The Test Doesn't Tell All
|High-stakes tests fail to catch the subtleties of incremental improvement that inform teachers' day-to-day curricular and instructional decisions.|
In this time of high-stakes testing, we rarely get a true picture of
what children know and need to know in order to meet high standards.
Test scores do not tell all. By investing significant resources in
improving our children's test-taking skills and making teachers better
at preparing students for these tests, we limit real education. The
limitations are especially evident in underresourced schools, where the
need for maximizing time and opportunity for learning is the greatest.
High-stakes tests fail to catch the subtleties of incremental
improvement that inform teachers' day-to-day curricular and
instructional decisions. The tests also offer no guidance on what types
of professional development teachers need in order to build capacity
for student achievement.
To capture these subtleties and design effective professional development for teachers, we need to figure out what is going on in classrooms where kids are learning successfully. If we are serious about accountability, we need to teach teachers how to assess their own work and its impact on their students. Teachers are in the best position to collect and use this data to shape their practice and inform educational decision-making. Teachers also are able to powerfully demonstrate the impact of a curricular or instructional mandate on their students.
Over the past two years, fellows of the National Teacher Policy Institute, a professional community of full-time K-12 classroom teachers working in urban, suburban, and rural settings, have been conducting action research in their classrooms and schools to help identify issues affecting student achievement and how these relate to standards. Their studies cover such topics as adolescent literacy, parent-teacher communication, portfolio assessment, cross-age peer tutoring, and teacher retention. They fall into four categories: teacher preparation and new-teacher induction, ongoing professional growth, teacher networks, and teacher leadership in school change."
Teachers are in the best position to inform educational decision-making.
In these studies, the fellows are crafting compelling designs of what needs to happen if we are truly to support students so that they reach the standards.
A middle school teacher in New York City's Community District 2, whose class has an Asian-American majority, provides one example. The teacher saw that meeting the "speaking and listening" performance standard was a great challenge for her students. She also recognized that the parents of her students needed to be aware of the standards to give support to their children in achieving them. She was able to identify a cultural mismatch between the speaking and listening standard and how her students had been raised. A bilingual school aide described the mismatch in this way:
For thousands of years, Chinese haven't talked during meals. Just put your mouth on the food. My daughter said one time, "Mommy, American families talk during lunch."
In fact, the New York state standard on speaking and listening requires students to participate in group conversations and meet several criteria, including voicing opinions that may be in conflict with those of others in a group. In the classroom described above, however, many students believed that vocal disagreement with others constituted disrespectful behavior. They have been conditioned to be quiet. One student wrote: "A lot of our parents tell us not to argue. That's not respectful. If I disagree with someone's answer, they might get their feelings hurt. It's like arguing or telling them that they're wrong."
The reality was that parents were unaware of the standards because they had never been explained to them in a language they could understand. After this teacher spoke to the parents (during parent-teacher conferences aided by the school's sole, part-time translator) about why discussions in school are important and offered suggestions to help their children become more vocal, student participation in group conversation immediately increased. Students' ability to meet the criteria for speaking continued to improve dramatically throughout the year.
|By participating in a teacher network, new teachers were able to demonstrate improved professional practice and were better able to implement standards-based instruction.|
In another example, a Los Angeles teacher did research on the impact of collaboration within a teacher network on new teachers' understanding and use of state language-arts standards in their classroom practice. More than half of the teachers in this researcher's primary school are noncredentialed and have taught for less than three years—a common situation in urban schools throughout the country. By participating in a teacher network called the Early Literacy Club, these new teachers were able to demonstrate improved professional practice and were better able to implement standards-based instruction in their classrooms. Teachers who did not participate in the teacher network left teaching at the end of the year. Those who participated stayed in the classroom.
A further demonstration of the value of teacher input in helping students reach high standards is provided by a New York City high school teacher who was part of a team that has instituted portfolio assessment as an alternative to the state regents' examinations. Teachers in every subject area worked together to compare their curricula and the state standards; they revised student assignments and grading criteria to better align with the standards. Over time, the result of their alternative to the exams was that student work rose to meet standards-based assessment criteria. In her study of seniors' work with portfolios, the teacher discovered that the portfolio approach provided a much richer profile of students' knowledge and skills than that provided by a standardized test. She notes that the best ways to "determine if students can apply the scientific method or conduct research is to look at students' actual experiments and research projects."
Teacher research can be a powerful tool for realizing change felt far beyond the classroom.
These examples of teacher research, as well as dozens of others from New York City; Los Angeles; Santa Barbara County, Calif.; Fairfax County, Va.; and the state of Illinois, have been published in a new book, What Matters Most: Improving Student Achievement. This Rockefeller-funded publication connects the action-research findings of National Teacher Policy Institute teachers with the recommendations of the National Commission on Teaching & America's Future, and demonstrates the power that teacher research can have on shaping teaching practice and student-learning outcomes. In her preface, Linda Darling-Hammond, the executive director of the commission, writes, "We at the commission through our partnership with NTPI plan to use the lessons gained to engage more teachers in our policy research and discourse ... and encourage all engaged in similar efforts to do the same."
Through this work, we have seen how teacher research can be a powerful tool for realizing change felt far beyond the classroom. Teachers who ask questions about their practice and try to determine the extent of their students' learning stimulate professional discourse about teaching and learning and move the conversation outward to include parents, communities, and other stakeholders.
Educational policy that is informed by teachers' voices becomes significantly more responsive to local needs for resources, professional development, and other supports—enabling all students to better meet high standards.
Ellen Meyers is the vice president of programs and communications for The Teachers Network in New York City. Frances O'Connell Rust is an associate professor of education at New York University. The books What Matters Most: Improving Student Achievement and A Guidebook for Connecting Policy to Practice in Improving Student Achievement are available through The Teachers Network, [email protected].
Vol. 19, Issue 38, Pages 34, 37Published in Print: May 31, 2000, as The Test Doesn't Tell All