Judge rules Boston police-lieutenant exam discriminated against black applicants
A federal judge ruled today that a Boston Police exam given in 2008 unintentionally discriminated against ten black Boston Police sergeants trying to become lieutenants and gave the city 30 days to come up with an acceptable remedy.
In his 82-page decision, US District Court Judge William Young wrote that while the city did not deliberately craft a multiple-choice exam to screen out blacks, the net effect of the multiple-choice exam was the same: They were at a disadvantage before they even sat down to answer the questions.
Young wrote:
This is a profoundly important case, one that evokes the finest of our nation's aspirations to give everyone equal opportunity and a fair shot. In deciding this case, the Court first emphasizes what this case is not about: this is not a case about conscious racial prejudice. Rather, the Plaintiffs' case is rooted in their allegation that the seemingly benign multiple-choice examination promotion process, while facially neutral, was slanted in favor of white candidates.
In 2008, 91 BPD sergeants took the promotional exam - 65 white, 25 black, 1 Hispanic. Some 33 sergeants earned promotions based on the exam, 28 of them white, only 5 black. Ten black sergeants sued in 2012.
Young said he agreed with contentions by the sergeants' testing expert and lawyer that the lower scores of the black officers could be due to several causes, including the fact that the white officers had had a greater chance of having taken the test before. And while emphasizing he found no overt discrimination by the BPD brass - including current Commissioner William Evans - who wrote the questions, he noted that BPD has "has a history of discrimination against minority police officers" that might have made minority officers reluctant to take earlier tests.
Young continued that "whatever the causes, the so-called 'achievement gap' is real," and statistically signficant enough to warrant a finding of "disparate impact" on the black sergeants.
He continued that the 2008 test wasn't even very good at its stated task of showing who was most qualified to become a lieutenant.
The Court finds that the City fell short at many stages of the test construction process. ...
Because job knowledge is only a limited part of the job analyses for the role of lieutenant, the Court agrees with [the plaintiffs' expert witness] that the 2008 exam skipped over critical skills and abilities, including interpersonal skills, presentation skills, reasoning and judgment skills, oral communication skills, analytical skills, ability to give constructive criticism, ability to speak in front of groups, ability to counsel subordinates, ability to counsel and comfort families of victims, and ability to make sound decisions quickly. ... The Court therefore also agrees with [his] opinion that, as a result of [the city's]'s decision not to test many critical skills, a high score on the 2008 exam simply was not a good indicator that a candidate would be a good lieutenant.
Young continued that if the city cannot give the sergeants an acceptable "remedy" within 30 days, they could propose their own, to which the city would then have another 30 days to respond.
Attachment | Size |
---|---|
![]() | 206.57 KB |
Ad:
Comments
This shit is getting old
A,B,C or D is not a racist question.
Then explain the racially-correlated results.
.
I will
But then you'll just call me a racist.
That very same achievement gap the ruling cites: the black applicant pool is (I assume) drawn from a population with statistically significant lower test scores everywhere else, so it isn't surprising that they would be less likely to have lower test scores here.
Now why is that? Well, a few centuries of oppression and neglect aren't consequence free.
Still doesn't make the test questions racist. For the same reason that a police physical fitness test isn't "ableist."
Discuss.
Blacks have same IQ distribution
The police should not discriminate against people because they've been screwed over before.
That said, it doesn't help when the white kids are given the answers ahead of time ... which used to be standard practice.
Kids?
Kids?
Well then
Maybe it might be a better idea to focus on education to close the achievement for future generations instead of filling jobs with people who are woefully unqualified. As form the exam, I'm fairly certain questions were dealing with actual job-related material, not with naming the past ten leaders of KKK.
I remember a study (a very
I remember a study (a very long time ago in a class) that was able to show that people of all races generally fared about the same on standardized tests. However, when a group of black students where told that white students did had better success rates/ grades at the test, the black student ended up with much lower results than both their control group. The same phenomenon was seen when white students where told that Asian students excelled on the test - the white group of students under performed the their control group peers. It was a pretty interesting double blind test. Just food for thought that sometimes it really can be very mental.
Priming studies (of which
Priming studies (of which your stereotype threat study is an example) are among the psychology studies that failed to replicate in the recent review of the reliability of psychology research. The effect is either very small, or non-existent.
