Posted in Education Politics

The Tribune didn't publish my OpEd, so here it is:

Granite School District get D Grade in Academics

Mark Twain said that “there are lies, ****ed lies, and statistics.” With the release of the 2006-07 school year Annual Yearly Progress results, the truth of this aphorism is brought home in light of the “spin” placed on its results by Granite District. Granite is relying on the complexity of the results to pull the wool over the public’s eyes.

In an email sent to various patrons, Granite makes the statement, “If less than 95% of a group takes one test, the entire school does not make adequate yearly progress.” What Granite omits is that this requirement only applies to groups with 40 or more students. They also leave out the fact that this requirement is included because a school could cheat by sending its poor performing students home on test day. Out of 41 schools that failed AYP, only two failed because of participation: Granite and Hunter Highs. Both schools failed whole school participation, not based on a single subgroup.

The other 39 schools failed to meet academic standards.

Granite is quick to point out that failure of any subgroup within a school will fail the school, but they neglect to point out that each school has four shots at passing each subgroup. For example, the target for math in elementary schools is 71% of students proficient. If the test shows that more than 71% are proficient, that group passes; if a sub-group is close, but not quite above 71%, then the group passes if it falls within the “confidence interval” or statistical margin of error. For example, only 64% of the “Limited English Proficient” students at Cottonwood Elementary were proficient in math, but they were statistically “close enough” that they “passed.”

If the subgroup can’t make the target within the confidence interval, they may pass if they can show progress toward the proficiency target from the previous year according the “safe harbor”10% rule. Using Cottonwood as an example again, only 53% of the Disabled group scored as proficient (goal of 77%), but since they had improved from 48% in 2005-06, this group passed.  

At Copper Hills Elementary, three groups passed based on this Safe Harbor rule by scoring within the confidence interval.  The Hispanic subgroup scored 59% proficient in language arts compared to 60% last year, but this was “close enough” to a 10% improvement. LEP students “improved” from 59% to 58%, and Students with Disabilities “improved” from 38% to 34%. In math, the LEP students went from 57% to 58% (hardly 10%) and disabled students “improved” from 49% to 40%.

When we look at a failing school like Hillsdale, which had two “No” boxes, we see that only 6 out of 10 subgroups have enough students to matter; two groups (Asian and Pacific Islander) passed in the confidence interval on Language Arts; Caucasian, Hispanic, LEP,  Low SES, and Disabled made Safe Harbor in the confidence interval; in fact, no group with 40 students passed the proficiency target. Hillsdale failed on a whole school basis because the “confidence interval” for the whole school would not let a 59% proficiency masquerade as 77%, nor let it pretend to be a 10% improvement over last year’s 64%. This school received only two “No” boxes: Whole School in Language Arts and Hispanic in Math, but only one of the “Yes” boxes was based on the subgroup actually making its target: 74% of Caucasian students proficient in Math, (goal of 71%). Of the remaining 17 academic checks, six passed by the safe harbor rule, with five of those in the confidence interval, seven passed the academic test in the confidence interval, and 4 groups had populations so low they weren’t counted. According to the Granite email, 21 schools have more “No” boxes than Hillsdale, seven shared the 2 No box status, and only 13 did better.

When Granite’s scores are compared to the state averages, whether by whole school or by subgroups, Granite scores below the average in every group but two. Graded on a “curve,” this is a ‘D’ grade for Granite.