You know I write things and people read what is in their heads....
I'm not a sicentist I'm a market researcher... nor am I a psycologist although I buy them in....
I have at no point claimed MBTI is science, it's just a segmentation - a bit posher than some but still its a classification system....which can be evaluated.. but it's expensive to do
It can be evaluated as much as those self-report tests will allow, and again, that's really not saying much. The fact that you can evaluate it does not mean that you can prove its accuracy. It's almost circular to say that those questionnaires prove anything.
ALL self completion questionnaire are liable to biase
All research is flawed... even medical research...
When you are dealing with quant, it's about limiting error and trying to reduce it down as far as possible.
MBTI isn't being used to save babies... so it only needs to be good enough for people to use it with some common sense to help them understand peoples behaviour... it's sutiably stable enough for doing that.
I wouldn't offer up conclusions on peoples lives off the back of it.... but I would say largely useful
I never said that MBTI can't be a useful tool in understanding behavior. I'm trying to say here that its accuracy cannot be proved because it's so non-falsifiable and there's no empirical way to measure its accuracy.
You can do as many questionnaires as you want, measure them for consistency by repeating these questionnaires, and then use the results as predictive/descriptive tools, but you can never prove that the results of such questionnaires are "right."
... and she's being pretty grounded and rational in her explanations, at least about this.
She's only explained how to test consistency in results, but has said nothing to show how the test actually relates to real life behavioral tendencies.
As for the test itself, it's two-choice forced selection, which introduces numerous problems on its own. Not to mention the rampant N-bias.
For instance one question said (paraphrased):
"When I'm reading, I like to:
A) Take the author's words at face value, or
B) Read between the lines and try to figure out the real meaning"
The implication is that the S answer (A) doesn't actually find the real meaning and is therefore somehow inferior to the N answer.
This happens all over the test and leads tons of people mistype themselves as Ns. You don't have to look far on the forum to find Ss who think they're Ns because numerous test questions are worded in a way that's biased toward the N answer.
Now, we could use this testing research to adjust the questions until 50% of people test S and 50% test N, but that doesn't work because we don't know how these statistics correlate to the neurochemistry involved in cognition and so we have no idea if 50% of people should test as Ns in the first place.
Nobody has explained the problem of relating these test questions to real cognitive processes.
Originally Posted by Mitzy
men are always hating!
regardless of what type i am, im allowed to have own interests even though theyre not the same as my male "club members"
Well, I guess you're outta the club too then.
If you could be anything you want, I bet you'd be disappointed--am I right?
It can be evaluated for consistency, not accuracy in its descriptions.[/QUOTE]
thats not science....
You want to actuarely prove that all ENTP's tested are bonefide ENTPS... my test wont do that nor does it have to be... it measures the overall segmentations reliability...and gives you a score of accurace over time... that is evidencing it's fitness for purpose....
As I said before ALL segmentation have errors, the fact it's self completion is irespective.... the test I suggested will actually identify if there are specific scales/questions that cause the grey areas (this is how you would IMPROVE the MBTI tool).
Providing that 80%/90% of all people would classify themslves as a specific type - is evidence of fitness for purpose (and would be very good....) .