STATEWIDE SAT RESULTS - WHICH HIGH SCHOOLS ARE EXCEEDING EXPECTATIONS?
The Idaho Board of Education released statewide SAT results last week. Idaho has provided funding for several years for administration of the SAT to 11th graders on "SAT School Day", which occurs annually in April. Each Idaho student is required to take a "college readiness" exam as a prerequisite for graduation. Most students take the SAT on the "school day" administration date. Some students may have already taken an exam, and others prefer to take the ACT and so do not sit for the "school day" exam.
This makes it difficult to compare high school SAT results in a year-over-year fashion because differing percentages of the junior class take the exam. So, for example, 85% of juniors may take the SAT "school day" administration in one year, but in the next year only 75% may participate.
For the 2017 comparison, we included only high schools that met the following criteria:
- Over 80% of juniors participated in the "school day" exam (we used April enrollment figures from the State Department of Education to come up with a participation percentage).
- Over 70 students participated in the exam (even with this criterion, performance can vary widely from year to year for schools at the low end of the participation continuum).
We compared high schools using a scattergram in which we charted free and reduced lunch percentages on the "x" axis, and on the "y "axis we charted the mean score on the SAT. We looked at the average composite score, and scores for Evidence-Based Reading and Writing (ERW) and Math. We use free and reduced lunch percentages as a variable because the SAT (and other norm-referenced assessments) are highly correlated with poverty.
What we did not consider was the percentage of Limited English Proficient (LEP) students at the high school. This is important to note, because a few high schools have high percentages of LEP students. As you may know, Borah High School has the largest percentage of LEP students in the state among sizable high schools. Because they are still learning the English language, and because the SAT assesses some sophisticated concepts which require high-level reading skills, most LEP students struggle with the SAT. This artificially deflates average scores at those high schools. Conversely, most northern Idaho high schools have few, if any, LEP students.
Several other large high schools have significant percentages of LEP students - Nampa - 7%, Declo (Cassia Cty) - 6.6%, Caldwell - 6.6%, and Centennial (West Ada) - 6.4%,
Composite Score Performance
As you may know, we really like scattergrams, because they allow us to look at data and consider it in light of the effect of another variable. We also typically add a trendline to give us an indication of how schools are doing compared to the state average and overall trends. Since the SAT is highly correlated with poverty, you see a negative trendline - in general, higher poverty schools do less well on the SAT.
Traditionally, Boise, Moscow, Timberline, Sandpoint, Century, and Madison have performed well on the SAT, and this year was no exception, though Boise's performance was especially impressive in 2017. The surprises in composite performance are likely Capital, which performed well above expectations, and Idaho Virtual Academy. which does well in ERW, and not as well in math. The two Vallivue high schools, Vallivue and Ridgevue, performed well for their demographic, and though Caldwell's raw average score is below most, the high school is an outlier demographically, and actually did very well on the SAT considering its free and reduced population.
Note that Borah and Declo, schools with large LEP populations, performed above expectations even as most of their LEP students were included in testing.
Evidence-Based Reading and Writing
On the ERW subtest, the landscape is very similar to what the composite chart shows. As you can see, Idaho Virtual Academy is one of the top performers in ERW, considering the online school's demographics.
The SAT math subtest drives the "college readiness" percentage calculated by the College Board (parent company of the SAT), because scores are quite a bit lower than on the ERW subtest, and the cutoff score for "readiness" is higher.
On the math subtest, the usual high performers are joined by a few schools that exceed expectations: Coeur d'Alene, Lakeland, Kimberly, Shelley, Emmett, and Burley. Caldwell and Boise exceeded expectations by a wide margin on the math subtest, as well.
It's interesting to look at these data and think about what schools like Century, Capital, and Twin Falls are doing in Math and ERW that other schools which perform less well could learn from. Since the College Board provides an item analysis and releases the test annually, a statewide analysis of skills would likely reveal that performance gaps exist because particular concepts are not learned as well by students in some high schools. What a valuable learning tool that would be, if the State Department chose to use the information to assist districts and teachers.