Randroth

I would be interested in the data surrounding bipolar (and its different types), schizophrenia, OCD, and depression.

^^Done. Thanks for the suggestions.

I dislike taking large tests for this type of stuff. Also, more choices can be beneficial in my opinion, only it'll take me longer to do the test/quiz. Truely, as long as the tests are fairly accurate then I don't really mind. Just giving feedback…

Originally Posted by Armageddon_Wasteland

I dislike taking large tests for this type of stuff. Also, more choices can be beneficial in my opinion, only it'll take me longer to do the test/quiz. Truely, as long as the tests are fairly accurate then I don't really mind. Just giving feedback…

Thanks for the feedback! This won't be one large questionnaire, it will be a series of polls I'll be posting here on PerC. And for each question, I'm planning on posting as many relevant choices as possible without cluttering the poll.

How are you planning to analyse the data? SPSS?

Originally Posted by friction

How are you planning to analyse the data? SPSS?

I had been planning to simply calculate the percentage of each answer with each personality type. I've never used SPSS though, is it something you would recommend for more in-depth analysis?

This sounds like some interesting work. I would request, though, that you mention in your final report the limitations of doing this type of research on an internet forum. Obviously, you're going to have issues with generalizing the results because of the lack of randomization and the less than ideal sample. It will still be interesting to see, though.

great! i have been lamenting such a lack of statistical data when it comes to Jungian types for some while now. will be very interested to peruse the results. great work! :)

Just out of curiousity, why do you think there is a correlation between MBTI and psychological disorders?

This is awesome! I really have high hopes for this study.

Originally Posted by justcait

Just out of curiousity, why do you think there is a correlation between MBTI and psychological disorders?

Jung hypothesized that a number of psychological disorders were related to psychological traits he described in Psychological Types; for example, he conceived of hysteria and neurosis as resulting from highly intense relation to external objects (i.e. extreme extraversion) and schizophrenia and paranoia as resulting from total or almost-total detachment from external objects (i.e. extreme introversion). As for the rest, I have no idea whether or not there's a correlation between them and MBTI, which is exactly why I'd like to study them in the first place.

There is an inherant error in your plan - the mental disorders screw up cognitive performance and thus influence the andwers selected on a test - Are you the one who has claimed that certain personality types are more prone to psychosis? All data demonstrates what you are attempting is erronous - the mental disorder skews the MBTI results. You are confusing results with causality.

Originally Posted by OldManRivers

There is an inherant error in your plan - the mental disorders screw up cognitive performance and thus influence the andwers selected on a test - Are you the one who has claimed that certain personality types are more prone to psychosis? All data demonstrates what you are attempting is erronous - the mental disorder skews the MBTI results. You are confusing results with causality.

1) "MBTI test results" =/= "MBTI type". Somebody taking a test to try and find their personality is probably going to get bad results, psych disorder or no. So saying that having a mental disorder will prevent people from answering test questions correctly is true, but irrelevant.

2) As far as I can recall, I've never claimed that certain personality types are more prone to psychosis. And even if I had, the entire point of this exercise is to FIND correlations between personality types and other factors. If there really are no correlations, then so be it. The exercise has still done its job.

3) Not once have I claimed correlation=causation in this entire project. No further explanation needed here.

4) MBTI is self-reported, and that's the whole point of the system. At this point in time, we have no way of empirically confirming somebody's personality type (although we're getting there), because it's just an inventory for certain psychological tendencies that are grouped together--and that's okay. Even if a mental disorder does actually skew what we would call a person's "MBTI type", it would be because the person in question has a different perception of their own psychological tendencies than they would otherwise, and that's no different than a non-mentally ill person finding their MBTI type by their perception of their psychological tendencies.

Originally Posted by Randroth

I had been planning to simply calculate the percentage of each answer with each personality type. I've never used SPSS though, is it something you would recommend for more in-depth analysis?

I would recommend it if you wanted to know if the results were statistically significant.

@Randroth -

I'm sorry, but I really don't think your poll results will hold any weight when it comes down to it, and I think you need to make this clear if you want to "publish" this data.

First of all, your results will be meaningless if you don't normalize the data. Professional polls have information at their disposal about the people answering the poll such as age, gender, socioeconomic status, nationality, etc. They then use this information to normalize the data in order to gain meaningful results. One of the biggest problems is not having the number of members of each type on this forum. For example, say there are 15 ISTJs that answer the poll as Christian. And then say there are 50 INFPs that answer the poll as INFP. At first it may appear that Christianity is more popular among INFPs, but in fact you are surveying thousands more INFPs than ISTJs. I suppose you could pretend your sample size was the amount of people who answered the poll, i.e. Out of 100 INFPs who answered, 50 are Christian, which makes a 50% Christian ratio, while 15/15 ISTJs were Christian, making a 100% Christian ratio.

However, you, and everyone who reads your study, has to remember that the results will still be skewed because you have a biased sample (most are frequent internet users who are interested in discovering themselves through personality typology) which might not correlate to real life at all. Plus, your true sample sizes are not even in the slightest. By giving out a poll to an entire forum, you are sampling thousands of INFPs, while only a couple hundred of ESFJs. Even normalizing for this probably can't cover the wide gap. Most surveyers carefully choose their samples to make it as unbiased as possible. My guess is that there will be so few extroverted data points that it will be meaningless. How many data points will you wait for to make the data meaningful? That's something you need to decide. I'm not a specialist in polls and bias, but I have taken a couple of statistics classes that have taught me a lot. Have you? And what I learned in those classes is that it's a lot harder to conduct accurate, unbiased, meaningful polls than it looks. If it were so simple, don't you think it would have already been done?

I just don't want everyone getting their hopes up about how profound these results are unless both the polls are conducted and the data is analyzed properly.

@MBTI Enthusiast

No need to apologize for speaking your mind. I'm in a bit of a rush so I can't address everything very thoroughly, but here are my thoughts on a couple of points you brought up.

1) Regarding the sample sizes, as long as there are enough data points from every type I'm not hugely concerned about polling way more of certain types than others, because it's the percent of each type that gives certain answers that I'm concerned about. That said, while I anticipated getting far fewer responses from certain types, it's blown my mind how few responses I've gotten from some of them. I haven't done an exact tally, but I think from the three total polls I've done (not including educational level) I've probably gotten no more than 50 data points from all of the ESxx types for all of the polls combined. Right now I'm holding out on gaining more votes as time goes on, but it's not looking promising and more than anything else it will influence my decision to continue or not, which I'll probably be making in the next few days.

2) I'm not planning on publishing anything in the sense of getting it into an academic journal; this project is primarily by and for PerC, so if I continue with this the final results would be put together in a google doc or something of that nature and posted on PerC somewhere.

I doubt this will lead anyone to understand their type better. Anyway because you said you only had 50 data points from ESxx types, I tried to add a few more :p