The decision to recruit largely via Twitter appears to have shaped the responses received, with many indicating Twitter as a key social media space for instance. This may reflect widespread adoption of Twitter in professional contexts but may be more likely to reflect recruitment methods. Recruiting volunteer participants through such a personal route does risk gathering a more limited variety of opinion (Rosenthal and Rosnow 1975). However one of the benefits of such a transparent recruitment process was that participants offered feedback on the questionnaire itself, providing insightful commentary on the relatively high incompletion rate. Their comments highlighted several concerns:
(a) The substantial length and involved style of the questionnaire:
“I tried, but it was too wordy. A simple tick box questionnaire is more my style. Sorry..!”
(b) The wording on the consent form:
“…could I ask about the native English speaker requirement in your survey? Why?”
(c) Compatibility issues with the use of Bristol Online Survey on mobile devices:
“I should never have tried to do your #mscel survey on my phone. As if typing wasn't hard enough, battery died on last screen”
Concern (a) was identified as a potential issue at the point of authoring and refining the questionnaire. It was, however, deemed more valuable to seek richer responses from fewer participants than to receive a higher number of quantitative responses. This approach does, however, further deepen the risk of only obtaining responses from particularly motivated and potentially atypical social media users.
The two remaining concerns were fair criticisms of the questionnaire. In retrospect the phrasing commented on by potential participant (b), on the consent form, should have been phrased as “fluent English speaker”. Concern (c) was the result of not testing the questionnaire on mobile devices where it did not display consistently or usably. Given the high proportion of participants who accessed the questionnaire via referral from Twitter it would have been wise, in retrospect, to test the questionnaire on major mobile browsers to ensure compatibility.
Previous Section: Features and limitations of volunteer recruitment | Next Section: Interviews