matt wrote:
I think that the test could be improved by adding a text box at the end of each line, so that people could elaborate on why they answered the way they did. This would allow you to see which questions people were having trouble answering.
I once had a comment field (maybe should add this again to the experimental version?). However, to have one comment field per line would make it impossible to track. I cannot go through 1,000s of answers possibly with some comments for different items. It has to be done with a general comment field in order to be practical. Possibly I could make a concatenation of question # with comment and then save it as a single formatted comment. That might be possible.
Besides, number of "?" is a good indication of the quality of a question. If it attracts many "?" answers it's typically bad in some way. Some questions do this even if they are not bad (like sex or relationship questions when the test-taker is too young or have no experience), but I typically allow higher percentage of "?" answers for such questions.
matt wrote:
Also, certain questions are asked as if the only responses are binary, like "Do you prefer to have friends of the opposite gender?" If the answer is yes, then someone would know that they answer yes, but how would a person answer if they preferred not to have friends at all? How would they answer if they preferred to have friends of their same gender? Or they could have interpreted the question as "Do you prefer to have [any] friends of the opposite gender?"
If you have no friends, you should use "?". If you never had a relationship, you should also use "?" on all the relationship related questions.