The Arkansas Department of Education noted in a July press release that Arkansas students are improving their performance on state Benchmark tests:

For the first time, more than 60 percent of Arkansas students at each grade level scored at or above proficient on both the mathematics and literacy Arkansas Augmented Benchmark Exams. The exams, administered in April 2009 to 209,000 third- through eighth- graders, are recognized nationally as some of the most rigorous in the country. . . .

“It’s a great day when a state can report that more than two-thirds of its students are achieving proficiency, especially when those results reflect continued academic growth at each grade level,” said Diana Julian, Ed.D., interim commissioner for the Arkansas Department of Education.

But the Arkansas Benchmark tests have come in for some criticism for allegedly lowering the proficiency standards in recent years. In other words, what if Arkansas students seemed more proficient only because the proficiency bar were lower?

Here are the tables correlating raw scores to proficiency for 2009; here are the same tables for 2006. As you can see from the following comparison tables I created, the raw scores needed to be proficient or advanced have dropped considerably in Math, especially in grades 7 and 8.

But does this really mean that the Benchmark tests are being scored differently? The other possibility is that the Benchmark tests have somehow become harder — so much harder that a raw score of 28 out of 80 points is now enough to mark an 8th grader as “Proficient” (compared to a raw score of 39 points in 2006). As can be seen in this Word document from the Arkansas Department of Education, there may some sophisticated mathematical calculations that underlie any changes to the raw scoring methodology, and the ADE may simply be resetting the proficiency standards after they were raised in 2005.

Nonetheless, if we look at NAEP exams (National Assessment of Educational Progress), Arkansas students aren’t doing nearly as well as the state Benchmark exams would suggest. In 2007, NAEP showed that a mere 28% of Arkansas 8th graders (and 41% of 4th graders) were proficient or advanced in math. That same year, Arkansas Benchmark tests purported to rank 48% of 8th graders (and 65% of 4th graders) as proficient or advanced in math. In other words, Arkansas has been claiming that 20% of Arkansas 8th graders and 24% of Arkansas 4th graders were proficient or advanced in math, even though those students couldn’t make the cutoff on more reliable national tests. This is at least an indication that Arkansas’ cutoff scores may be set too low.

Thus, it would be very useful if the Arkansas Department of Education would put out a short document describing in plain English why the raw score cutoffs are so much lower today than they were in 2006, and why the Benchmark cutoffs for proficient or advanced are so much lower than the national standards. Such a document would allay the fears and rumors that can otherwise circulate.