Arkansas Test Score Scam of 2009
Scores Improve on all Arkansas "Homegrown" Benchmark Tests; BUT on nationally Normed Test, Arkansas Scores as Low as 35 % -- 65% of Students Across The Nation Do Better Than Arkansas Third Graders in Language, etc.
by Debbie Pelley & Iris Stevens: Once more, as is done every year about this time, the Arkansas Department of Education and the media are touting the successful K-12 scores in Arkansas. As usual when the government control is involved, the truth is twisted into propaganda. It is true that students in Arkansas keep improving on one measure, the benchmark tests designed, controlled, and with scoring percentages set by the state. But on all other measures compared with students across the nation Arkansas fall far short. [See highlighted list below.]
Note that on these Arkansas math benchmark tests, students in 8th grade only have to get 28 points correct out of 80 points to be Proficient. (Proficient is equivalent to a B or in ADE terminology indicates students have met standards); Grade 7: 27 pts out of 80; Grade 6: 40 out of 80; Grade 5: 34 out of 80; Grade 4: 40 out of 80; Grade 3 - 35 out of 80. No wonder their scores are so high! The tests are graded on the curve; ADE sets the curve and changes it from year to year. (There are no scores for grades 1, 2 & 9 because there are no state designed benchmark tests for them. NCLB doesn't require it.) The categories are Below Basic, Basic, Proficient and Advanced. [Source]
In addition to such low scores required on the Arkansas Benchmark Tests, in many districts teachers' lesson plan books are monitored and their classes observed to make sure they are teaching every day the released items from former state annual tests. Teachers are now being instructed to teach nothing else but the test sample questions or the released test items. In other words, the Test Released Items Booklet becomes the entire curriculum. Then teachers are required to give several practice tests at regular intervals in preparation for the annual Benchmark Test. [Memos instructing teachers to teach only released test items & Download the "Released Items Booklets" for all grades]
Anyone who studies the facts and compares the Arkansas benchmark scores to any national measure can see that the so called educational improvement in Arkansas is just a scam. For example, the gap between Arkansas benchmark test scores and the nationally normed test scores (the test that compares our students to other students across the nation) is staggering. The scores in half the grades tested are more than 20% lower than
the percentage on the Arkansas benchmark tests. This is the second second year for this particular nationally normed test, and scores for last year were very similar and showed no significant improvement even though the Arkansas Benchmark test scores keep going up and up.
First note how low the language score is for the nationally normed test (SAT 10) comparing our students to the rest of the nation. How could anyone possibly think Arkansas has had any success when we have scores like the following in any area? For example,65% of the students tested across the nation scored better than 3rd graders in Arkansas; 66% scored better than 4th grade, etc. This is the 2nd year for this test. While the scores last year were almost the same. And these scores are the result of more than doubling educational spending in Arkansas from $1.4 billion in 1996 to $4 billion in 2004-05 (federal, state, and local spending) which equals at least a 3% sales tax increases in Arkansas.
2009 Scores on Nationally Normed Language Tests (SAT 10)
(comparing Arkansas students to other students across the nation - Source for these scores:)
3rd grade - 35%
It doesn't appear that all that money spent on pre-school has helped the scores.
4th grade - 34%
(In 1988 the language score for 4th grade on nationally normed test was 65%)
5th Grade - 43%; 6th Grade - 45%; 7th Grade - 48%
(In 1988 the language score for 7th grade on natioanlly normed test was 61%)
8th Grade - 43%; 9th Grade - 48% * All the scores in 1988 were similarly higher than 2009, but these are the only two grades in elementary school where all students were required to take the test as they are now so these grades, are the only ones that can be compared that far back. [ADE document with the scores in red font above]
* Review this link to see how high our nationally standardized tests have been in the past in table form. This table includes all nationally normed test scores from 1988 to 1995 for the grades that Arkansas used for accountability, 4, 7, & 10. In the fourth grade in 1988 on the nationally normed test, every score was above 60 %, and dropped down to 50% in 1995 after the state took over education and started teaching to the tests.
Note: language and reading are incorporated into one score called the Literacy score on the Arkansas State Benchmark Test. If the Language scores above were averaged in with the Reading Score on the Stanford Achievement nationally normed test below, reading scores would be even lower and the gap wider between the nationally normed Stanford Achievement Test (SAT10) Reading scores and the Arkansas State Benchmark Test Literacy Scores.
Reading Scores
Note: The gap between the math test scores on nationally normed tests (the one that compares students nationally) is not as wide as on reading, but 3rd grade nationally normed score is still 23 percentage points lower than State Benchmark test score; 4th grade 10 percentage points lower, 5th grade 11 percentage points lower, 6th grade, 14 percentage points lower, 7th grade, 12 percentage points lower, Grade 8 had 5 points higher on state test.
The following facts were gathered in 2007 but facts are still the same except for a point or two here or there. Again these are the results of increased educational expenditures that would equal a 3% sales tax increase in Arkansas.
The following paragraph about Kentucky sums up what is happening in Arkansas.
Debbie Pelley, Retired Arkansas Teacher of 27 years, and Iris Stevens, Retired Arkansas Teacher of 31 years. [Article with additional information]
Tags: Arkansas, Arkansas Department of Education, benchmark tests, education, K-12, Kentucky, Scam, test scores, testing To share or post to your site, click on "Post Link". Please mention / link to the ARRA News Service. Thanks!
by Debbie Pelley & Iris Stevens: Once more, as is done every year about this time, the Arkansas Department of Education and the media are touting the successful K-12 scores in Arkansas. As usual when the government control is involved, the truth is twisted into propaganda. It is true that students in Arkansas keep improving on one measure, the benchmark tests designed, controlled, and with scoring percentages set by the state. But on all other measures compared with students across the nation Arkansas fall far short. [See highlighted list below.]
Note that on these Arkansas math benchmark tests, students in 8th grade only have to get 28 points correct out of 80 points to be Proficient. (Proficient is equivalent to a B or in ADE terminology indicates students have met standards); Grade 7: 27 pts out of 80; Grade 6: 40 out of 80; Grade 5: 34 out of 80; Grade 4: 40 out of 80; Grade 3 - 35 out of 80. No wonder their scores are so high! The tests are graded on the curve; ADE sets the curve and changes it from year to year. (There are no scores for grades 1, 2 & 9 because there are no state designed benchmark tests for them. NCLB doesn't require it.) The categories are Below Basic, Basic, Proficient and Advanced. [Source]
In addition to such low scores required on the Arkansas Benchmark Tests, in many districts teachers' lesson plan books are monitored and their classes observed to make sure they are teaching every day the released items from former state annual tests. Teachers are now being instructed to teach nothing else but the test sample questions or the released test items. In other words, the Test Released Items Booklet becomes the entire curriculum. Then teachers are required to give several practice tests at regular intervals in preparation for the annual Benchmark Test. [Memos instructing teachers to teach only released test items & Download the "Released Items Booklets" for all grades]
Anyone who studies the facts and compares the Arkansas benchmark scores to any national measure can see that the so called educational improvement in Arkansas is just a scam. For example, the gap between Arkansas benchmark test scores and the nationally normed test scores (the test that compares our students to other students across the nation) is staggering. The scores in half the grades tested are more than 20% lower than
the percentage on the Arkansas benchmark tests. This is the second second year for this particular nationally normed test, and scores for last year were very similar and showed no significant improvement even though the Arkansas Benchmark test scores keep going up and up.
First note how low the language score is for the nationally normed test (SAT 10) comparing our students to the rest of the nation. How could anyone possibly think Arkansas has had any success when we have scores like the following in any area? For example,65% of the students tested across the nation scored better than 3rd graders in Arkansas; 66% scored better than 4th grade, etc. This is the 2nd year for this test. While the scores last year were almost the same. And these scores are the result of more than doubling educational spending in Arkansas from $1.4 billion in 1996 to $4 billion in 2004-05 (federal, state, and local spending) which equals at least a 3% sales tax increases in Arkansas.
(comparing Arkansas students to other students across the nation - Source for these scores:)
3rd grade - 35%
It doesn't appear that all that money spent on pre-school has helped the scores.
4th grade - 34%
(In 1988 the language score for 4th grade on nationally normed test was 65%)
5th Grade - 43%; 6th Grade - 45%; 7th Grade - 48%
(In 1988 the language score for 7th grade on natioanlly normed test was 61%)
8th Grade - 43%; 9th Grade - 48%
* Review this link to see how high our nationally standardized tests have been in the past in table form. This table includes all nationally normed test scores from 1988 to 1995 for the grades that Arkansas used for accountability, 4, 7, & 10. In the fourth grade in 1988 on the nationally normed test, every score was above 60 %, and dropped down to 50% in 1995 after the state took over education and started teaching to the tests.
Note: language and reading are incorporated into one score called the Literacy score on the Arkansas State Benchmark Test. If the Language scores above were averaged in with the Reading Score on the Stanford Achievement nationally normed test below, reading scores would be even lower and the gap wider between the nationally normed Stanford Achievement Test (SAT10) Reading scores and the Arkansas State Benchmark Test Literacy Scores.
| Year 2009 | Stanford Achievement Test (SAT10) Reading Scores (comparing students nationwide) | Arkansas State Benchmark Test Literacy Scores | Gap between scores - percent point lower on nationally normed test (SAT10) that compares students to rest of nation. |
| Kindergarten | Not given on ADE report | Not given in AR | |
| Grade 1 | 41% | Not given in AR | |
| Grade 2 | 41% | Not given in AR | |
| Grade 3 | 43% | 66% | 23% pts lower on nationally normed test |
| Grade 4 | 61% | 70% | 9 pts lower on nationally normed |
| Grade 5 | 53% | 68% | 15 pts lower on nationally normed test |
| Grade 6 | 44% | 67% | 23% pts lower on nationally normed test |
| Grade 7 | 53% | 63% | 10% pts lower on nationally normed test |
| Grade 8 | 50% | 71% | 21% pts lower on nationally normed test. |
| Grade 9 | 53% | Not given on ADE report |
Note: The gap between the math test scores on nationally normed tests (the one that compares students nationally) is not as wide as on reading, but 3rd grade nationally normed score is still 23 percentage points lower than State Benchmark test score; 4th grade 10 percentage points lower, 5th grade 11 percentage points lower, 6th grade, 14 percentage points lower, 7th grade, 12 percentage points lower, Grade 8 had 5 points higher on state test.
The following facts were gathered in 2007 but facts are still the same except for a point or two here or there. Again these are the results of increased educational expenditures that would equal a 3% sales tax increase in Arkansas.
- Only nine states scored lower than Arkansas on the ACT test in 2007. Arkansas dropped slightly in 2007, but the national average improved. (The ACT test is the culmination of K-12 instruction and is used by colleges all over the nation to award scholarships.)
- The graduation rate dropped slightly more than 10% in one year according to the last educational report in Arkansas (2005-2006) the next report won't be available for a few more months.
- College Remediation Rate is 51.6% (46.6% in 1995); National Average is 33%
- Only Nine States Scored Lower Than Arkansas in 8th Grade Math on NAEP Test; only sixteen states scored lower than Arkansas on the 4th Grade Reading NAEP Test, etc. in 2005. Other scores were similar. In 2007 our scores were even lower. Arkansas Department of Education Director said, "We have slipped a bit in terms of state rankings in all areas save mathematics at the fourth grade.
- Only 28% of AR students pass test on Advanced Placement (AP) courses, while 59.4% pass nationwide.
- Test Scores on nationally normed tests are 11 Percentile Points lower in 2006 than in 1990 (a 22% decrease) (Later scores are not yet available).
- In its latest State of State Standards in 2006, Fordham Foundation gave Arkansas an F on its Math Standards, F on U S. History Standards, an F on World History Standards, a D in Science and a C in English for an average of F.
- The National Report Card on Higher Ed. gave Arkansas a D+ in High School Preparation for College in 06 in its latest report.
- Educational Reforms Have Cost $1.4 BILLION since 2003, which equals about a 3% sales tax increase in Arkansas.
The following paragraph about Kentucky sums up what is happening in Arkansas.
In 1993 US Education Secretary Richard Riley called Kentucky a lighthouse for the rest of the nation and said President Clinton's Goals 2000 would help states duplicate what Kentucky has done and is doing. [Herald Leader, 1993 by Lucy May.] Kentucky has shown great improvement on their state tests over the years. However, note that five testing experts from five universities said Kentucky had been "misinforming, misleading, exaggerating and overstating student achievement."--------------------
Also, in 1994, in Jefferson County, the largest school district in the state, parents demanded that their children be tested for national comparison. (For four years, the students had been tested only with the KIRIS test developed by Kentucky) The national tests revealed that Kentucky students' comparative scores, after four
years of Kentucky's state test, had declined from an average of 7% below the national average to an average of 27% below that standard in reading and math across all grade levels in which the tests were given – second, sixth and ninth grades. In no area had scores improved on the nationally normed test. ("Test-score drop was worse than reported" by Beverly Bartlett in the Courier-Journal, 1995 and "Tests: State, national levels are different" in the Herald Leader by Lucy May 4/28/95) [Link to Kentucky info.]
