internet-based.org/questionnaires-as-a-poll-instrument/
Web-based research and questionnaires are essential epidemiologic tools that provide crucial information about public health and disease. They are a common method for collecting data. They are generally more affordable and efficient than face-toface interviews, mailed questionnaires, or automated phone menu systems. However questionnaires, surveys and Web experiments are not without limitations that need to be addressed in order to ensure reliable and valid results.
A questionnaire can be affected by response bias. This is the tendency of respondents to answer questions based on their own opinions and not on research goals. Additionally, questionnaire design can influence responses in a variety of ways. For example the wording of the question can affect whether respondents comprehend and interpret the question in the same way (reliable) and also, determine what you’re interested in (valid), or can answer with accuracy (credible).
Respondents can also experience survey fatigue or a lack of engagement with the questions and reduces the chance of them providing honest answers. Additionally, a lack of incentives or compensation could discourage participants from taking the time to complete a questionnaire.
Online questionnaires also pose challenges for some experimental designs for example, positioning or reaction time studies. The varying settings of browsers as well as screen sizes and operating systems makes it difficult to control and measure the same variables across different participants.
Finally, Web-based surveys are only available to those who are keyboard and Internet literate, which currently isn’t a large portion of the population. It’s also difficult for Web researchers debriefing participants after the experiment window has ended.