Friday, December 17, 2010

Multiple Intelligence and Life in general

My wife sent me a couple of links on something called “Multiple Intelligences” or MI which is a multi-modal methodology for the classification of intelligence.

Essentially the idea is that using a single classification for intelligence “IQ” does not capture an entire person so the doctor involved subjectively developed various groups of “intelligences” and designed a series of tests to assess an individuals placement within these classifications.

This is actually a pretty good idea, however, like most things in psychology (including IQ) it suffers from subjective assessment and the internal filters of the people involved in the development of the program and the assessment of the intelligences.

Multi-modalities exist in nature. If you take a salt shaker and dump it out on a table you get a pile of salt. If you take a cross section of that salt pile you will get a “curve” that looks a lot like the grading curve your teachers may have used when you were in school. This is called “normal” distribution.

Here is the thing most people do not understand. The entire curve, even at the far edges is “normal”. Most people tend to think of the center section of the curve as “normal” and the edges as “abnormal”. That couldn't be farther from the truth. “abnormal” is something that does not appear on the curve.

Suppose we jerk the salt shaker slightly sideways as we are pouring it out. We now have what looks like two intersecting piles of salt. If we take a cross sectional curve we still have a “normal” distribution, but, we call this distribution “multi-modal” because it has two “humps”.

There are many multi-modal distributions in nature and in manufacturing. One of the biggest mistakes people make in understanding multi-modal distributions is tossing out “flyers”. Flyers are typically some strange result the researcher collected, didn't understand and then pitched. The problem is that those little grains of salt farthest from the center of the pile tell us more about the limits of the distribution than all the grains of salt in the center pile.

I won't get into the math behind that. It is kind of like averaging or being on a curve. Maybe one of the kids in your class at school did so much better that everyone else became angry because it made it harder to get a “C”. Some people “drop” those “flyers” just because they “skew” the curve”. This is how cultures are destroyed, the lowest common denominator. The “stronger” mutation the group tries to destroy to prevent the evolution and development of the group. The kid, or any other flyer, that skews the curve.

The idea of multiple intelligences has an emotional draw to people. It is a way for them to assure themselves that they are not stupid and help them develop a better self image.

The problem is the subjective analysis. Teachers and psychologists (and sometimes engineers) typically grade subjectively based on their experience. This will skew the results of the analysis based on the personal prejudices of the reviewer. The teacher who gave you crap grades because you didn't get along with them. We have all had one of those.

Intelligence tests are typically not very intelligent. They require language comprehension skills that are a learned skill and so people who have strong language comprehension skills do well on tests regardless of their actual intelligence. The same is true of these MI tests.

Multi-modal distributions occur naturally. In this situation it appears that some researchers have subjectively determined multi-modal groups and then created tests which skew respondents into specific categories based on a subjective assessment of what the researcher believes they mean. For example, “I enjoy categorizing things by common traits” and “I easily pick up on patterns” are essentially the same thing and yet these questions are in different intelligence groups. In the logic section I found the phrase, “Structure is a good thing”. Understanding different structures can be difficult. For example understanding the structure of interpersonal relationships in the management of a large organization can appear chaotic to some. Some might not see the structure at all. Rembrandt vs Picasso. To me, this structure phrase means nothing. I would need to understand the structure and it's intended purpose to subjectively determine if the structure is a “good” thing, or likely to produce the intended results. The Tacoma Narrows Bridge is an excellent example of a structure many thought was good, well, before they built it they thought it was good. After it collapsed not so much.

So how do I answer “Structure is a good thing”? I would select a “middle” answer and the results would skew me away from being a logical person.

However!

The world wide web does offer a wonderful opportunity at really understanding the multi-modal development of personality (intelligence is a stupid term, and I have an IQ of 157 so you should believe me).

Different websites will have more visitors than others. By just capturing the visits to different websites and subjectively analyzing the “content” we (people in general) can develop a better understanding of the real, not the subjective, categorization of these multi-modal groupings.

Multiple Intelligences is a good idea, in the current development of subjective analysis it has it will probably die away for some other “pretty butterfly” that captures the imagination of researchers unless real objective analytical development is accomplished.

Here is a paper on MI: http://pzweb.harvard.edu/PIs/HG_MI_after_20_years.pdf

Here is an overview: http://tip.psychology.org/gardner.html

Here is an interesting paper on MI and culture: