- Despite the movement toward authentic assessment, most states in this country still use multiple choice, standardized tests for student placement and to judge the effectiveness of public schools.
- Many children first encounter standardized tests in the 1st grade, and they continue taking them through early adulthood.
- Most four year colleges use SAT or ACT scores as part of their admission criteria and give little indication of changing in the foreseeable future.
- Students from lower socioeconomic levels perform less well on standardized tests than do other students, threatening their access to higher education and many professions.
- All of the above are true.
- None of the above is true. Performance based tests are fast replacing multiple choice, standardized tests.
As these statements suggest, the use of standardized, multiple choice tests is a complicated issue. One problem is that of equity: Some groups of students tend to perform better on these tests than do others. African American, Native American, and Hispanic students; girls; children from lower socioeconomic levels; and children with particular learning styles generally perform less well than do middle class, Euro American boys, and Asian American students.
This equity issue has broad implications for two reasons. First, public officials and school administrators often use these scores as a basis for important decisions about school funding and access. Second, nationally normed, psychometrically sound instruments are used pervasively in the United States, in part because they are easy to administer and score and the cost per student is low. This is true even in states like Vermont, where portfolio assessment is gaining ground. Given political realities, standardized tests aren't going to go away any time soon.
Some educators maintain that because these tests don't fit the educational philosophy underlying authentic student performance, we should simply ignore them. Taking this high road may be philosophically comfortable, but it could harm children who have little choice about participating in standardized tests. Further, for most children, taking these tests is an authentic experience as in real, not imaginary. Students will face these tests over and over again in the course of their lives.
This brings us to our position. We believe that we must find ways to use standardized tests and the skills upon which they depend to engage students in authentic teaching and learning. Only by doing so can we mitigate the inequities of standardized testing. As one African American graduate student put it:It doesn't matter whether the tests measure knowledge or not. When children of color can't do well on them, they can't get into good colleges. They can't get good jobs. We need to know how to help children do better on these tests so the tests won't act as barriers to educational and economic opportunity.
No Surprises?
One possible approach occurred to us two years ago after the annual publication of test scores in our local newspaper. Sure enough, there were few surprises. Nearly all the schools composed primarily of middle and upper middle class kids delighted in scores that were quite acceptable. The other schools including a school that was home to most of the town's homeless children, a school with the highest English as a second language population, and an elementary school we'll call the "Jane Doe" school did not do so well.
In some of these schools there was talk of changing the curriculum to more closely approximate the test format. Some teachers developed workshops to reteach the skills on which the children scored poorly. Others maintained that the tests were irrelevant and should be ignored. Meanwhile, administrators were fielding questions from concerned parents and central office officials. Principals felt pressured, teachers felt frustrated, children were upset, and many parents were simply confused.
What was surprising about the results at Jane Doe was that (1) students were predominantly white and middle class and (2) the school's test scores did not reflect the quality and quantity of the learning going on there. The teachers at Jane Doe had developed a coherent, educationally sound curriculum based on learning as a developmental process. The students, parents, and teachers were involved in well planned, authentic learning experiences.
Significantly, no one asked whether these students understood tests as a particular kind of thinking, problem solving, and literacy. The underlying assumption was that whatever the tests were measuring, kids in some schools just didn't have enough of it. But the question that we found most compelling was this: Did these children and children at the other low scoring schools understand how to demonstrate what they knew under timed, multiple choice conditions?
Putting Performance in Context
Following conversations with teachers at Jane Doe, we began to explore whether they could enrich their curriculums by using standardized test formats to get students to use new problem solving strategies and approaches to learning. We also began to consider how teaching for authentic student performance might be related to equity issues and test taking.
We ended up collaborating with teachers at the Jane Doe School. As we worked with them, we learned about the school's educational environment and the students' test taking knowledge and problem solving skills. We also came to understand the children's emotional needs and the teachers' needs, concerns, and insights.
Based on this information, we designed a series of workshops to help teachers and students explore what various types of test questions really ask of the test taker. We also developed workshop materials for the students and extensive scripts for the teachers to use in coaching the students. Unlike other test preparation programs, ours required engaging students in improving their problem solving skills by examining correct answers.
During one workshop, for example, a child had been attempting to answer the question, "The freight train carried apples and oranges in its ______." The answer was boxcar. The child asked plaintively, "How did you know the answer was boxcar?" A classmate then related how he associated information from the Boxcar children's book with this question. That response helped other children explain that answering a question with information from outside the school is a legitimate strategy.
Situations like this were repeated many times as students helped one another discover new problem solving strategies. The problems students brought up helped to clarify why they chose certain answers reasons that often were very different from what teachers assumed. For example, one question listed four months November, September, April, and December and asked students to select the one month that didn't fit in this group. One child chose December, the correct response. But when asked to explain her choice, she did not give the "right" reason December alone has 31 days. Instead, she said that December is when Christmas comes.
When children chose correct answers for nonstandard reasons, teachers and students could explore other explanations. Such incidents moved the workshops beyond simple test preparation and helped teachers use assessment as a catalyst for learning.
Authentically Good Performance
The workshops were successful beyond anyone's expectations. When the school's test scores were published the following spring, Jane Doe was the only school in the district whose scores improved significantly. And the improvement was dramatic: many children's scores had risen by as much as 20 or 30 percentile points between 1994 and 1995. In reading, the median percentile rank rose from 56 to 75; in math, it went from 33 to 57; and in language, from 35 to 66.
The curriculums in these content areas had not changed from one year to the next. Neither had the school's developmental approach to learning. What did change was the children's knowledge of how to take a standardized test.
Even more important than the scores were the children's and teachers' reactions to the workshops and their test taking experiences. Children who had reported having few test taking strategies, as well as considerable anxiety and little confidence in their ability to perform, became confident mentors for younger children. They told their fellow students how to remain calm, how to avoid getting stuck, how to eliminate choices, and how to approach particular types of questions efficiently. They explained why they didn't need to worry about the way other kids were rustling the pages of their test booklets. And they reassured them that it was okay to not know everything.
In general, the children spoke of feeling happy about having performed a challenging task well. The teachers, many of whom took a dim view of standardized tests, reported that the workshops helped them to better understand their students as learners and problem solvers. By paying attention to the tests as one type of literacy and performance experience, they were better able to teach for authentic student performance and better able to help their students understand their ability to successfully negotiate the standardized test.
Assessment as a Learning Experience
At Jane Doe we collaborated further with the teachers and students as they prepared for a different test. We also went on to coach teachers at an elementary school with a number of homeless students, a rural elementary school, and an elementary school with very good test scores. The average scores for three schools improved substantially, and results are not yet in for the fourth.
What we taught the children to do through the workshops was simple but fundamentally different from many test preparation programs. We had no new test strategies to offer. Rather, what we offered was a process that combined practice in a number of things: attempting unfamiliar formats, receiving immediate feedback, exploring reactions and feelings, and generating and actively discussing different problem solving approaches. Through this teacher guided, but student centered, process, children learned how to show what they knew.
