[lbo-talk] 'American kids, dumber than dirt'

[Miles Jackson]

Wojtek Sokolowski wrote:
> You know as well as I do that correlation does not equal causation, and in
> order to predict something you need a causal model, not just statistical
> correlations.  The IQ and 'aptitude' testing lacks such causal model in a
> scientific sense of the word, save for a bunch of racist beliefs about
> 'general intelligence' or g.  Since typical academic tasks differ rather
> substantially from those included in 'aptitude' tests (which you seem to
> acknowledged in your reply), the existence of g is a necessary assumption to
> claim that the latter is a good predictor (in a causal sense) - without such
> an assumption the predictive claim is a nonsequitur.  I would be interested
> to hear how you causally link test scores to academic performance without
> assuming some kind of "latent intelligence" they are supposed to measure
>
>
>   
To descend into methodological wonkiness, predictive validity is 
distinct from internal validity.  To effectively predict something does 
not require a causal model.  Thus if "g" is a statistically significant 
predictor of taking cruise vacations, then it has predictive validity 
for that criterion, regardless of the causal relation (or lack of it) 
between "g" and cruise vacations.  Your question "is some kind of latent 
intelligence causally linked to academic performance" is completely 
distinct from the question "do scores on an IQ test effectively predict 
academic outcomes".  In my view, it's important not to conflate these 
two different questions.


> The difference between a spurious correlational swindle and a bone fide
> causal model is that the former gives you a lot of false negatives, whereas
> the latter does not.  A false negative may not matter that much in money
> grubbing schemes - as overidentification of risk is good for the bottom line
> - but it matters quite a bit when it wastes actual human lives by baring
> them access to societal resources.  In that sense, 'aptitude' testing not
> only obscures the true nature of human cognition, but it is akin to fascism
> that gives access to a good life to a select few and bars it for the
> "untermenschen" based on pseudo-scientific beliefs of the 'worthiness' of
> human life.
>   
I don't follow your argument here.  The predictive validity of IQ tests 
is higher than the predictive validity of SSRIs for the treatment of 
depression.  (Proportion of variance accounted for is higher in the 
former case than the latter.)  According to your criterion of science, 
most drug research is not scientific!  There is lots of error variance 
floating around in studies with human beings, so we'll have a lot of 
false positives and false negatives in "real" scientific research.

This is your second fascism analogy in the thread; I still don't quite 
understand.  For the most part, aptitude tests have high predictive 
validity for their criterion, and they are continually revised to 
maintain and improve their predictive validity.   Thus the use of these 
standardized tests with high predictive validity is facilitating 
meritocracy, not undermining it!  --Consider the alternative: we rely 
solely on the subjective assessments of teachers and other authority 
figures to determine access to some scarce resources/opportunities.  

 From many long years of teaching, and can assure you that these 
subjective, anecdotal assessments will produce far more injustice and 
denial of opportunities to capable people than IQ tests ever will.  One 
example: the English profs at my college recently got together to do a 
"grade norming" check on research papers in Eng 102.  A group of tenured 
English profs each graded the same set of research papers, and they 
found that the assessment of a particular paper varied wildly depending 
on the prof (e.g., a paper that received an A from one teacher received 
a D from another!).  Given the fact that the research paper was the 
primary basis for a grade in this composition class, a student's grade 
would dramatically differ depending solely on which teacher they had.  
To put it bluntly, this is what happens when you have people without 
training in scientific measurement trying to assess outcomes.

I share your concern about capable people being denied access to 
opportunities; that is one of the fundamental tragedies of capitalism.  
However, the greatest danger here is not standardized assessment; it is 
nonstandardized, subjective assessment.  Again, if a particular 
standardized test does have predictive validity, it will effectively 
identify capable people and increase their likelihood of getting access 
to valuable resources.  If we rely on subjective assessment, people in 
positions of power can apply arbitrary judgements to deny opportunities 
to the capable and divert resources to the faithful and incompetent 
(that pretty much sums up the Bush Administration, yes?).

Miles