SJC-1103 minority police versus state exam
My thoughts regarding LOPEZ v. COMMONWEALTH Pedro LOPEZ & others v. COMMONWEALTH & another. SJC-1103 (decided November 9, 2012).
Case is described at 'Minority officers win round in legal battle over promotion exams - Metro - The Boston Globe'.
Assuming that exams in use do not predict job performance, shall those who utilize exams that do not predict job performance be at this time punished by way of plaintiffs and courts and financial penalties?
How guilty shall an employer be, if he utilizes exams that do a moderately effective job of predicting employee performance, if the exams he uses are inferior in terms of predicting job performance compared to exams he does not use?
Should government agencies be obliged to measure job candidates, using measurements that do a superior job of predicting job performance?
How shall it be possible for govt to discover which exams predict job performance, if govt is inhibited in terms of experimentation, due to fear of being sued for making a mistake in choice of exam?
An exam might be a success, in terms of weeding out those who are unfit to be sergeant, even if looking at those who score high enough to become sergeant, the score differences do not correlate with job performance.
What if a situation is created wherein performance on the exam corrrelates with job performance, but also results in persons unfit to assume the post, being promoted to the post?
What if the exam that correctly predicts job performance, costs a hundred million dollars more than the exam that does not predict job performance? Would it be worth it to utilize the expensive exam?
You could probably find those who evaluate performance in a certain way who find a correlation between job performance and the exam in use. And you might also find those who evaluate performance in such a way that they see no correlation between job performance and the exam in use.
One could argue that performance as a police sergeant is not a thing wherein sergeants end up being ranked from 1 the best, to 2 the 2nd best, to 3 the 3rd best, all the way down to 100 the worst. Rather one might say, that with police sergeants, performance as realistically measured is a simple split between the inadequate and the adequate.
The number of performance-level classes you create in your stats (i.e., 'excellent', 'very good', 'good', etc), could effect your conclusion regarding whether the exam accurately predicted performance.
An exam that is good at weeding out misfits could be a bad predictor of job performance amongst those whose performance is at least adequate. An exam that is bad at weeding out misfits could be a good predictor of job performance amongst those whose performance is at least adequate.
Elimination of an exam in use, because the exam fails to correlate with job performance could result in the use of an exam that is even worse, one in which high scores correlate with low job performance.
Could be, that plaintiffs and courts exerting power over the decisions re what is considered to be an exam that predicts job performance, by way of intimidating multi-million dollar lawsuits, would produce inferior results compared to simply demanding that government agencies make good faith efforts to develop exams that accurately predict job performance. Did anybody higher up in the command-chain, ever order Mass Human Resources Division to develop exams that predict job performance?
In police-type systems, persons get promoted to the next grade up in rank. Could be that the exams don't predict sergeant performance well, but do predict lieutenant or captain performance fairly well.
A person's value to society could be said to consist of more than simple 'job performance'.
By exams I mean in this context, all systems that weigh various criteria used in judging who shall be given a job.
Case is described at 'Minority officers win round in legal battle over promotion exams - Metro - The Boston Globe'.
Assuming that exams in use do not predict job performance, shall those who utilize exams that do not predict job performance be at this time punished by way of plaintiffs and courts and financial penalties?
How guilty shall an employer be, if he utilizes exams that do a moderately effective job of predicting employee performance, if the exams he uses are inferior in terms of predicting job performance compared to exams he does not use?
Should government agencies be obliged to measure job candidates, using measurements that do a superior job of predicting job performance?
How shall it be possible for govt to discover which exams predict job performance, if govt is inhibited in terms of experimentation, due to fear of being sued for making a mistake in choice of exam?
An exam might be a success, in terms of weeding out those who are unfit to be sergeant, even if looking at those who score high enough to become sergeant, the score differences do not correlate with job performance.
What if a situation is created wherein performance on the exam corrrelates with job performance, but also results in persons unfit to assume the post, being promoted to the post?
What if the exam that correctly predicts job performance, costs a hundred million dollars more than the exam that does not predict job performance? Would it be worth it to utilize the expensive exam?
You could probably find those who evaluate performance in a certain way who find a correlation between job performance and the exam in use. And you might also find those who evaluate performance in such a way that they see no correlation between job performance and the exam in use.
One could argue that performance as a police sergeant is not a thing wherein sergeants end up being ranked from 1 the best, to 2 the 2nd best, to 3 the 3rd best, all the way down to 100 the worst. Rather one might say, that with police sergeants, performance as realistically measured is a simple split between the inadequate and the adequate.
The number of performance-level classes you create in your stats (i.e., 'excellent', 'very good', 'good', etc), could effect your conclusion regarding whether the exam accurately predicted performance.
An exam that is good at weeding out misfits could be a bad predictor of job performance amongst those whose performance is at least adequate. An exam that is bad at weeding out misfits could be a good predictor of job performance amongst those whose performance is at least adequate.
Elimination of an exam in use, because the exam fails to correlate with job performance could result in the use of an exam that is even worse, one in which high scores correlate with low job performance.
Could be, that plaintiffs and courts exerting power over the decisions re what is considered to be an exam that predicts job performance, by way of intimidating multi-million dollar lawsuits, would produce inferior results compared to simply demanding that government agencies make good faith efforts to develop exams that accurately predict job performance. Did anybody higher up in the command-chain, ever order Mass Human Resources Division to develop exams that predict job performance?
In police-type systems, persons get promoted to the next grade up in rank. Could be that the exams don't predict sergeant performance well, but do predict lieutenant or captain performance fairly well.
A person's value to society could be said to consist of more than simple 'job performance'.
By exams I mean in this context, all systems that weigh various criteria used in judging who shall be given a job.
posted by David Virgil at 4:34 AM
0 Comments:
Post a Comment
<< Home