Why would the use of Google docs impact the reliability of a study in comparison to any other online data collection modality? In online survey questionnaires, the reliability of the instrumentation used is easily tested using statistical software. The primary advantages of online data collection lie in convenience, rapidity, and reduction in data entry errors. The disadvantages are obvious in that the respondents are non-random.

As I often tell my students, when you ask one person the first question you have more information that you had before you asked it. The question is not whether experimental designs are better. Sometimes they are and sometimes they are not. It depends entirely on the extent of knowledge in a specific area and the research hypothesis being tested.

I agree with our colleague. The issue is always going to be representativeness/response bias. At the end of the day even the best online study which is well disseminated, with completion potentially rewarded, easy to understand and collecting good social data, will not be representative for reasons such as digital exclusion, differential use of  online by age and ethnicity etc. The first thing I look for in such studies is acknowledgement and analysis of the possible shortcomings and biases. But there are shortcomings in all methods. Ideally I see online as part of a wider data collection effort where you triangulate or develop different data sources. But of course resources do not always permit.

Thank you both for your answers!!! They are very helpful for me.