Web-based surveys and questionnaires are a vital epidemiologic technique that provides crucial information about the state of health and disease in the public. They are a standard method of collecting data that is typically less expensive and time-consuming than face-to-face interviews, mail-in questionnaires or automated menus for telephone systems. However questionnaires and Web experiments are not without limitations that should be addressed to ensure that they are reliable and valid results.
A questionnaire may be affected by response bias. This is the tendency of respondents to answer questions based upon their opinions and not on research goals. Moreover, questionnaire design can influence responses in a variety of ways: for example the language used in the questions can affect whether respondents are able to comprehend and interpret the question in the same way (reliable) and also, determine what you’re interested in (valid) or are able to answer accurately (credible).
Respondents can also experience survey fatigue or lack of engagement with the questions asked, which reduces the likelihood of them providing honest responses. A lack of incentives or compensation can also deter respondents from filling out an application.
Online questionnaires can also pose a challenge for some experiments, like reaction-time or positioning studies. It is difficult to control and measure variables across participants because of the differences in settings for browsers, operating systems, and screen sizes.
Finally, web-based surveys may only be accessed by those who are keyboard and Internet literate. This excludes a large part of the population. In addition, it’s often difficult for Web researchers to explain the results of their research after an experiment’s window closes.