On the other hand, there is a standard deviation difference in the mean of the IQ distribution between Asian, European, and African, populations and that is not specific to the U.S. and has been replicated countless times over several decades under the strictest scrutiny, because the finding is *so* uncomfortable.
"Discuss"...
Could you be any more patronizing?
Read the decision. Please. This is more than just an ABCD thing.
I can be
but I choose not to be. There's subtlety in everything. Crying racism is a blunt instrument. Very few black American Nobel laureates, too. Must be that the Nobel committee is racist?
most notable line in this decision
If white officers, either through deliberate design or inadvertently, had more chances to take the test than non-whites, and greater chance to retake it after initial failure, then you don't need to exercise your bigotry muscles to explain why they did better.
My bigotry muscle
Says they failed the first time and tried to take it again next time the exam was offered. Should we not let them do it?
I won't call you a racist
I won't call you a racist (yet), but I will call you lazy. Read the article. Note what is said about the flaws of the test, which only addressed "job knowledge" and none of the many skills that you'd like a police sergeant to have. Do you really think it's a good idea to promote officers based on "A, B, C and D" and ignore their skills relating to decision-making, dealing with subordinates, communications, etc.?
Oh
So now you don't need to know your job to get promoted?
Let's see
So, let's suppose your job is waiting tables in a restaurant. In order to decide if you should be promoted, we test you on your knowledge of where the soup spoon goes. We do not test you on your customer interaction skills, or your ability to deal with stressful situations, or your ability to balance a tray full of dinners, even though all of these things have a great deal to do with whether you're a successful waiter. No, all we care about is that soup spoon.
Stop trying to pretend that this is a no-brainer. It's a victory for those who exploit ignorance and bigotry, and that's a very different thing.
Riiiight
Because all the exam tests is how many bullets you load into the mag, and how to properly wear your holster. To hell with all those cops who know their job, let's hire a gang leader because he has great leadership skills. Or a conman because he can interact with people really well.
So, you didn't read the article, and you don't intend to
OK, so we've established that you're what we call a "low-information" individual who refuses to read what's right in front of you. In other words, you're willfully ignorant. If you weren't, you'd read the article and you'd see that you're completely wrong about what the exam does and does not test. You've also established that you don't want to know what the truth is, because if you did, you might have to abandon your stupid prejudices, or at least stop blathering them as if they were fact. Ignorant and loving it -- that's you, and talking to you is like talking to a coffee table. Done with you.
It is possible for an
It is possible for an assessment instrument to be statistically biased, even if the content or structure of the instrument is not obviously biased itself. Statistically biased means that there are performance differences between the candidates, even after accounting for their true ability differences. For example, there are often real differences in K12 reading and math that follow demographic subgroups, but if there are differences in test scores beyond what can be explained by those true differences, then the test is statistically biased and the results are not valid. With a large enough sample size, one can even detect statistical bias for individual questions on a test. The MCAS technical reports have these analyses at the end for anyone who wants to read some really boring stuff.
What probably made this such a simple case is that these candidates are not randomly selected from the US population. They all presumably met the same eligibility requirements; i.e., they were already found to be competent enough to be police officers by previous tests or interviews. That a score difference by racial subgroup would suddenly manifest in just the lieutenant exam from such a homogenous group is a huge red flag, and from there it's a very simple statistical analysis to demonstrate the bias.
What blows my mind is that BPD did not catch this and remedy it themselves. Almost all high stakes tests have a test developer or psychometrician involved in the design and analysis, and I would have thought this bias would have been identified immediately following the test and that the candidates would have had an opportunity to retest on a new instrument.
Four thumbs up? That's a new
Four thumbs up? That's a new record for me!
My Opinion
Multiple choice exams are the easiest exams one can take, they GIVE you the answer right there and if you studied it should be pretty easy.
I don't know anything about police exams but I do know a question with a choice of answers with the correct answer included is not discriminatory.
How does the judge know his
How does the judge know his extra testing wouldn't have resulted in 65 White, 0 Black, 0 Hispanic? Seems a bit of a hypothetical ruling. Maybe they should just have a black person write the multiple choice questions. Is the only qualifyer the multiple choice test? I can't imagine any job not having an interview process, but that's gov't I guess.
