Lets Talk Distributions

[INTJMom]

Thanks.

 

[edcoaching]

This sort of distinction is among the few times, I have felt the familiar feeling of "science" happening in personality theory.

You mean that last paper trying to distinguish between trait and archetype? What Mitchell aimed for in his paper?

There's also the real life side. For example I'm filming students doing math tasks to try to verify any definite differences in approaches taken by ES, IS, EN, IN students, the learning styles I use. Best-fit type, of course.

The differences among the groups are striking. Here's a summary of what we found...

Introversion and Sensing (IS)
• Used squares paper and markers; none used tiles unless the facilitator suggested it
• None used numbers to find common denominators

Introversion and Intuition (IN)
• Only students who drew shapes other than rectangles or used isometric graph paper
• One student built shapes with markers rather than the tiles
• Worked quietly for up to nine minutes on a task
• All used numbers to find common denominators

Extraversion and Sensing (ES)
• Altered the materials to make sense of problems (only ones who shaded tiles, divided graph squares in half, etc., to fit in thirds and sixths)
• Used trial and error without asking for help in between experiments
• None used numbers to look for common denominators
• Used square graph paper and tiles

Extraversion and Intuition (EN)
• Careless mistakes; used colors that didn’t match problem or counted tiles and squares incorrectly
• Unaware of the denominator they were illustrating, i.e., talking about 12ths while illustrating 10ths.
• So confident in their answer that they didn’t see mistakes even while explaining their solution
• Long verbal explanations

This is just the tip of the iceberg--we're repeating the whole thing with more controls this year. Already, though, some of teh "math experts" I'm working with are realizing that the curricula they're advocating took quickly take away concrete representations, etc. It isn't trait but an actual split in what kids need to learn...

 

[ygolo]

You mean that last paper trying to distinguish between trait and archetype? What Mitchell aimed for in his paper?

Yeah, that one. I had become rather skeptical about the whole notion of psychological type (even though it matched experience rather well, most of the popular web-sites were saying that Jung and Meyers-Briggs theories were being replaced more by the Five Factor Model in personality research), and the trait theory did make much sense either (plus it was nearly impossible to find any coherent description--making it look like an unfalsifyable, shape-shifting theory that could not be pinned down to be tested). I was strarting to think the whole notion of "personality" is an artifice, and that we are all just social meme machines.

The fact that people have actually worked out the implications of two sets of theories and are doing measurements that will differentiate between them is very encouraging.

There's also the real life side. For example I'm filming students doing math tasks to try to verify any definite differences in approaches taken by ES, IS, EN, IN students, the learning styles I use. Best-fit type, of course.

[...]

This is just the tip of the iceberg--we're repeating the whole thing with more controls this year. Already, though, some of teh "math experts" I'm working with are realizing that the curricula they're advocating took quickly take away concrete representations, etc. It isn't trait but an actual split in what kids need to learn...

Interesting. So the idea is to use the results to tailor teaching methods to students and to teach them to be as flexible in the working styles as possible, despite their preferences?

 

[ygolo]

75% validity... and more thoughts on pop. facts about MBTI

I just realized that if the confidence level on each dichotomy was even as high as 93%, the likelyhood of being correct on all 4 would still be below 75% (assuming no correlation on correctness on dichotomies).

(0.93)^4=about 0.748

Actually, knowing that it is meant as a sorter means I conceptually understand the number given being a confidence level fairly well now.

So if I take my scores on a sorting instrument and multiply the confidence scores (taken as fractions), I'll get a confidence of my type based on the instrument (assuming no correlation on correctness). Of course, even a high confidence level could still be wrong.

Just curious how close my deductions are on this.

 

[edcoaching]

Interesting. So the idea is to use the results to tailor teaching methods to students and to teach them to be as flexible in the working styles as possible, despite their preferences?

Exactly. Students need to learn in all 4 styles. A very crude example:
IS: memorizing math facts
ES: measurement, applying mathematics to practical situations
IN: synthesizing math learning to solve new problems--logic, etc.
EN: math discussions that further understanding.

Students need all 4, but when they don't "get" a big concept (such as...1/16 is smaller than 1/6) they generally need some instruction in their own style to get the light bulb to go off. But...for the students, knowing there are different styles and this activity matches a different style is far different than thinking you're stupid and can't learn...

 

[edcoaching]

I just realized that if the confidence level on each dichotomy was even as high as 93%, the likelyhood of being correct on all 4 would still be below 75% (assuming no correlation on correctness on dichotomies).

(0.93)^4=about 0.748

Actually, knowing that it is meant as a sorter means I conceptually understand the number given being a confidence level fairly well now.

So if I take my scores on a sorting instrument and multiply the confidence scores (taken as fractions), I'll get a confidence of my type based on the instrument (assuming no correlation on correctness). Of course, even a high confidence level could still be wrong.

Just curious how close my deductions are on this.

Yeah, I think so? I usually tell groups that about 75% will agree with all four letters. And when we're done, when I ask for a show of hands it's really close to that. Unless...

• It's a multicultural group. My experience is that more of their scores will be close to the midpoint because they're adapting behaviors all the time. So odds are they'll be less clear.
• If it's a group of educators. They're so afraid of negative labeling that they can't make up their minds what they are. I like having them for 2-day workshops because eventually a light bulb goes off...