GothamSchools — daily independent reporting on NYC public schools

Not only is New York State proposing a “proficiency plus” accountability model that will take both absolute proficiency and student growth towards proficiency on state reading and math tests into consideration, it is also looking at creating a “growth for all” system to reward schools who move already proficient students to even higher levels of proficiency, and, perhaps, penalize schools where proficient students do not make additional gains. This part of the accountability system would not require federal approval, presenters stressed at last week’s public forum on the model.

According to the presentation by Ira Schwartz of the state education department, many ideas are still on the table for where to set the bar for growth, how to compare students and schools, and what positive or negative incentives schools could expect under the new system.

First, the state must determine what schools should strive for in educating students who already test proficient. Is it enough that students continue to test above the proficiency cutoff, or must they show one year’s growth or more when scale scores are compared? The question echoes this summer’s debate over whether to emphasize the “proficiency gap” or the “achievement gap” in looking at student performance.

Asked whether some parents and educators might choose to improve the teaching of non-tested subjects such as art, music, and physical education rather than devoting more resources to helping proficient students score even higher, Schwartz responded that the Regents had specifically asked for a way to hold schools accountable for the growth of all students.

Next, the state must decide how to compare schools. New York City’s Progress Reports got a nod for their peer group comparison, one possible method for pushing schools with large numbers of proficient students to constantly improve their students’ scores. One alternative mentioned would compare the scores of students with “similar growth histories,” as Colorado’s new growth model does by calculating a student’s “growth percentile” compared to students with similar test scores in previous years; another would calculate “statistical expectations for what the student would learn with a typical teacher/school.”

Finally, the state will create rewards and consequences based on the new measures of student growth. Though details were scarce, it seems that incentives could be as simple as changing the rules for recognizing schools as high performing or rapidly improving, or as complex as “differentiated consequences” which would allow the state to classify schools and provide more or less intensive interventions based on the specific weaknesses of the school.