Monday, March 30, 2015

the achievement gap, testing, and accountability

In recent months, the Colorado legislature has repeatedly debated reforming Colorado's standardized testing system. One issue that has frequently come up is the importance of preserving equity in education through testing. For instance, in an editorial last week, the Denver Post wrote:
"In all the furor surrounding testing that has erupted at the statehouse, the equity issue sometimes gets lost. Yet, it's immensely important to be able to tell whether kids in poverty and minority children are making progress and that their instruction is effective."
Equity in education is a goal everyone seems to agree on. The disagreement is how to achieve this. I believe that Colorado's current standardized testing system doesn't achieve equity. So we can't talk about "preserving equity through testing" because that doesn't actually exist in the first place.

Here are the issues I see with (in)equity in Colorado's standardized testing system:

1. Computer-based testing, especially at lower grade levels, naturally benefits students students with more computer experience, which are usually students from higher income brackets.

In Colorado, computer-based testing is encouraged beginning in fourth grade. This is problematic because kids don't necessarily have the computer skills in elementary school to type answers. This form of testing also exacerbates inequality because students who live in poverty are much less likely to use computers at home than students from a middle-class or wealthy backgrounds, and thus wealthier students and wealthier schools will score better on the tests.

2. The current science and social studies testing unfairly favors large schools.

As we discovered last fall, the high school science and social studies testing targets all of science and social studies. This includes subjects like economics, which small schools do not offer. Naturally, students who haven't taken these course will score lower on the standardized tests.

The task force recommended eliminating science and social studies testing for high school seniors in January; thus far, the state legislature has not taken action.

3. Some schools can afford to purchase test-oriented textbooks, while others cannot.

Colorado students have not fared well on the new standardized tests thus far. Pearson, which makes the PARCC and CMAS tests, also makes textbooks which are aligned to the tests. (I rip on Pearson a lot, but the other standardized testing companies all do the same thing.) This matters because, as Temple University professor and data journalist Meredith Broussard wrote last July:
"[S]tandardized tests are not based on general knowledge. As I learned in the course of my investigation, they are based on specific knowledge contained in specific sets of books: the textbooks created by the test makers. ... Pearson came under fire last year for using a passage on a standardized test that was taken verbatim from a Pearson textbook."
Even overlooking the obvious issue of whether Pearson textbooks are actually best for students' education, it's easy to see how this system contributes to further educational inequity.

Textbooks aren't cheap: some schools can't afford them whatsoever. Some schools can afford enough for classroom copies but not for students to take home; some schools can afford to give every student a personal tablet and e-book copies of the textbooks.

Yet all those students take the same standardized test.

4. The timing of testing (currently March and either April or May for most Colorado students) doesn't account for the achievement gap. 

Proponents of Colorado's standardized testing system have repeatedly advocated standardized testing as a measure of student growth from year-to-year.  However, studies have repeatedly shown that the income-based achievement gap in the United States primarily occurs over the summer.

Take the following (very oversimplified) scenario:
Students A and B both score "100" on a standardized test in May.
Student A is of high socioeconomic status.  Over the summer, he/she participates in a few summer camps, visits the library every week, or takes part in other summer learning programs. Over the summer, Student A improves 20 standardized testing points; if an exam was given on the first day of school, Student A would have scored 120. 
Student B doesn't have access or exposure to the same educational resources as Student A. Over the summer, Student B loses 30 standardized testing points; if an exam was given on the first day of school, Student B would have scored 70. 
The following May, Student A and Student B once again partake in standardized testing. Student A scores a 130; Student B scores a 110.
According to Colorado's current standardized testing system, Student A progressed more than Student B. However, when accounting for the summer learning achievement gap, Student B actually made miraculous strides, whereas Student A hardly improved.

Of course, in Colorado we don't have the data to determine how students gained or lost educational experience over the summer.  Consequently, while the achievement gap undoubtably affects Colorado students and their test scores, the current testing system doesn't measure how.

So what does this all mean?

I think the idea that standardized testing contributes to inequity can be difficult for some people because we'd like to think that measuring student growth is simple and that the data collected actually means something. Unfortunately, this logic is highly problematic.

Under Colorado's current system, the data collected via standardized testing is flawed. So when this data is then used to inform everything from teacher salaries to classroom standards to school comparisons, it leads to bad policymaking.

Some of the debate over testing in Colorado has been linked to the Common Core standards Colorado adopted in 2010. To me, these issues are only loosely related.  Holding Colorado students to high standards isn't controversial; the only question is how best to do it.

I'd be remiss if I didn't at least try to provide a solution, so the following are a few suggestions for Colorado's standardized testing system:

-Change the timing of testing. Minimize the amount of class time spent on testing. One option: hold 1-2 hours of math and 1-2 hours of literacy testing in August when students return to school, and another 1-2 hours in each subject in April or May. This tracks student progress, accounts for the achievement gap, and is still way less time than Colorado students currently spend testing.

(My middle school, a charter school, did this and I always found it way more useful than the state-administered standardized tests.)

-Making testing feedback more readily available. Right now, schools, parents, and students don't get testing feedback for 4-6 months. That's too long.

-If social studies testing is really desired, use an "exit exam" format. If Colorado has set standards for what students are supposed to learn in high school government, why not test them on those standards once they have completed their high school government class?

-Make learning materials equally available to all Colorado schools. This requires no explanation.

-Make paper-and-pencil exams the norm, at least in elementary school. Initially, schools were required to use computers; the state eventually relaxed this requirement. Nonetheless, 85% of schools still used the computer option. I'm doubtful that comparing computer-based scores and paper-based scores is perfectly accurate.

I think there is a tendency to characterize the movement against Colorado's standardized testing as anti-accountability.  To me, that couldn't be further from the truth.  Accountability in education is incredibly important, but it goes both ways.  We can't go on holding our teachers and schools and students "accountable" while we take for granted that testing works. My hope is that those who actually have the power to influence Colorado's standardized testing system will reach this conclusion, sooner rather than later.

-Jessica Piper

Like usual, if you have questions, feel free to ask in the comments or hit me up on twitter (@jpiper303)

No comments:

Post a Comment