I want to conduct a reaearch without that typcal students sample and a broader variety of professions included. Does anyone have experience regarding response rate and representativity?
To my understanding, anything advertised within your network (any social network) would be regarded as convenient sampling. However, if you have LinkedIn premium membership, you can access to vast number of professional, whom you can short list using search function. I have used LinkedIn in first case but not in the latter case.
I believe in latter case, you can have total number of professional short listed based upon their job description, so that would be sample frame and the number of responses that are returned can be used to calculate response rate.
Another option could be making use of Google URL Shortner or Bit.ly to track number of clicks form LinkedIn and other social network and number of actual response. Then you can say use the No. of actual responses/No. of clicks to show how many people accessed the questionnaire and how many responded. I have done the same in my recent data collection to track the respondents. You can more information on respondents using both of above mentioned services.
Thank you for your answer!
What about posting in Groups on LinkedIn treating the interests of the desired target group?Is it still biased?
Theoretically it does not look biased, however, in reality the approach cannot amount to be a random sampling. Not all the group members read posts within the group and thus are not covered. For example, I rarely check the message digest from a group on daily basis, rather delete them without ready. Personally, I would still consider it a convenient sampling.
Another approach that is used these days by researchers in digital divide is to use representative sampling, where they collect a sample of same characteristics as that of the population and then try to generalize the results for the population.
In addition to that you may make use of some survey sights, such as Qualtrics or Amazon Mechanical Turk (https://www.mturk.com/mturk/welcome) for data collection to collect data from your target population. Of course it will cost you some money (depending upon your responses).
You can also use snowballing technique, where you identify several primary respondents, ask them to fill in the questionnaire and then ask them to given you references. In this way you can reach number of respondents. However, response rate can be an issue in this technique too.
In short, if you are doing a data collection as a student, with limited resources, adopting random sampling is a bit difficult. So I would suggest to use convenient sampling or snowballing technique.
- Badges
- Science topic
- Similar topics
- Methodology
- Research Methodology