Last month they were ecstatic in Georgia because 90% of the third graders had passed the reading test and were gonna be able to go on to the fourth grade. Since Georgia ranks 50th in SAT scores this was finally some good news, at least to the Georgia educational establishment. I wrote about it here. I was not very impressed. I didn't think that only nine out of ten students passing the reading test was sumpin' to crow about. We should be able to do better.
After writing that post, I got mail from many of my readers who sent me articles from their states with the same type of results. The 10% failure isssue seemed to be pretty common across the country. Now I find out that it was worse than I thought according to this article in Wednesday's Atlanta Urinal and Constipation.
Georgia third-graders needed to answer correctly fewer than half the questions on the state's reading test to be promoted to fourth grade, educators disclosed Tuesday.The "cut score" — the number of correct responses needed to pass — was 17 of 40 questions, or 42.5 percent.
Huh? WTF? 17 out of 40? 42.5%? That's a passing grade? And still one child out of ten failed the test? What kind of bullshit is that?
Such a cut score is not unusually low for a standardized curriculum test, which measures how well students know the material covered in the classroom. While classroom tests require 70 percent for passing, a cut score between 40 percent and 60 percent is typical on standardized tests.
Yeah but scoring less that 50% is considered passing? Even on a standardized test I think that's bullshit.
The state declined Tuesday to make immediately available information that would show how difficult, or easy, the questions on the third-grade reading test were. Officials said they would provide that data later this week.
Yeah I want to see that. I'll see if my cats can get better than 42.5 fucking percent.
After repeated requests for the information from The Atlanta Journal-Constitution, the state school board voted Tuesday to release the cut scores. Ashe said the state's initial refusal to release the information had raised questions about the test, which was used for the first time to determine whether students in Georgia got promoted.
Yeah. If I had been the state school board I sure wouldn't want to admit that 42.5% was a passing grade. I also wouldn't want to admit that the schools were so bad that one out of ten students couldn't even get 17 out of 40 questions right.
So how is the cut score determined?
In July 2002, groups of teachers gathered to set cut scores. The teachers took the test and then evaluated each question, estimating how many children would likely get it right. Their responses were used to set cut scores, which varied depending on the test. In fifth-grade social studies, students needed to answer one-third of the 60 questions correctly. Fourth-grade math required 63 percent correct answers.
Whoa! Back up! In fifth grade social studies answering one third of the questions is considered passing? No wonder our population thinks we live in a democracy rather than a constitutional republic. The teachers ain't teaching, the students ain't learning, and the state school board is fixing the test results. Here we see gummint schools and the teachers unions in action. No performance and no accountability. What a disgrace!
How about other states?
Georgia officials attempted to survey the policy of all 50 states and Washington, D.C., on making scores public. Of the 26 states that responded, 14 released cut scores to varying degrees, and 12 kept the cut scores a secret. Texas, often cited as a model for Georgia, makes cut scores public.
And ....
This year, Texas third-graders needed to answer 23 of 36 questions correctly — 64 percent — to be promoted.
64% compared to 42.5%.
Tuesday's vote by the school board authorized the release only of the cut scores for Georgia's curriculum tests, which are given in grades 1 through 8. A school board committee will consider at a meeting July 7 whether to release cut scores for End of Course Tests, given in high school, and the Georgia High School Graduation Test.
I hope they're higher than 42.5% but I'm not holding out much hope.
And John Fonda Kerry just pandered to the National Education Association (Motto: We take education out of the schools) and said no way would he support vouchers.
Most of the kids who passed fifth grade social studies did not have the slightest idea what he was talking about.
Mr. Wilson,
The test scores you are seeing are not unusual for a standardized test. It takes skill or incredable bad luck to score less than 25% (pick a letter and answer everything that one). Now you want the bell curve to fall in the middle because you are tring to find at just what level the students are at (25%+35%=60%) and the passing grade will be lower than the middle of the bell curve (45% to 35%). If you were to give this test to various ages of kids you would expect the first graders to score about 30% while the collage kids should hit around 90%. Writing these tests are a royal pain and I never enjoyed making them up but when you are tring to find out the level of knowledge of students coming into a class (for me it was Navy Electronics Techs, Sonarmen and Radiomen at NavSubScol, Groton CT) there is no other way. If you write the tests right there are questions nobody misses and some that nobody get right. Now what I have said up until now is for a test written to determine the knowledge level, if you are talking a test to determine if information covered is being in at a useful level then the acceptable score goes to 75% or 80% for a passing score. (Some of the tests we gave required a 90% to pass)
Sadly, I am not surprised by any of this. Then we wonder why somany high schoolers are not prepared to enter the business world, let alone going to college. The poodle's opposition to vouchers, shows he is in the pocket of the NEA and taking the minority vote for granted. After all, it is the minorities who are faring the worst in the NEA fifdom called public schools.
Sailor
Posted by: A Sailor in the Desert on July 1, 2004 05:25 AMKeith,
I had a Physics professor in college who came right out and told us he was going to write our tests to get a median score of 50%. How the class actually performed is a different story but that's how the tests were written.
A distribution curve like this not only allows you to identify the idiots but also the gifted.
Posted by: Ralph Gizzip on July 1, 2004 07:23 AMIf you can't understand the meaning of 50%, you won't even know what majority means, particularly in the sense of a democracy. Now, given that audience, can you imagine trying to explain the electoral college?
Notwithstanding, when they reach 18 they'll still be able to vote. Should they vote? I think not, but my default rule is if they don't understand the 3 branches of government and their respective responsibilities they shouldn't be allowed to register to vote. Of course, the USSC won't let that happen - they won't even allow a literacy test as a threshold for voting.
"Get out the vote" drives are mostly "dimocrat" (Denny's accurate word) efforts to flood the polls with those milking off the teat of the welfare state. Wanna do something really useful? Channel those efforts instead to voter education. Not partisan issue based education, but civics and fundamental government lessons, plus maybe a dose or 2 of U.S. history. Oh, thats right, they "learned" all that in school. Silly me.
Posted by: Daniel on July 1, 2004 12:19 PM