In online surveys it is common to include measures to help respondents. The problem is that these can harm the integrity of the results by allowing respondents to answer questions without having read them.
Very often in a multi-response question there will be a None of these response option. It makes no sense for someone to select this answer together with any other answers, and therefore None of these will be made “exclusive” by the scriptwriter. What this means is that if someone clicks on one or more answers (other than None of these), and then clicks on None of these, those other answers will automatically be de-selected.
So the result of this is that the data, and associated tables and charts, look nice and tidy. But what this measure has done is allow the respondent to answer the question without reading it. What would be better is that if they select various answers, one of which is None of these, then they should either see a warning message, or be removed from the survey.
Another example would be a single response question such as At which of the following stores do you do most of your grocery shopping? followed by a multi-response questions such as And which other stores do you also do any grocery shopping? Very often the answer given in the first question will automatically be hidden in the second question so that respondents cannot give the same answer twice.
But would it not be better to allow respondents to give the same answer twice? As with the previous example, those who do so could be shown a warning or excluded from the survey. This is more work for the scriptwriter, and it could mean the survey takes a little longer to complete. But the results will be more meaningful and, as a result, will take less time to analyse.