Dear fellow researchers,
we recently conducted an online survey seeking to ask physicians about their perceptions and adoption behavior regarding telemedicine in primary and specialized care. We noticed that it seems to be difficult to acquire large sample sizes and achieve acceptable response rates.
Hence, I am asking you for your experiences in achieving high(er) participation rates when conducting online surveys in health care, especially with general and specialized practitioners as the target group. During the ongoing pandemic, time to participate in research seems to be scarce.
Can you provide any best practices?
Thank you and best regards,
Marius
Dear Marius,
the most important thing is the survey to be SHORT and the questions to be SHORT. Surveys that try to answer/solve too many research issues are not attractive.
Yours, Eliana
Thank you Eliana Charalambous! I will put more effort in formulating questions more leaner. Constructing short and precise surveys is always a challenge, especially when the theoretical framework comes with many constructs and hypotheses that want to be tested.
The problem is that we try to yield max information/data to be statistically analyzed in ONE survey. We don't have the time and patience after 10-12 questions or 2 pages....
I agree with the above. Also to
- be clear about the relevance to science, practice or life
- to say the results will be available to participants in some form
- to have a strong, motivating, and short explanation about the research
- also to have a strong short intro to the survey itself saying average completion time
- generally follow good practice in survey design especially cognitive testing with clinicians
- make sure it saves partly completed surveys for completion later
- offer an incentive if one is available (does not have to be monetary, just relevant)
- if possible refer to clinical involvement in the research (approval, involvement in design etc) as doctors are motivated by peer behaviour and involvement
Thank you Andrew William McCulloch for the valuable advice! Some of the points you mentioned have not been incorporated in our previous survey designs. I am sure these will help to increase participant motivation and resulting response rates!
There are three golden rules:
1. Keep it short and simple, so it could fit on a single sheet of paper. We routinely measure and publish wordcount and reading age. Microsoft word has built in word count and readability statisics, such as the Flesch Kincaid Grade (FKG). Reading age is roughly FKG + 5 (see https://bmjopenquality.bmj.com/content/9/1/e000789).
2. Use tested questions. It may be tempting to do it yourself, but so often terms are ambiguous. If so, some people answer the "wrong" question and the whole exercise is a waste of time.
3. Ask people directly to do it. Email does not count as asking. If you can't ask them they won't do it. If you can't get through to the target respondents yourself, collaborate with those who can. If asked nicely, most people will help.
- Badges
- Science topic