Rethinking Student Assessment: #1
May 15, 2009, 3:27 am
by Susan Horowitz, NYC Public School Teacher
In a 2008 memo, Joel Klein wrote that many teachers felt it would be useful “to better understand how teachers’ efforts are influencing student progress” (personal communication, October 1,2008). As a result, the Department of Education (DOE) introduced Teacher Data Reports, a new system designed to track teachers’ strengths and weaknesses based on the results of their students’ standardized test scores. In the new system, teachers would be assessed based on the level of growth their students demonstrated from one year to the next. However, growth analysis would be determined solely on the difference, from one testing year to the next, between the students’ raw scores on the English Language Arts (ELA) State Exam.
These tests, while important tools to measure students against norm standards, should not be the only tool used to determine teacher and school efficiency. Klein himself acknowledged this concern when he stated “reports like these can never perfectly represent an individual teacher’s contribution to student learning” (personal communication, October 1,2008). This statement is especially true because the format of the tests is different from grade level to grade level. For example, the third and fifth grade tests assess editing skills and reading and listening comprehension, while the fourth grade test assesses these three skills along with writing proficiency. It is therefore impossible to compare testing data from one year to the next. In the school at which I teach, the students’ scores consistently drop from third to fourth grade and then climb back up in fifth grade. This is not because of a lack of effective teachers in fourth grade, but because the skill sets required to pass the fourth grade test are very different from the skill sets required to pass the other two tests.
Moreover, not only is the data gathered by these means an inaccurate assessment of student growth, but these tests also misjudge student ability. In my experience, because the anticipation leading up to these tests is so psychologically burdensome, many students become overwhelmed and do not perform to the best of their ability. Conversely, some students who do not understand concepts can still accurately answer multiple choice question and so score higher than their critical thinking skills warrants.
These tests should be considered as only one limited form of data. Yet, the DOE (through the Learning Support Organization) trains teachers to use the results from these tests when planning instruction. By making these tests so important, the DOE forces teachers to teach to the test and abandon genuine instruction that is based on triangulated data. We suggest that , instead of using one test to determine successes and failures, the DOE train teachers to collect and analyze multiple forms of data and create portfolios for each student that reflect student progress. Before such a system can be designed, teachers and the DOE must agree on what constitutes legitimate data and how this data can be used to drive instruction.
We do not know a flawless way to track teacher efficacy or student data, but the following blog entries will be an investigation into how teachers and schools can use assessment data to track student progress. We hope that through member suggestions and responses we can open a dialogue about the future of assessment.
Tips for Parents (21)
May 15, 2009, 3:27 am
by Susan Horowitz, NYC Public School Teacher
In a 2008 memo, Joel Klein wrote that many teachers felt it would be useful “to better understand how teachers’ efforts are influencing student progress” (personal communication, October 1,2008). As a result, the Department of Education (DOE) introduced Teacher Data Reports, a new system designed to track teachers’ strengths and weaknesses based on the results of their students’ standardized test scores. In the new system, teachers would be assessed based on the level of growth their students demonstrated from one year to the next. However, growth analysis would be determined solely on the difference, from one testing year to the next, between the students’ raw scores on the English Language Arts (ELA) State Exam.
These tests, while important tools to measure students against norm standards, should not be the only tool used to determine teacher and school efficiency. Klein himself acknowledged this concern when he stated “reports like these can never perfectly represent an individual teacher’s contribution to student learning” (personal communication, October 1,2008). This statement is especially true because the format of the tests is different from grade level to grade level. For example, the third and fifth grade tests assess editing skills and reading and listening comprehension, while the fourth grade test assesses these three skills along with writing proficiency. It is therefore impossible to compare testing data from one year to the next. In the school at which I teach, the students’ scores consistently drop from third to fourth grade and then climb back up in fifth grade. This is not because of a lack of effective teachers in fourth grade, but because the skill sets required to pass the fourth grade test are very different from the skill sets required to pass the other two tests.
Moreover, not only is the data gathered by these means an inaccurate assessment of student growth, but these tests also misjudge student ability. In my experience, because the anticipation leading up to these tests is so psychologically burdensome, many students become overwhelmed and do not perform to the best of their ability. Conversely, some students who do not understand concepts can still accurately answer multiple choice question and so score higher than their critical thinking skills warrants.
These tests should be considered as only one limited form of data. Yet, the DOE (through the Learning Support Organization) trains teachers to use the results from these tests when planning instruction. By making these tests so important, the DOE forces teachers to teach to the test and abandon genuine instruction that is based on triangulated data. We suggest that , instead of using one test to determine successes and failures, the DOE train teachers to collect and analyze multiple forms of data and create portfolios for each student that reflect student progress. Before such a system can be designed, teachers and the DOE must agree on what constitutes legitimate data and how this data can be used to drive instruction.
We do not know a flawless way to track teacher efficacy or student data, but the following blog entries will be an investigation into how teachers and schools can use assessment data to track student progress. We hope that through member suggestions and responses we can open a dialogue about the future of assessment.