I’ve taken a few different mental assessments and they all follow the same formula for the most part. One question will say “have you ever thought about cutting yourself” and then 14 questions later, “have you ever thought about harming yourself” and then some weird question right after like “do you like school?”
And it’s like 60 questions like this on the sheet it just seems so weird to me. Do people really not notice the pattern? Is it meant to be visible? Is there some psychology behind it?
In: 3
I was diagnosed as schizophrenic a few years ago and every session with my counselor always begins and ends with are you going to hurt yourself or anyone else. Bitch it’s been years if I was I would have done it already
If you’re answering honestly, you’ll have a consistent pattern. If you’re verysmart and answering what you think you should, those questions will trip you up.
The proper way to design such assessments is to start with a large question bank and then have a large number of people with known characteristics take it. You can then apply any number of statistical techniques to identify what questions provide independent information (and what information they provide).
In practice, this sort of process is expensive and time-consuming. Far more often, a prominent researcher just makes up something that sounds good and uses it for their research. Other people copy them, making the assessment more popular and no one much bothers to determine whether the assessment has any underlying validity.
This sort of approach had led to a ‘replication crisis’ in fields such as psychology where the majority of published results appear to have no underlying validity.
So the real answer to your question is that the assessment you’re seeing probably has little behind it except a wild guess – probably incorrect – by some researcher who didn’t have the funding to build an assessment that could measure more accurately than simply asking someone “so what kind of crazy are you?”.
Great responses here, but there is also a meta level of them seeing how the test biases the respondent. Think of three questions in a row
1) Would you call yourself depressed?
2) You’re holding a floofy pink puppy with adorable like snookum eyes, what do you name her?
3) Are you sad today?
You would expect the answer to question 3 to be now biased by the pleasant thoughts evoked by question 2. If you only had question 3 you might be missing people who are in fact depressed, just under-reporting it. By have the same-ish question in multiple phrases and places in a survey you can get a better base-line and hopefully remove interactions like this.
The different wording allows you to catch edge cases and so forth. So say you’re testing for kleptomania, and you have one question which is “do you feel like you need to steal things?”. And then someone thinks “oh, I don’t need to, I just choose to” so you put in “do you sometimes steal things you don’t need?” or whatever. (And in the reverse way, someone could feel a compulsion but not go through with it.)
There are also statistical ways to catch liars with these sorts of multiple-questions, but my grandfather designed a lot of those tests and he says he puts little stock in that.
Latest Answers