Statistic of the Month: PISA for Schools – A Model for an Effective School Reporting System

By Jennifer Craw

In 2012 the Organisation for Economic Co-operation and Development (OECD) and America Achieves collaborated to launch a pilot of the OECD Test for Schools, a test based on PISA that allows participating high schools to compare their school level results with results from schools around the world.  Like PISA, the OECD Test for Schools measures 15-year-old students’ applied knowledge and competencies in reading, mathematics and science.  For the pilot, 126 schools from varying socio-economic and demographic backgrounds participated between April and October, including 105 schools in the United States, 18 schools from the United Kingdom and three Canadian high schools.  Each participating school received a summary of results, allowing that school to compare its performance with the performance of countries and regions from around the world on the PISA 2009 assessment.  This type of reporting gives individual secondary schools an opportunity to see their standing in a global context and to set targets for themselves beyond local and even national expectations.  Below we have described key sections of the results’ report for Herndon High School in Fairfax County, Virginia.  Students at Herndon High School, located in a middle-class suburb of Washington, DC, come from relatively advantaged socio-economic backgrounds, with 13 percent of students qualifying for federally subsidized school lunches.  This compares to the average U.S. high school where 40 percent of students receive free or reduced price lunches.  An example report for another U.S. high school, North Star Academy in New Jersey, which has a much different socio-economic profile that that of Herndon High School, is available on OECD’s website as well.

Each report provides school results, detailed information about what the students in each school know and can do in reading, math and science, information about the school’s learning environment compared internationally, a comparison of the school’s results to similar schools in their country and the school’s results compared to their international counterparts.

In the figure below, the distribution of students scoring at each proficiency level in Herndon High School (shown on the chart as “Your High School”) is compared to the distribution of students at each proficiency level in the United States overall for reading. While a slightly higher ratio of Herndon students scored at levels 3 and 4 than did students in the U.S. overall, the difference in students reaching proficiency level 2 (which the OECD considers a baseline) is not statistically significant between Herndon (88 percent) and the U.S. in general (83 percent).

StatChart1

In the figure below, the disciplinary climate in mathematics lessons at Herndon High School is shown to be significantly inferior to the climate in mathematics lessons at the schools with the highest performing students in mathematics in the United States.  StatChart2

The report also highlights the relationship between student background and learning outcomes.  In the figure below, Herndon High School’s performance on reading is compared to other schools in the United States. Circles in the gray vertical bar represent schools with similar socio-economic profiles to that of Herndon, and the size of each circle correlates to the size of that school.  Herndon student performance, represented by the red circle, is just below the average for what would be expected for a school with its socio-economic profile, resulting in no real statistical difference.  Corresponding charts for math and science show similar results, with Herndon students performing neither significantly better nor significantly worse than would be expected of schools with similar socio-economic profiles in the United States.  The report also points out that in PISA 2009, the socio-economic background of students has a higher impact on their performance in the United States than the average across OECD countries—a troubling trend that high schools in the United States ought to keep in mind as they track student scores and work to improve equity within their own schools.

StatChart3

Schools are able to benchmark their performance on the OECD Test for Schools internationally with PISA 2009 results from countries around the world.  In the following figure, the distribution of Herndon students at each proficiency level is compared with distributions for 12 countries that participated in PISA 2009.  Here Herndon students perform slightly better than U.S. students in general, and at similar levels to students in Germany, with just over 80 percent of students scoring at or above level two.  However, Herndon’s performance is well below the top performing nations on PISA 2009, with significantly more Herndon students failing to reach level two and significantly fewer students reaching level six than students in Shanghai-China, Finland, Korea, Singapore, Canada and Japan.

StatChart4

In the next figure, Herndon High School’s performance on Science is compared to that of the United States as a whole, Shanghai-China (the top performing education system on PISA 2009) and Mexico (the lowest performing OECD nation on PISA 2009).  While Herndon students perform above the top 10 percent of students in Mexico and above 50 percent of schools in the United States, Herndon students would be below the bottom 25 percent of students from schools in Shanghai-China.  This suggests that while Herndon High School is doing well nationally, it has a long way to go if it is to prepare students to compete with the best in the world.

StatChart5

Leaders from schools participating in the pilot have commented that the data from such a reporting system is very useful because it allows schools to see themselves in a national and an international context and to draw lessons from other top-performing education systems and schools.  Other education systems may want to consider 1) modeling their own school reporting system on this one, 2) augmenting their own existing system with these reports (perhaps every few years), or 3) integrating their own school reporting systems with this PISA option.  More information about the OECD for Schools Test can be found at http://www.oecd.org/pisa/pisa-basedtestforschools/.