In an Op Ed in C-J Friday, Sheldon H. Berman correctly opines that
- assessments ought to be diagnostic
- CATS is about accountability and is not diagnostic
- a proper test would tell us how students improve over time
- overreliance on the current test distracts from the larger educational mission
- parents aren't wild about the present testing circumstance
- CATS should move toward a growth model of assessment (a value added system perhaps?)
- limit the amount of time spent on assessment
- investment in the schools must include efforts toward "small class size, early childhood services, [and] additional instructional time..."
Then he gets something wrong. Berman suggests that a new CATS assessment should include what folks refer to as multiple measures, "authentic measures of assessing student performance such as portfolios, exhibitions and extended writing samples that are better able to reveal higher-order thinking."
Here's the problem.
The essential question of accountability is, "What will the public accept as trustworthy evidence that schools are performing up to expectations?" I have real doubt that the public will put enough stock in highly subjective measures. Portfolio assessment, exhibitions...etc., suffer from threats to interrater reliability. In a high-stakes assessment environment subjectivity is poison. If the assessment system was redesigned to assist teachers only - not for accountability - it would be a different matter.
Otherwise, Berman has the right idea. You don't fatten a calf by weighing it. But when growth data are consistently kept, smart beef producers can better monitor the progress of the herd and account for the effectiveness of their various efforts to improve the livestock.
The current education reform movement is built on the belief that setting high standards and holding educators and schools accountable will lead to improved student performance. I am one who holds that belief.
However, if our goal is to bring all students to a level of proficiency that reflects depth of understanding, the measures we use to assess progress must be of sufficient quality to give us more than a general sense of what students know. They have to be diagnostic tools, enabling educators to accurately assess individual student needs and curricular strengths and weaknesses.
There is a significant difference between testing systems used to monitor performance and systems that provide diagnostic information on individual students. Diagnostic tests are both more elaborate and more expensive than tests used to monitor because they have to provide a finer grain of analysis on the understanding individual students have of particular concepts.
For example, a high quality diagnostic test in math has to provide educators with detailed information on individual student understanding in such subcategories as number sense, computation skill, algebraic concepts, geometry, data analysis, measurement, etc. Diagnostic information also has to tell us how students improve over time, what growth in understanding they've made in these particular areas, and what value specific educational programs and services have added to that growth.
To date, the instruments we've used to assess school improvement have been relatively crude and have focused primarily on monitoring performance rather than providing diagnostic information. In addition, the consequences of our overuse of these instruments have detracted from our larger educational mission...
...First, we need to move toward a growth model of assessing performance, assessing each student's growth for each school year. The Kentucky Instructional Data System (KIDS) project, which will allow schools and districts to look at students' progress from year to year, is a modest start in this direction but more needs to be done to align tests across years so we can track individual performance over time.
Second, the Commonwealth Accountability Testing System (CATS) needs to become more useful as a diagnostic tool by releasing more test items along with detailed item analyses so we can precisely assess students' strengths and needs.
Third, we need to carefully weigh the amount of time schools spend on local and state testing with the need for high-quality instructional time, and then begin to limit the amount of time devoted to testing.
Finally, we need to further develop and give greater credibility to authentic measures of assessing student performance such as portfolios, exhibitions and extended writing samples that are better able to reveal higher-order thinking...
...With so much attention focused on accountability, we need to be careful that the measures we use don't become ends in themselves....
...Our primary goal, after all, is for students to become thoughtful, responsible adults who are active and productive members of our community.
As the NCLB reauthorization debate continues in Washington and as we deliberate on the next steps in Kentucky's accountability system, it is time to enhance the sophistication of our assessment system so schools can focus on depth of understanding while ensuring that students become well-rounded individuals.
There is an old saying that the cow doesn't get fatter by weighing it more often. If we want to realize real instructional gains from our assessment efforts, we need to focus our attention on better diagnostic tools that enable us to advance individual student performance. In addition, we need to think beyond assessment and seriously invest in the solutions the public and most educators believe will effectively move education reform and student performance forward -- small class size, early childhood services, additional instructional time, and increased funding for schools.