DC Schools Release Test Scores, US Education Department Joins Investigation

Amid an investigation into alleged cheating on Washington, DC standardized tests, the school district released figures today that show a relatively flat trend in progress over the last year.

Preliminary results from the 2011 District of Columbia Comprehensive Assessment System show that 46.4 percent of secondary students scored proficiently in math and 44.2 percent scored proficiently in reading -- up 2.7 and 1 percentage point from last year, respectively.

For elementary school students, 42.3 percent scored proficiently in math and 43 percent were proficient in reading -- down 0.8 and 1.1 percentage points from 2010, respectively.

When compared to 2007 scores, however, both secondary and elementary students have shown strong, steady improvements. Seventh and 8th graders, specifically, were the lowest performing grades four years ago -- now they're the highest performing grades.

"We will work hard as a school district in the next school year to build on the progress we have made. We have a long way to go in reaching our goals, but our seventh and eighth graders have proven we can make steady progress," DCPS Chancellor Kaya Henderson said in a statement Friday.

The DC CAS is administered to all public and public charter school students in the 3rd through 8th grade, and in the 10th grade.

The results released today come the day after a district official revealed that the U.S. Department of Education has joined the local investigation into allegations that the district's steep improvements on the DC CAS between 2008 and 2010 were the result of widespread cheating.

Officials called for the probe in March after USA Today reported on excessive erasures on answer sheets from more than 100 schools.

In Atlanta, a two-year investigation concluded this week, revealing widespread cheating among teachers who corrected students' answers on standardized tests. The report traces cheating back to 2001 and implicated 38 principals. Another 82 educators also confessed to dishonest practices.

Comments

[Broknrekord3]

04:08 PM on 07/12/2011

Foucault's panopticism is the only reasonable idea to keep people from cheating, but that requires big brother that's much too much for most of us. Either you can have your freedom, or you can have more honest results.

[Dlollar67]

10:10 PM on 07/10/2011

If such increases in test scores is considered suspicious enough to investigate for cheating, then how can Rhee and Duncan say with a straight face that they expect all school to be 100% proficient in 3 years without cheating? Baffling.

[poeticjustice4all]

01:30 PM on 07/09/2011

Yawn.

[TINA ANDRES]

04:50 PM on 07/08/2011

The only way to end cheating scandals is to have an independent group come in and administer the tests. I'm sure there are plenty of companies that would line up to provide the service at astronomical rates. If this is what is important than let's go ahead and pay them more than teachers, we could just fire a few thousand more teachers and pack the classrooms to pay for it. We're giving it all away to private interests anyway.

[sawyer0413]

08:25 PM on 07/08/2011

Actually, there is no way to eliminate cheating. Cheating can occur within the system you described. As long as the stakes are high, cheating will occur. Read Freakonomics. It has an excellent discussion on the topic.

[TINA ANDRES]

08:46 PM on 07/08/2011

Yes, but there needs to be a way to keep the accusations away from the teachers in the current climate. I'm sorry if anyone thinks I was serious with that post. Keeping the accusations away from the teachers who are actually teaching the students the information for the test should be a priority though. I've read Freakonomics, cheating is basically human nature anyway.

[Bolan]

04:08 PM on 07/08/2011

I feel sorry for all those teachers in DC and Atlanta. They have an insurmountable job teaching all those young democratic "youngins" anything.

[UnRed]

09:26 PM on 07/11/2011

"democratic"?? You mean, as opposed to "socialist youngins"? Think someone should tell this brain trust that the political party is capitalized? Nah...kind of like talking to a dining room table...it's not especially capable of understanding

[SmotPoker]

04:01 PM on 07/08/2011

No need for that silly book learning in Montana and other red states.

[oilfield]

03:39 PM on 07/08/2011

pretty bad when the cheat scores are still bad.

[Gregory Gull]

03:28 PM on 07/08/2011

As Deming often said, there are 3 ways to realize better results
1) fudge the numbers
2) rig the system to show better numbers
3) improve quality

All the talk and emphasis on raising standards, results/test scores, accountability for results promote #1 & 2 above. It is not a surprise there are scandals!

http://www.forprogressnotgrowth.com/2011/06/18/why-educate/
http://www.forprogressnotgrowth.com/2011/03/07/want-to-improve-quality-listen-up/

[demisfine]

02:44 PM on 07/08/2011

Michelle Rhee comments in 3, 2, 1

[Computer Geek]

02:31 PM on 07/08/2011

The problem with testing is that memorization isn't cognition. Memorizing to pass the test and then forgetting it all a week later doesn't help.

Yes, I know this story is about cheating in reporting the results or the tests taking itself, but the bigger point is that the methodology of learning by rote to pass a test is incorrect to begin with.

[Wuggly Ump]

02:49 PM on 07/08/2011

Couldn't agree more.

[bertha112]

04:57 PM on 07/08/2011

Best comment yet! I agree totally.

[mlaiuppa]

02:03 PM on 07/08/2011

So, have they stopped cheating or are just better at hiding it?

[sawyer0413]

08:29 PM on 07/08/2011

Neither. They simply move it higher in the chain. What isn't shocking is cheating. What is shocking is how long it took to identify. Which then raises the question, how high did the cheating go to possibly conceal the test cheating? Seems awfully suspicious that it took a reporter to find cheating, and she isn't part of the system. It is a thing that makes you go ... hmmmmm.

[acajoe]

02:18 PM on 07/08/2011

THATS COLD

[Arthur L]

03:30 PM on 07/08/2011

no. it's thinly veiled racism.

[AntagonisticallyYours]

01:43 PM on 07/08/2011

No matter how the scores have increased (based on truth or cheating), these scores are still embarrassing. Forty-six percent proficient in math? Forty-four percent proficient in reading? How do you learn any other subject (i.e., history, social studies, science), if you can't read? And sorry, it's the parents, not the schools. Unless these kids have some kind of learning disability, there is no reason for such low scores, except for the fact that the parents and the students don't care. I'm not saying there aren't bad teachers our there, but teachers can only do so much.

[mlaiuppa]

02:06 PM on 07/08/2011

Set the bar high enough and you can fail as many as you want.

NCLB is designed to make all schools look like failures so that public education can be privatized for profit. That was the original goal; to funnel federal education tax dollars into private corporations. So far so good. Mandatory standardized testing is really sucking up the money.

[capitalismisntamoral]

02:27 PM on 07/08/2011

Exactly! No wonder you have so many fans.

[demisfine]

02:47 PM on 07/08/2011

If I could fan you five times I would.
Once will have to be enough.
The GOP sends their kids to private schools already.
They want vouchers to give them back their tax dollars so they can use those funds (public funds) to pay for their private schools.
And just for kicks, they'll have their buddies open for profit schools to get some money for themselves out of the deal.
They will gut and dismantle our schools in very short order.

[VA Jill]

01:31 PM on 07/08/2011

You can't win. Low scores = your schools are failing. High scores = unless you live in a wealthy area, it's automatically cheating.

[Arthur L]

03:31 PM on 07/08/2011

No. In Atlanta, the high scores meant they were cheating.

[sawyer0413]

08:32 PM on 07/08/2011

Actually, high scores do not automatically indicate cheating. In fact, with analysis, the demographics, and even the right answers, are irrelevant to detecting cheating. The analysis looks for patterns that should or should not be in the data. Once those patterns are detected, further analysis can help determine if cheating occurred. Even so, a pattern can occur naturally, just ask Jaime Escalante and Garfield High.

[giono]

01:29 PM on 07/08/2011

The real legacy of Michelle Rhee......