if you feel the potential disadvantages are outweighed by the potential advantages.
Survey hosts
There are several organisations that will host your survey, help with its design, data collection and some of the analysis. Most cost money, but some have free options, albeit probably too limited for your purposes. Some are popular among social scientists (e.g. https:// www.surveymonkey.com) while others may be primarily designed for market research purposes but could be useful to you, e.g. http://www.objectplanet.com/ offers several useful tools.
Direct approach: A direct approach is taken when you approach a potential participant saying something along the following lines:
‘Good morning/afternoon. My name is..., and I am (a student at.../ working for...I/we are conducting a study of (topic) and I/we would be very grateful if you spare a few minutes to (answer a few questions/ look at this questionnaire and see if you would be willing to answer the questions).’ People may be approached by door-to-door calling (carry some identification and credentials) or in the street, provided you choose a safe area and time of day. If you are a student, your campus is probably safest.
Mail: Mail shots produce notoriously poor response rates (normally well under 50%) (see earlier).
Telephone: Some researchers call telephoning a stranger out of the blue cold-calling. Telephone surveys may produce disappointing response rates (often less than 10%) and can use a lot of researcher time unproductively. However, a possible way to boost response rates is to first send a letter or email explaining the purpose of your investigation, and explain that you are going to telephone shortly to ask if the person is willing to help. When you phone, you will have saved yourself a lot of time explaining the basics, and if the answer is to be a ‘no’, at least it is usually given quickly.
Use of institutions: It is always tempting to consider this, since you may have access to large numbers of the type of people you are looking for, and it may even be possible to test groups of people together. However, institutions must protect their members, and you will usually have to get through a lot of red tape before you get permission to approach people through institutions. Some (especially hospitals and other health service institutions) will have their own ethical committees. Be prepared to submit a written proposal, explaining and justifying your research and explaining why you need to approach people through the institution. You will also need to submit a
Testing 63 copy of your questionnaire, and explain who you are, and you may have to meet one or more bosses. Be prepared for a long wait - sometimes several months - and a possible rejection. Do not, therefore, rely on one institution. If you have ‘contacts’, this can be useful, as personal acquaintance may reassure those responsible that you can (hopefully) be trusted, are doing something useful and will behave responsibly. If your work will produce something that is useful to the institution, that is an advantage.
Ethical issues
The most important underlying principle is respect - respect for the right to choose to help, respect for privacy, respect for confidentiality. It is also important to express thanks for help, and to explain how the help that has been given will be used and will be useful.
Opting out and opting in: People recruited to do your questionnaire or test should always be invited (not coerced) to participate, and it should always be made clear that they should omit any question they prefer not to answer, and are free to stop at any time without giving a reason (opting out). You should deal with the ethics committee of your institution if you are conducting an independent piece of research; if you are a postgraduate or undergraduate student consult your supervisor about gaining ethics approval. Some ethical committees may require participants to sign a consent form agreeing to participate (opting in). This is more commonly needed in medical than in psychological research. Opting in can produce a lowered response rate but may be an important safeguard when researching sensitive topics, or special groups of people.
An important set of safeguards is that you should make sure that testing conditions are private (ensure that the participant’s answers cannot be overseen or overheard) and undisturbed, and that participants know that their answers are to be treated confidentially and anonymously.
Feedback may be given (see later). Consult the guidelines in Appendix 4. Appendices 2 and 4 offer other ethical considerations and further suggestions about how to put ethical principles into practice. The British Psychological Society’s Code of Ethics is a useful source of advice about this important topic (visit https://bps.org.uk and search for the Code of Human Research Ethics), and other sources are given in Chapter 8.
