So, this test has dichotomy and cognitive functions on a single test, which makes it one of my two favorites cognitive function test. I would say that a test combining cognitive functions with dichotomy is more "powerful". This is so nice, because combining the results you all post it, it is possible to me to do a very good evaluation of the test, and I can easily detect it flaws. Its a shame that it seems that the author of this test doesnt seem to become aware of this, a test with dichotomy and cognitive functions can be quite "tuned" by slowly fixing the issues. So, Im not going to post my results but Im going to point the interesting flaws.
With dichotomy on its own, it is important that no dichotomy correlates significantly with each other, for example Extraversion should not correlate significantly with Sensing for example. This is this tests first flaw: Intuition is correlating significantly with Perceveing, or, to be more clear, N is correlating with P. That flaw doesnt affect any NP or SJ person, but it does affect SPs and NJs. The why is a little complicated, but observing the pattern of results it seems that S causes an increase in J and N causes an increase in P. So, a SP person, like [MENTION=29903]Mayflower[/MENTION], will get borderline on J/P or even J despise being P. And, on the other side, a NJ person, like [MENTION=34350]Norexan[/MENTION], will get borderline J/P or even P despise being J.
Cognitive functions, on its own, have very discussable musts, and for that reason I dont do a evaluation. However, in cognitive functions crossed with dichotomy, there is one rule that must be followed:
- Fi and Fe must correlate significantly with feeling, Ni and Ne must correlate significantly with iNtuition, Te and Ti must correlate significantly with Thinking and and Se and Si must correlate significantly with Sensing. What "significantly" means is subjective, however general standards for Google Scholar (scientific articles) says its 0.6, however I do use myself 0.4 for cases like this.
It doesnt matter what cognitive function stack you follow, or if you dont follow any cognitive function stack at all. This rule and what I said for dichotomy has to be true in evaluating different types (if you cross data with only the same types, these rules dont apply, for example this rule doesnt apply if we are evaluating only INFJs).
Inside this cognitive function to dichotomy rule, there is a three other flaws on this test. Se and Ni are not correlating with N/S at all, and that could be related to the N correlating with P issue. Whatever who is significantly affected by the N-P unproper correlation, is affected by this issue either. The third issue is that Fi is not correlating with feeling. So, any F that relys on Fi significantly more than on Fe will get borderline T/S or T instead of F. FP=Fi is not exact and is quite a raw conversion, but it is expected that most FPs should be affected by this issue. This is why what [MENTION=31348]Peter Deadpan[/MENTION] pointed out happens and it happens with [MENTION=7991]chickpea[/MENTION] and [MENTION=33955]Ashtart[/MENTION] and with me. I usually oscilate between T/F all the time, because Im "borderline" on that, but in this test I get T for MBTI and F for the other two methods all the time. Speaking of which, I can explain them (the methods). These 4 different MBTI results gives some doubt, right? Such as this doubt:
Norrsken said:
I don't understand my results, honestly.
I wont explain exactly your results, but rather Ill explain what these 3 methods are. Basically, instead of giving a single result, this test gives 3 different points of view.
The simplest is "myers-briggs function type". Its the equivalent of dichotomy, basically it measures E/I, N/S, F/T, P/J regardless of cognitive functions, like it is measured on 16personalitities. This method does not use cognitive functions. Sakinorva has some MBTI interesting reads and at their opinion (at least though my own interpretation) the most reliable method is through dichotomy and not cognitive functions, and for that reason the "most likely myers-briggs type" is always equal to the "myers-briggs function type" result.
"axis-based function type" is an orthodox method which I never saw (and perhaps most typology community neither) before, but Sakinorva claims that it is "commonly used by magicians and typologists today". By magicians, it is quite clear that they dont really approve much that method (I dont either, but it is interesting). I think it is better explained by them:
"The axis-based algorithm will assume that there are no inferior functions in your stack, and that functions on opposite ends create axes that you would either prefer or not prefer, so in other words, your scores for Ne/Si are compared to Ni/Se, and the same thing goes for Se/Ni and Ni/Se. The algorithm then tries to figure out which one of those four "valued" functions you prefer should be dominant, and voila! You get your type."
So, basically they will do a Ne vs Si, Se vs Ni, Te vs Fi and Fe vs Ti duels. There will be 4 winners in this duel, and using the winners they will determinate which type is yours (they were unclear what cognitive stack they use for that evaluation but I guess its the Grant one).
The third one is one of my favorites. It claims to be "The Grant/Brownsword model" but the author mistakelys creates a new typing method on its own (although seems aware of this). They have, for every type, a formula like this:
Dominant+Auxiliary+Tertiary-Inferior = Type score
The dominant, auxiliary, tertiary and inferior are all from Grant/Brownsword model, which is, for example, Ne-Fi-Te-Si for ENFP. They put weights on this, meaning that the dominant counts more than auxiliar and that the tertiary count less tha dominant and aux. However, this is not Grant/Brownsword method at all. There is one important and missing "detail" here. Summing and subctracting functions doesnt imply into a fixed cognitive function stack. Its entirely different thing that appers to be equal. If we sum E+N+F+P, we are "piling" up all the 4 dichotomies althogheter, forming 1 dichotomy on ENFP orientation. When we sum Ne, Fi and Te, and minus Si, we do the same thing, we are "pilling" up 4 cognitive functions. The Grant Stack expects that Ne-Fi-Te-Si will show for every ENFP (and as Sakinorva observes in their own website that doesnt usually happen), where Ne should be the highest function for ENFP (highest value), Fi in second highest, Te in third and Si, depending on which community/point of view you follow, is either the 4th or the 8th function (the lowest). So, for Grant, Ne must be at first position for ENTP and ENFP, while for Sakinorva Grant-Brownsword algorithm that doesnt really need to happen. Its more flexible. But it creates a very interesting method, I prefer to call the sum method, since it can be used unweighted and it is possible to use other stacks as reference. I already managed to put cognitive function stacks to compete with each other, but not for this test.
So, my final saying is that the most reliable method of these 3, for this test, its case-dependent. Because of the N-P correlation, the dichotomy one is not really the most reliable all the cases.