Statistic of the Month: Trends in the Performance of the Top Performers on PISA 2003-PISA 2012


By Jennifer Craw

The last time the PISA assessment emphasized mathematics was ten years ago.  With the most recent PISA assessment also emphasizing mathematics, we can now take a closer look, not only at how countries’ and economies’ performances compare to each other, but how their students’ performance improves, declines or has stayed steady overtime.  This month’s statistic takes a look at the ten top performing countries on PISA 2012 and compares their performance over the past four administrations of the PISA mathematics exam, looking specifically at how each country or economy has increased or decreased their proportions of low and high performing students as well as the change in the proportion of resilient students (students from low socio-economic backgrounds who score at high levels)  based on PISA results.

Note: Each of the following charts contains trend data provided by the OECD for PISA 2012.  The OECD assures users that when comparing trends in mathematics, reading and science on PISA, only those countries with valid data to compare between assessments are included. (PISA 2012 Results: Volume I pg 53.)


In addition to the mean scores PISA generates for each country or economy, PISA ranks students into six proficiency levels according to the skills they demonstrate in answering the assessment questions.  The OECD consideres a proficency level of 2 as the baseline for proficiency in mathematics.  The above chart demonstrates the rate at which each country has either increased or decreased the proportion of its students who are not considered even baseline proficient, or at Level 2, in mathematics using the PISA rubric.  The percent of low performers in the US has first increased slightly then decreased slightly, but compared to the PISA assessment in 2003, the US is right where it started with just above 25 percent of students not meeting the proficiency standard.  That means that one in four US 15-year-olds that took the PISA test was not able to do much more than follow obvious mathematical instructions, let alone interpret data from multiple sources or recongnize math contexts that are not obvious.  Meanwhile, in Shanghai and Singapore, where already a much smaller proportion of students failed to meet proficiency standards, the percent of students scoring below level two has dropped significantly even since the previous administration of PISA in 2009.

The most dramatic change in the above chart is in Poland, where the proportion of low-performing students has declined sharply, even since the 2009 administration of PISA.  In 2003, 22 percent of Polish students failed to score at level 2 or higher in mathematics, but in 2012 that number had decreased to just over 14 percent.  Poland is one of three new countries whose average scores across the three subject areas measured by PISA have placed them on CIEB’s list of Top Performers; look for a profile of the Polish education system and how they have managed to improve so dramatically in such a short time period on our website soon.


This next chart shows the change in the proportion of students who score at the highest proficiency levels on PISA mathematics, levels five and six.  Students who score at these proficiency levels can develop and work with models for complex situations, work strategically using broad well-developed thinking and reasoning skills, and can formulate and communicate their interpretations and reasoning.  Once again, the performance of US students flat lined between 2003 to 2012, with even a small decrease with just over 10 percent of students scoring at level five or above in 2003 and 8.8 percent meeting that standard in 2012.  Poland, which also started out at around 10 percent in 2003 has managed to increase the number of students scoring at the top levels of PISA by almost 7 percent, with 16.7 percent of its students meeting proficiency level 5 or better.  Similarly, Hong Kong, Japan, Estonia, Taiwan and Singapore have increased their proportion of top-performing students.  Shanghai, already out-performing all other countries and economies on PISA 2009, increased its proportion of top-performing students by 5 percent on PISA 2012.

One final chart shows the change since 2003 in proportion of resilient students in the top performing countries.  These are students who come from low-income families who score at the highest levels of PISA proficiency.  We previously looked at the performance of resilient students on PISA 2012 in our December 2013 issue of Top of the Class.

Resilient trend data

Two of the top-performers, Canada and Finland, show an increase in their proportion of low-performing students on PISA and a decrease in top-performing students despite maintaining relatively low levels of low-performers and a good proportion of high-performing students overall.  Yet neither of these countries is ignoring the data.  For a more in-depth look at Finland’s reaction to the most recent PISA results, see our interview with Pasi Sahlberg in last month’s Top of the Class newsletter.  Canada’s decrease in resilient students, as well as its decrease in high-performing students has motivated national and provincial education policy leaders to look closely at Canada’s mathematics curriculum and teaching resources.  In Ontario, Education Minister Liz Sandals has committed CAN$4 million (USD$3.69 million) to help improve math skills of teachers.  The funds will be used to subsidize enrollment in math courses, particularly for their primary and junior school teachers.  Read more at CTV News.

While the US has not declined in its performance on PISA since 2003, it is clear most top performing countries have improved, leaving the US further behind.  Along with the lessons from top-performers like Poland and Shanghai who have improved performance for all students, a consideration of how Canada and Finland respond to their decline could also help the US in its approach to mitigating this trend.