He actually discusses that
I know it's 82 pages and all, but if you dig down into the decision, he talks about possible outcomes of testing like that.
The last promotional exam
The last promotional exam given by the BPD was broken up into three sections- multiple choice, oral board and essay. Was the breakdown different from the 2008 test? Or was it similar?
You sure did dig deep to find
You sure did dig deep to find anything that could seem anti-cop didn't you? Why do you hate the police Adam?
Oh, yeah, I dug reeeeeeel deep to find this
It's just amazing how far down I had to dig to find this obscure ruling by an obscure judge in an obscure court located soooo far away from here.
Like, it's not everybody who has access to the Mass. Lawyers Weekly RSS feed, where I read yesterday afternoon about this ruling.
And God only knows how the Boston Globe found out about it.
This sounds like a hypothetical comment...
from someone who hasn't read the ruling
ridiculous
Ridiculous.
These are men (or women) that have already passed 1 civil service exam to become sergeants. They know what the test will entail because guess what? They have already taken an exam AND all candidates are provided the SAME list of study materials and books to study from.
The white officers were not provided any additional information which enabled them to get higher scores, nor were those black officers that scored well enough to be promoted from this test. Those that scored better STUDIED MORE, PREPARED MORE, AND EARNED THEIR HIGHER SCORE.
I haven't read the full 80 something pages of the decision yet, but the rundown above does not seem to provide anything other than a almighty Judge providing people who could not be bothered to study hard with an excuse as to why they didn't score well.
Study more and get ahead, don't blame your own laziness or lack of studying on your race. Everyone has the same opportunity when they are given the same materials.
Higher score at WHAT?
Higher score at WHAT? At a test that evaluates only one area of knowledge required for the job, and that disregards many other areas of skill and knowledge that are arguably much more critical? This is like the kid who can't do basic math bragging about the medal he won in the first grade spelling bee.
So the material that the exam
So the material that the exam focuses on may not be the important material and the exam should be rewritten, what does that have to do with the racial discrimination aspect?
That's in the ruling
That's addressed in the ruling. Did you read it, or even just the excerpts in Adam's article?
Just Adam's excerpts;
Just Adam's excerpts; "whatever the causes, the so-called 'achievement gap' is real," those who did get promoted may have in fact, studied harder and were better prepared.
Cure worse than disease?
So, I didn't get through the entire ruling, but...
- My first instinct is a lack of understanding why or how a test like this can have racial bias.
- But for sake of argument, and because of the test results and judge's ruling, I guess I'll go along.
- A practical test is probably a better way to evaluate many of a senior officer's duties (vs a multiple choice).
- But that leads to the next problem; a practical test is also likely to be more subjective.
So guess where that brings us? If for any reason, the results of the next test aren't perfectly representative of the tested population, the next lawsuit will complain that the testers were too subjective.
Frustrating.
Absurd ruling for black Lieutenant candidates
It should be noted that black officers are hired from the same entrance exam (affirmative action notwithstanding) and graduate the same Boston Police Academy as their white colleagues. The only time that officers snuck onto the BPD in the modern era without a full Boston Police Academy was when Mayor Menino and Ed Davis authorized "lateral transfers" to get a few friends from the suburbs onto the BPD through the back door and a handful of the former Boston Municipal Police who qualified.
In order to take the exam for Sergeant, a patrolman must be in grade for at least two years. In order to take the exam for Lieutenant, a Sergeant must be in grade one year but most Sergeant/Lieutenant candidates have been on for a decade or more. During those many years, black and white officers receive equal in-service training and are availed of the same amount of free help toward the promotional exam. Any officer that pays attention to the academy, in-service training and the listed reading material has a good chance of promotion. To lower the standards for blacks is another example of the soft bigotry of low expectations.
Social justice demands
That Boston have less competent police lieutenants.
It's totally not fair that jobs should be given to people who give the right answers on tests, because all animals are equal, but some animals are more equal than others.
All levity aside, it's an interesting question. Think of the children. Will the futures of Boston's children be better served and secured by having more black police lieutenants or having more competent police lieutenants? Ideally, I guess, it'd be both, and I hope the solution can be that. Perhaps racially targeted tutoring services could be part of an acceptable remedy.
