Much has been said and written about "psychological testing" as a means of identifying and weeding out potentially lethal pilots (and other mass murderers).
Here is the sad truth: It can't be done with an acceptable degree of accuracy, and if it were attempted and implemented on a large scale there would be hell to pay in the form of a great deal of litigation and astronomical damage awards on the grounds that those identified and "weeded out" were slandered, wrongly accused and prevented from serving in their chosen profession. People can't be punished for what they haven't done (yet).
The reason for this is that, as psychiatrist Dr. Richard Friedman has pointed out (Star Tribune, March 31), murder-suicide is an extremely rare event, and thus defies prior identification — many, many "false positives" would be identified for every "true positive" discovered. (For a scholarly explication of this problem, see a seminal paper on it by Meehl and Rosen in the 1955 psychological literature.) To make the picture even more bleak, suicide is not really being "prevented," despite the proliferation of hot lines, suicide-prevention (sic) centers, and public forums on the causes of/reasons for suicide, as well as the availability of effective treatments for depression, which is a major precursor of suicide.
The annual incidence of suicide in the U.S. population has hovered around 11 per 100,000 for the past century or more. Add to this an estimate that the 30,000-plus people who committed suicide (1980s data) took more than 5,000 others (relatives and strangers) with them, and it is clear that we have a major public health problem that defies easy answers and solutions.
Makes one wonder if suicide, and the collateral damage it causes, is part of the human condition.