I am trying to capture the reasons why incidents happen to divers using a validated human factors model. I have been doing this research outside of formal academia for a while, so am relatively well known across parts of the industry and have a level of credibility within parts of the population.
However, I recognise that the best way of getting a broad sample across the population would be to use the organisations involved in the industry to promote the survey on my behalf, but they do not appear to actively support my work.
As such, I have had to promote the survey myself through a number of social media outlets (forums, Facebook, LinkedIn, Twitter and hard copy magazines and their webpages) asking for responses through an anonymous survey. I also know that I have had 1st and 2nd level network contacts promote the survey on my behalf because of who I am and the work I have done in the past.
However, there is a concern that I might be biasing my results by using self-promotion and those who know me may also introduce biases by selecting their population to promote to.
Fundamentally the community is sceptical of higher organisations introducing more regulation or legislation for what is a recreational activity, but because of the work I have been doing in the past, I have a level of trust as I do not believe that legislation is the answer, rather more personal responsibility. Crucially, I believe that if I used an independent data collection company there would be fewer responses (and would likely cost more than I can afford as I am self-funding this 5 yr research programme).
With all of the above in mind, I am trying to find evidence to defend my position: whilst the number of responses may be influenced because I am a credible researcher and have a positive reputation within the community, the specific answers as to why incidents have occurred will not be influenced.
Generally, people do not lie in surveys. I would go for the method where you get more responses. I think most responded will be unbiased even though they know you! They will provide their true responses. If you 100% know that they will be answering some questions falsely, I am sure, those would be considered as facts because you already know it. You may not need a survey for that. A focus group would be a better choice to clarify the doubt if you have any. You can have some questions by which you can test the validity of the survey.
Without knowing the context, it is very hard to discuss about a human-factors issue.
Regards,
Shaheen
I agree with Shaheen. MANY researchers have reputations and they still are able to conduct valid research. Make sure the questions are parsed well to avoid leading, bias, double-barreled questions, etc. and I think you have nothing to worry about. Academic surveys are done with human subject mandates and that usually include knowing whom is asking the questions. No big deal. If you believed that the results were unreliable you could (and should anyway) confirm with triangulation, as Shaheen also suggested, with focus groups,
etc.
Talke care.
MLK
I understand your concern. Clearly the sample that would be included in your survey is not likely to be a random sample. This is a limitation of your study and there is no getting away from it. However, it depends how you present your study or what you intend to use it for. If you try to present it as representing a larger population you may need simply to state how self-selection factors may have affected it. You can also make comparisons within your sample that are meaningful. For example, if you know the age or gender of respondents you can legitimately compare age groups or genders and these are less likely to be distorted by self-selection into an online study, Also, if your sample is large, it may be compelling even if it is not a random sample.
Hi Gareth,
What I think is that there is not a perfect recruitment method. Participants will be always influenced by something, such as money (if you pay them), length and type of survey and of course the researcher. If your sample is big enough, these influence should be minimized. Of course, be sure that the participant do not know your hypotheses, because they can answer accordingly.
- Badges
- Science topic
- Similar topics
- Biological Science
- Anatomy
- Tissue