The problem with current personality testing..

The problem with all the personality tests online these days is that they type people based upon their ratios,not their dominant function.

Bipolar spectrums for answering between feeling and thinking or intuiting and sensing are handy,yes,but inaccurate if you don't consider that they're merely aids of determining dominant function,
but there are some variables even beebe left out in his theoretical framework of typology.

http://www.greatlakesapt.org/uploads/media/beebe1.PDF

For example, what if somebody was an ambivert?


If somebody was an ambivert, to say that they are especially Xe or Xi would be completely inappropriate, yet their strength may be found in one particular function, be that Nx,Sx,Fx,or Tx. So technically, if somebody who as an ambivert has a definite strength are, then couldn't they switch between types with that general dominant function? I'm an ambivert with a consist dominant Ni/Ne function.. So I have been typed infj,intj,enfp,and entp in the past. It was annoying that my results were not consistent between tests,and I was very confused and frustrated. The only consistent variable was that I always scored extremely high on the intuition part.

For example, I took the personality test twice located here: http://www.mypersonality.info/

The first time I took the test, I got:
Introversion(x):37%
Extraversion(x):63%
Sensing(Ns):5%
Intuition(Ns):95%
Thinking(Tf):47%
Feeling(Tf):53%
Judging(Jp):53%
Perceiving(Jp):47%

I was typed INFJ.

The second time I took the test, I got:
Introversion(x):89%
Extraversion(x):11%
Sensing(Ns):0%
Intuition(Ns):100%
Thinking(Tf):53%
Feeling(Tf):47%
Judging(Jp):53%
Perceiving(Jp):47%

I was typed INTJ.


The reason I was typed INFJ in the first test and INTJ in the second was because of the difference in the Thinking and Feeling percentages. I'm going to guess,also,that those differences were the product of one change in answer selection. If type is supposed to be something that's fairly consistent,how could it be that one difference in answer selection would suddenly change the whole result? I'm probably pretty strange in the sense that I don't have natural preference towards thinking or feeling, OR introversion or extraversion that is consistent. But I do know how test could be revised to provide more consistent results in how people are categorized, I believe.