Treatment of First Resort

      As one would expect from the heterogeneity of the class
      "depression" having no covering laws, and as is evidenced from
      numerous postings on this NET, some people benefit from cognitive
      therapy and some do not. Some benefit from drug therapy and some
      do not.

      So we must approach the problem of treatment on an individual
      case-to-case basis.  What rationally coherent procedure might
      an individual coherently adopt to weigh the merits and
      drawbacks of a treatment for HIM NOW in his present circumstances?
      (Note the dual indexicals in capital letters).

      I suggest that he construct for himself the following
      Evaluation Table of non-zero values important to him multiplied
      by his subjective probabilities (degree of belief), the product
      of which will yield the expected values of the compared treatments
      for depression.

      The VALUES in the leftmost column represent the desireability
      of a treatment consequence.  Each VALUE is given a WEIGHT of
      1 - 5, 5 meaning a high value and 1 a low value.  The PROBABILITIES
      (Pr .1 - Pr .9) represent the individual's estimate that the
      treatment will realize the desired VALUE.  Multiply the WEIGHT
      times the PROBABILITY and then sum the products to get the method's
      expected value. For example, in the following table I represent
      MY personal assessments NOW comparing three treatment options -
      cognitive therapy mediated by a computer program, (CMT), cognitive
      therapy mediated by a human therapist (HCT), and antidepressant
      medication ({*filter*}).

                            Evaluation Table

  VALUE           WEIGHT     Pr(CMT)       Pr(HCT)       Pr({*filter*})

  Effectiveness     5     5 x .7 = 3.5   5 x .7 = 3.5   5 x .7 = 3.5
  Safety            5     5 x .9 = 4.5   5 x .6 = 3.0   5 x .6 = 3.0
  Economy           3     3 x .9 = 2.7   3 x .2 = 0.6   3 x .8 = 2.4
  Prevention        5     5 x .8 = 4.0   5 x .3 = 1.5   5 x .1 = 0.5
  Availability      3     3 x .9 = 2.7   3 x .5 = 1.5   3 x .7 = 2.1
  Ease              3     3 x .2 = 0.6   3 x .3 = 0.9   3 x .9 = 2.7
  Nonstigma         3     3 x .9 = 2.7   3 x .7 = 2.1   3 x .8 = 2.4
  Learning          5     5 x .8 = 4.0   5 x .8 = 4.0   5 x .2 = 1.0
  Convenience       2     3 x .9 = 2.7   2 x .5 = 1.0   2 x .9 = 1.8
  Records           3     3 x .9 = 2.7   3 x .1 = 0.3   3 x .1 = 0.3
  Catharsis         5     5 x .9 = 4.5   5 x .9 = 4.5   5 x .1 = 0.5
  Human contact     3     3 x .1 = 0.3   3 x .9 = 2.7   3 x .3 = 0.9

  SUMS OF PRODUCTS    (CMT) = 34.9     (HCT) = 24.6    ({*filter*}) = 21.1

      The expected value of computer-mediated cognitive therapy is 34.9,
      that of human-mediated cognitive therapy is 24.6, and that of {*filter*}
      is 21.1.

      Hence I would conclude, based on my personal values and
      probabilities, that (to no surprise to my critics ) computer-
      mediated cognitive therapy is, overall, the treatment of first
      resort for ME NOW.

      Each individual makes up his own table comparing treatments
      and decides which method is of greatest expected value for HIM NOW.
      Again note the important indices in capital letters.  A different
      individual will have different values, weights, and probabilities
      and the same individual may have different entries at a later time.

      (To be continued)

                 Ken Colby