Debbie Pelley, Retired Arkansas Teacher of 27 years, and Iris Stevens, Retired Arkansas Teacher of 31 years. [Article with additional information]
Tags: Arkansas, Arkansas Department of Education, benchmark tests, education, K-12, Kentucky, Scam, test scores, testing To share or post to your site, click on "Post Link". Please mention / link to the ARRA News Service. Thanks!
Posted by Bill Smith at 7:02 PM - Post Link
4 Comments:
This is correct Bill. I'm a high school math teacher in Arkansas and I get the results of this "successful" primary education system. A large majority of my students in the 10th through 12th grade lack basic math skills.
They don't know how to add, subtract, multiply, or divide without a calculator. When they can't understand these basic concepts, it's impossible for them to learn the more advanced mathematics that we're making all kids take now.
If you look at the Arkansas "Frameworks," the list of topics and concepts teachers are required to "cover" at each level, you'll see a mind-boggling jumble of concepts that are often too advanced for the age-level of the kids. Rather than seek to give the students a firm understanding of basic concepts, the standards are designed to "show" them a wide range of advanced topics. The result is very little understanding of anything.
We often say we're asked to teach a curriculum that's "a mile wide and an inch deep." And our kids' performance on national and international tests reflect it.
I agree that if you're going to measure performance using standardized exams, all exams must be scored by similar rules. Otherwise it's just "lies, damned lies, and statistics". Standardized exams have great power when used correctly. They shouldn't be warped around political agendas.
Although I agree with the premise that Arkansas has education problems, I find that there is a fundamental error in the reasoning. A colleague wrote about it back in July and said:
"In an entry made several days ago, Debbie Pelley made a common mistake and a very serious one by comparing two completely different ways of getting test scores. AACTAP is a criterion-referenced test. This is the same type of scoring you have in every classroom when a teacher hands out one of their own tests: how much of this material did you learn and what percentage out of 100 did I get for knowing X number of problems out of Y number of problems.
SAT9 and its kin are norm-referenced scores. They are just rankings and are not measurements of knowledge per se. You are simply being compared to other students nationwide. The flaw in this thinking is that it is a crime to be below 50% which people confuse with how well you know the material. It is entirely possible to know 75% of the material, yet rank below 50% because so many more know 76% or more of the material.
An easy example is the selling apples examples: I have a 100 people, each with 100 apples to sell. John sells 82 apples out of his 100. That is his criterion referenced score. His ranking however is the 52nd percentile, meaning that 48 apple merchants sold more than he did. That is a norm-referenced score.
It is fallacious to compare the two numbers and insist one number is in error because it doesn't agree with the other."
Thanks for being concerned about Arkansas education but please don't confuse norm and criterion-referenced tests scores again because it is such a basic difference and no expert would use such as a basis for a theory for fear of ridicule.
Anonymous says SAT9 and its kin are norm-referenced scores. "They [norm referenced tests like SAT] are just rankings and are not measurements of knowledge per se." My answer, " If norm referenced tests [NRT] are not measurements of knowledge, then why have states and countries used them for at least the last 50 years to see if student achievement is measuring up to the expectations of the public and the schools. Norm referenced tests were e used for a number years as the basis of knowledge in Arkansas under their accoutnability system implemented in the 80's.
Norm referenced tests (NRT) are used to see if the criterion referenced tests (CRT) are measuring what they are supposed to measure. For example, if students of an individual teacher in a school system are making mostly A's, and yet make below 50% on a NRT test which meassures students to other sudents across the nation, the school administrator would believe the teacher was inflating her grades. If her students scored say a 35% on an NRT test as Arkansas students did in a couple of grades in language, then it is positively evident that the teacher is inflating her grades. That is the point I was making about the Arkansas CRT and the NRT tests in Arkansas. Arkansas is inflating the scores on the criterion referenced tests and deceiving the public. And not only is Arkansas not doing well on nationally normed tests, but by almost every other measure of knowledge where they are compared to they nation, they are falling behind. Therefore, there is only one conclusion. The scores on the Arkansas Benchmark Tests are highly, highly inflated.
Yes, the NRT and CRT are different types of tests, but there should be positive correlation. The NRT is a comparison of students to other students in the nation. Their tests are designed by private companies, not the government and have been around for at least 50 years (I know from my own experience that fact). I would much rather see a student score well on that test than on the Arkansas test that is designed and controlled by the state, where the political leaders have a vested interest in the scores.
The table comparing the NRT scores and the CRT scores was to show the lack of correlation.
Post a Comment
<< Home