Of course, any solution that is agreed upon must result in more black lieutenants, or it (and its results) can be scrapped and everyone sent back to the drawing board, because the ultimate definition of appropriate job-related police lieutenant testing is now "black sergeants can pass it."
Again, I know it's a long decision ...
But the judge addresses these points and explains why the specific 2008 test was far from the best method to find qualified lieutenant candidates. If you do want to read about that, call up the ruling (the link is at the bottom of my post), skip the intro and all the boring statistical stuff that is really the heart of his decision, because of that pesky federal law, and head towards the bottom, where he discusses specifically why this was a bad test regardless of who passed it.
You're right
I should have read closer before posting. And I should have read up on Teal vs. Connecticut, on which it's based. I would have posted something different then. Because I would have known that the problem was never discrimination against minorities in actual promotions, but just on that test.
The city came into this with the outrageous argument that because analysis of the actual promotion rates demonstrates that the 2008 exam did not result in a statistically significant adverse impact in promotions, the court should rule that the city did not discriminate.
It's the same basic argument as in Teal, where the fact that a higher proportion of black candidates failed the test had _no_ impact on promotions, because black candidates were actually promoted at _double_ the rate as white candidates despite test outcome.
This is called the "bottom line" argument, because the bottom line is that a higher proportion of white candidates was not promoted. The rulings both threw out this argument because, basically, mean tests hurt fee-fees.
The court said, "Yes, you [statistically] got the job anyway, but so what? The test was still unfair, so you win."
Of course, this means that because the problem, in both cases, was never a problem with actual minority promotion, the solution does not need to involve actual minority promotion either, just changing the test component so more minorities pass it. And then back to the same rate of minority promotion - the "bottom line," which was never at issue.
The case should have come with a disclaimer - but instead of "no animals were harmed in the making of this movie," it would be "no greater proportion of minority candidates will actually be promoted if this case wins."
One for one promotions
Maybe the fact that some of these black sgts were promoted to sgt when the one for one practice was in play ? In the past a white patrolman promoted to Sgt with a score of 100, the next person to be promoted was not the white patrolman with the 99. It was the highest scoring black patrolman. So fast forward to the this lieutenant exam and you have black sgts who were promoted with lower scores previously but now it isn't one for one anymore .
I think everyone is missing the bigger point:
Doing well on these tests do not mean you will be a good supervisor. It is really that simple, and I think that is what the court is saying here. Assessment Centers were used on previous tests, and later tests, but were not used on these exams, and that is the issue.
Ever wonder why the public has never seen a police exam or promotional exam online? It is because these tests are horrible, and are basically designed for people who want to memorize 4 text books, 3 law books, and a regulation book for 8-10 hours a day for 200-300 days. You can't blame them, the salary increase is worth the 2 months of your life wasted studying, but if you actually saw the tests, you would see the problem.
(Disclosure: I have been promoted on two of these Boston exams, as a white male)
I've never taken one of these exams
but my profession is full of certification exams, most of which are full of silly "trick" questions that have nothing to do with your ability to perform the skills you are being certified to do. Yes, you can parrot back the certifying agency's core moral values, or list off some tangentially-related vocabulary terms, but actual skills for doing what you're certifying for? Very minimal.
I work in IT and certification exams are mostly irrelevant.
I think they are there mostly because the people who do the hiring have no idea what you do or how to judge your skills. Ridiculous tests of memorization provide them with a metric to at least ensure you had to read a book about your job.
I expect testing is also the easiest way to cover your ass in promoting people. Multiple choice tests are not a good way to judge fitness for a job, but they definitely help you avoid the accusations of favoritism, politics, and prejudice that can easily be raised when you evaluate people based on interviews and/or performance reviews.
To sum up the judge's
To sum up the judge's decision
How do I reach these cops?
Of the test takers, 71% were
Of the test takers, 71% were white and 27% were black. Of those chosen, 84% were white and 15% were black. The numbers are not significantly off from the racial makeup of the test population so how is this discrimination?
You did the t-test?
You know, the test of differences?
You could also read the report.