During the pandemic I started some web survey. It's convenient and very economical. However, it creates very important sampling biases. Would you like to know what are the best practices for getting the best out of these search devices? Are there articles or reference books that can help improve the design of surveys with this type of device (web surveys)? Furthermore, to what extent do sample biases invalidate the analysis of relationships (associations) between variables (through regressions, for example)?
Thanks in advance
There would be selection bias , as you already guessed, just due to the fact that not everyone has internet. Also, what is the incentive for people to respond to the survey? this may lead to more bias.
Best practice is to use something else. You have no idea of the sampling distribution. for example this is the electronic version of a convenience sample which should never be used for anything that is important. David Booth
Vikas & David
Thank you very much for the comments, even though it throws some "cold water" on this type of device.
Best
- Badges
- Science topic
- Similar topics
- Geoscience
- Asia
- India