GRGR(6) Discussion Opener

[David Casseres]

>19) Anyone have anything to say about the MMPI and how it gives infor-
>    mation different from a Rorschach test?  Structure vs. lack thereof.

I dunno about structure vs. lack of, but the MMPI is distinguished by a 
statistical objectivity that is lacking in the Rohrschach test.  Or is it?

In the Rohrschach test the subject makes a verbal response to the 
ink-blot stimulus (well, all right, that's unstructured), and the analyst 
personally decides what it means.  The stimulus is considered to be 
devoid of content, and the idea is that the subject reveals his own raw 
content and the analyst analyzes.  So the hierarchy is

  arbitrary blot --> automatic response --> intelligent interpretation

and of course the weakness is that we can't tell how subjective, how 
irrational, how prejudiced the analysis might be.

So then we get the MMPI, where the stimulus is a long list of 
first-person statements ("I love to walk barefoot through soft gooey 
mud"), the subject responds by grading her response on a scale (strongly 
agree -- agree -- don't know -- disagree -- strongly disagree), and 
(here's the cool part) there isn't any interpretation of the responses, 
only of the statistical aggregate in comparison to other statistical 
aggregates.

Again the stimulus is considered arbitrary, devoid of meaning.  But now 
the response is severely restricted.  The subject reveals herself only by 
checking boxes on the test form, not by blurting out anything like "It's 
not my feet, it's my hands... the mud is flesh... it's my own flesh...."  
And there's no analyst to smack his lips over that.  Instead, the pattern 
of responses is compared automatically to patterns that have been 
compiled automatically from thousands of other subjects, and from within 
those thousands of patterns, clusters are seen to *emerge*, statistically 
measurable resemblances among various collections of subjects.  And so 
our subject can be associated with a collection of other subjects, and we 
can ask "So what is it already about this whole collection?  Why, they 
are all passive-aggressive depressives with paranoid fantasies."  So our 
test subject is identifiable as likely to be one of those herself.

All much more scientific than the Rohrschach, structured, quantitative, 
objective, all that good stuff.  There are a couple flies in the 
ointment, natch.  For one, how do we know all those tankers and feebs 
whom the subject so resembles in her pattern of answers actually *are* 
passive-aggressive depressives with paranoid fantasies?  Well, they were 
*diagnosed!*  Oh.  By who?  Well, um, analysts.  But the subjective 
biases of all those analysts are kept under control by the law of large 
numbers.

We hope.  Trouble is, it's only their personal, individual subjectivities 
that are trampled flat by the weight of statistics.  If they're all 
members of a society that's as crazy as a shithouse rat, then the 
statistics may reify and institutionalize their biases instead of 
"averaging them out."  A-and of course, the more objective and rational a 
testing technique is seen to be, the more it will be adopted by 
beady-eyed apparatchiks whiose principal interest lies in manipulation, 
and the better it will serve Their purposes.

So here's the paranoid's summation: the unstructured Rohrschach is useful 
for the manipulation of individuals, one at a time, by trained analysts, 
whereas the highly-structured MMPI is useful for the wholesale 
manipulation of statistically significant preterite populations by an 
Elect who require only the most superficial training.

[PS: I've probably misremembered some details of the MMPI; if so I hope 
someone will set me straight.]


Cheers,
David