Larc,
This does indeed sound a fascinating area of exploration and would seem to attack at root the difficulty that most of us have is assessing the worth of a persons work without the introduction of the common 'human' predjudices that often mar our judgments.
It has to do with the accuracy of human jugement. The specific data I am looking at, has to do with the judgement managers make when they appraise their employees performance. The standard way of determining their accuracy is to compare their ratings with someone else's ratings who also knows the employees.
Just a question or two to clarify the MO as in conventional analysis it seems that 'knowing' the employee is crucial to the accuracy of report.
If the work that a person does is more easily assessed, for example a salesman hits his target of 30,000 widget sale per month, is punctual good natured and dedicated, would the system work as well for example with a social worker, whose qualities are measured against much more abstract assesments and more difficult to quantify?
Would a specific model need to be set up for 'types' of employment? If so how would it address the change in social attitudes toward performance that that alter with time. For example, the sucessful manager of 20 years ago would perhaps be judged sorely lacking in life skills in todays business world. Does the model 'average out' performance?
Forgive my ignorance, this area of expertise is way outside my own, but it does interest me.
Best to you Larc - HS