Tuesday, May 3, 2011

Alternative conceptions of "IQ"

Standard IQ tests measure a particular set of intellectual skills.  But this leads to a big question about what exactly is being measured.  The notion of "IQ" implies that the test is measuring an innate, relatively fixed intellectual capacity.

But consider the following analogies:

-Suppose reading skill in English was being tested.  Obviously, we could see that those who score very well in a test of rapid reading and comprehension would most likely be classified as more "intelligent."  Yet, the test would have massive potential sources of bias or inaccuracy.  Many people with excellent language skills would do poorly on an English reading test simply because of their lack of knowledge or experience with English.  Others with excellent minds might not have had much reading education or experience of any sort, therefore they would not have acquired the skill to read or comprehend written language quickly.

-If "athleticism" was being measured, suppose the test involved having subjects shoot free throws on a basketball court.  People who would score very highly on this test would most likely fit reasonably into a "high athleticism" category.  But many very athletic people would score poorly on this test, simply because they had very little experience shooting basketballs. If a  skill that was brand-new to all subjects was being tested one time (e.g. archery target practice), those who performed very well might well be more "athletic" but perhaps they had more experience in a similar type of skill in the past, or they just happen to be more calm with novel activities. 

Acquisition of skills is a process that grows over a period of hundreds or thousands of hours of experience.  If a test is measuring a skill, it may simply be measuring how many hours of experience that person has with a similar activity.    Of course, an intrinsic strength in something is more likely to lead to a person spending more time developing the strength -- talent leads to passionate commitment; it facilitates and makes enjoyable the thousands of hours of work

So, an IQ test may be measuring--to a significant degree--the number of hours of experience the subjects have with similar types of activities (e.g. memory tasks, logic puzzles, arithmetic, etc.).

On a related note,  recent findings have suggested that doing cognitive exercises such as brain-training video games does not really improve intellectual function significantly.  But--the manner in which this conclusion was reached was testing subjects before and after using tests that did not directly relate to the cognitive games which were being played.  Therefore, I believe the findings are spurious. 

Here is my thesis on this issue:

"Intelligence" could be understood not as an absolute quantity of a skill (since this quantity would depend mainly on how many hours of practice or experience the person has had, which arguably should be viewed independently from the concept of "intelligence" as an innate trait).  Rather, "intelligence" could be viewed as the RATE at which brand-new skills improve with a given period of practice.  For example, the amount of improvement after 10 hours of practice of a brand-new activity would be a much better measure of "athleticism" than simply measuring the absolute performance one time.   Even this type of measure would be influenced by a person's past experience:  for example, a person with thousands of hours of experience learning different languages would probably be able to learn a completely new language more quickly with 10 hours of practice, compared to someone with very little past experience of this sort.   The proviso that this be a "brand new" activity is important, because if a person has a very high level of expertise in a particular skill, then their rate of further improvement will be very low (since there is not much further improvement humanly possible).

So, for example, with athletic tests, it would seem a reasonable measure of "innate athleticism" to introduce a completely new sport or fitness activity, and to measure how quickly subjects could improve or master it with limited practice time.

Another complicating angle on this theme is that intellectual experience and focus in ANY area is likely to improve "innate intelligence."  If you have practiced music for 10 000 hours, you will probably have strengthened a variety of other "innate" intellectual capacities.  But this strengthening effect would not be obvious if measured conventionally.

Similarly, I suspect that "brain training" video games could have a significant effect of strengthening various intellectual capacities.  But this positive effect would not be captured well by simply repeating single "before" and "after" skills tests.  A much better measure would be the following:
1) "pre" test:  subjects would have to put in 10 hours of practice learning a game having to do with verbal memory.  The absolute scores, and the rate of improvement over the 10 hours, would be recorded.

2) Subjects would put in 100 hours of practice doing either a battery of memory-focused brain training games, or a "placebo" set of games not focusing on memory.

3) "post" test:  subjects would put in 10 hours of practice learning a completely new game having to do with verbal memory.  Once again, the absolute scores, and the rate of improvement, would be noted.

Here is my hypothesis about the findings of such an experiment:

1) The absolute scores at the beginning of  the "pre" and "post" tests would be quite similar.  This would be like testing the basketball skills of a group of volleyball players before and after volleyball training season.  You would not expect any difference.  This is the conventional type of assessment, which is bound to lead to the conclusion that there is no beneficial training effect.

2) What I would expect to be different is that the subjects who had done 100 hours of memory games would be able to much more rapidly IMPROVE their scores in the "post" test.  In the athletic analogy, a group of people who have spent the summer in volleyball training would be able to much more rapidly learn soccer skills, compared to a group who spent the summer watching TV, even though both groups might have very similar soccer skills at the beginning of a 10-hour training period.  


I believe my hypothesis is supported by observations of individuals acquiring expert skills, such as playing chess.  While the absolute number of hours practicing chess correlates directly with performance, there are some individuals who advance more rapidly with the same number of hours of practice.
(see the following reference: http://www.ncbi.nlm.nih.gov/pubmed/17201516)
I believe it is this "rate of change" that is the most reasonable measure of "innate" cognitive ability.  Even this "innateness" could be more malleable than the term implies, since I suspect that there could be a type of "meta-training" which could improve one's rate of skill learning or acquisition.  This would involve practicing ways to use time more efficiently, and, most importantly, practicing ways to bring devotion, joy, and energetic attention to one's learning experiences.

No comments: