I keep seeing this, but how often is it true? I always answer these questionnaires honestly. The person who made it just wants some good data for their science. I've never felt a need to lie to them.
> response errors in surveys can occur because respondents misunderstand the questions, cannot retrieve all the relevant information, use inaccurate estimation and judgment strategies, round their answers, or have difficulty mapping them onto one of the response categories.
These go into survey responses to sensitive questions:
- https://www.learnlab.org/research/wiki/images/a/a8/Tourangea...
- http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.504...?
I'm straying off topic, they will try in various ways to bullet proof these surveys. I don't know, I think if a group of people are quizzed, there will be more incentive for people to not give accurate answers. Hm:
What if they come from a family where a guardian is an alcoholic but don't have a point of relation to compare to, and say it's fine at home? They haven't processed it yet.
Outside of surveys, there are blown personality/psychiatric assessments designed to detect over-reporting and malingering: https://scholarsrepository.llu.edu/cgi/viewcontent.cgi?artic...