Page 9 - Priorities 7

Academics
Caution:
Test Scores
Ahead!
The standardized testing numbers — SATs, achievement tests and others — seem to dominate the education world. Often, we can mislead ourselves with test results. Let me share some examples.
School achievement scores are
often reported in the media in tables listing district and state average scores. Many similar tests, such as the SAT, are reported with nationwide scores as well. Consider this:
* At Woodside Priory, four or five high-scoring students can lift our school’s average. At a typically large high school, the impact of four or five students wouldn’t show. If we looked at a chart showing two or three years’ average test results, how many of us would assume that Priory must be improving and the big school not changing? (Or, if the situation was reversed, that Priory programs were failing?)
* In California, about 45 percent of high school grads take the SAT. In Kansas, the figure is nine percent - mostly the top students bound for elite colleges. How many of us would look at Kansas’ higher average score published in a table and assume that Kansas’s students are better educated? Or at least perform better on the SATs?
* If someone mentioned that he scored in the top quartile in California, and another student said she scored in the bottom quartile in Kansas, many of us would assume that the California student has the higher score. (Not necessarily so!)
* A student scoring in the mid-range for the Stanford Achievement Test might take the test again, get a few more right, and move ahead in the district, county or state ranking. Would you assume that a student whose first score was at the top of the rankings would move ahead the same amount if he/she got the same number of additionalquestionscorrectonthesecondtry? (It won’t happen.)
These kinds of misjudgments can be avoided by applying simple, common-sense filters we all know.
• When two groups of test scores are being compared, the results are meaningless if the groups are dissimilar. In my example, Kansas and California are dissimilar; another example of dissimilarity is populations with exceptionally high
or low numbers of limited-English- speaking students.
* Averages are strongly affected by the size of the group. In a big group, a change is hard to see; in a small group, a change may look bigger than it is.
* Percentile rankings are affected differently in the middle, where most scores fall, than at the tail ends of a bell curve.
It’s important to realize that national achievement tests are based on what is generally taught at a given grade. If a school teaches the same material at a different time, or teaches something else, students’ test scores will probably be low. That doesn’t mean students aren’t learning anything - only that the test didn’t cover what they learned.
The testing companies are not trying to mislead. They provide both schools and parents with a sound explanation of their testing procedures. They understand that their tests are an imperfect measure. It’s our responsibility to decide how much validity to give a set of scores.
Last spring, I met with a group of Priory administrators and Edward Haertel, a consultant from Stanford University, to discuss these high-stakes tests. We looked at what we as educators can learn from these test results, what may be misleading about them, and what other options we have. We discussed our role in helping students achieve the best scores of which they are capable. It is an important topic that we will continue to pursue.
Grades and standardized tests serve different purposes and give us two very different views. We need both. Together, they provide a balanced understanding.
But let me end on another note. This month, I am leaving the Woodside Priory School to join my husbandinTexas. AlthoughIamheartbrokentobe leaving my Priory family, I am looking forward to livinginthesamestateasmyhusband! Thepast three years have been a wonderful adventure, and I willcherishthememoriesItakewithme. Iplanto continue my ministry in Catholic education as the Academic Dean of St. Agnes Academy, an all-girls DominicanhighschoolinHouston. Thankstoallof you who have made my time at the Priory so unforgettable. Godbless.
9
Testing companies are not trying to mislead us. It’s our responsibility to decide how much validity to give a set of scores.
Deborah Whalen, Academic Dean