Friday, June 02, 2017

Social Programs & Science

Almost every week brings news of some new program as to various "Social Problems" (eg Homelessness, Domestic/Street Violence, Under-Achieving Schools/Students).

What is lacking for most (Almost all? All?) of those proposals/programs is some, built-in, plan to scientifically determine if each such program is effective and a wise use of both public and private monies. (Please remember that private contributions are often used to reduce the tax-liability of donors which reduces government's funds and places increased burdens on our fellow citizens.)

There are two basic models for any such scientific evaluations.

One is the "Control Group Vs. Experimental Group(s)" where random selection of subjects (eg
From the pool of all persons alleged to need a proposed program) is made and like subjects are assigned to the studied program. After the unmodified programs are in-place for some reasonable time, a statistical analysis is done to see if there is any, positive/negative differences can be found to exactly determine if that program had: A positive effect; Or, a negative effect; Or, no effect.

The other might be termed "Multi-Factor Analysis". In this mode, all in a larger number of subjects are involved in the considered program; But, all such persons are classified by very many characteristics (eg Age, gender, race, history of  AODA, age of mothers at first birth, academic testing, results of MMPI testing). Thereafter, a more convoluted study can be done to determine any outcomes as can, mathematically, establish those results noted above and may provide more and useful information as to which personal characteristic result in those outcomes.

Unless such evaluations are built-into such programs (And exactly followed), the only sure results of such programs are: The wages/salaries paid those who are employed by them; A waste of public/private funds; And, to make some people "feel good"---Without due cause.

No comments: