Here we are.  After a few posts describing the history, structure, and reforms of schools in New Zealand, I finally get to my first hypothesis regarding the falling PISA scores.  Just as a recap, here’s what the reading scores look like over time for New Zealand and comparable countries.  The PISA scores have been falling since the first PISA in 2000.

The PISA math scores also fall from 2000 to 2012.  I still need to look at the science scores.

Changing demographics

The elimination of school boundaries allowed parents to vote with their feet regarding schools to send their children to.  Schools that once had a mix of families from different SES backgrounds were changed by school choice.  Now, high SES families flocked to certain schools while low SES families were either left in other schools or actively chose to go to other schools to be around families more like them.  And, as stated before, many times SES background is linked strongly to academic achievement.  School choice didn’t change the teachers.  It didn’t change the school facilities.  School choice did, however, change the student populations.

At the same time of school choice, families started moving closer to their chosen schools.  This caused the neighborhoods around the schools with high SES student populations to have high housing prices.  High housing prices are only affordable by high SES families.  Then, school zones were reinstated.  Essentially, the re-establishment of school zones has locked attendance at high SES schools to just the families that can afford to live in the zone.  This isn’t very different than high SES schools in Illinois.  The difference, though, is that in Illinois most high SES areas have been high SES for quite some time.  In Auckland, the change happened, and still is happening, quickly.  The change happened just as the 15 year olds from the 2000 PISA entered the school system (1990ish).

How are students chosen to take the PISA test?

The sampling method in New Zealand is based on a random selection model.  This is the explanation from May, Cowles, and Lamy’s PISA 2012 New Zealand summary report (2013).

Schools and students are randomly selected to ensure the sample is representative of the New Zealand 15-year-old population. Schools are selected by the international consortium that carries out PISA based on the following characteristics: size, decile, location (urban or rural), authority (state or independent) and type (co-educational or single-sex). Students are selected randomly from all students in these schools within a specific age group (between 15 years 3 months and 16 years 2 months).

With the sampling method mentioned above, the international consortium randomly selects schools from different size groups, decile groups, locations, authority, and type.  Of those factors, decile rating (a rating the government establishes to rank schools by the SES backgrounds of enrolled students) changed the most due to school choice and neighborhood gentrification.  In fact, the decile changes started in the decade before the first PISA in 2000.

Remembering back to this post, we know that major reforms happened in New Zealand in 1989 which caused students to move around between schools and eventually for entire neighborhoods to go through gentrification and establishment of school zones once the major gentrification process was done.  Gentrification is still happening.

Hypothesis 1

Hypothesis 1 is very simple.  I believe sampling error is contributing to the falling scores.  A random school selection model is built upon some major assumptions.  The most significant assumption for the school selection for PISA is that the SES background diversity from families stays the same between each administration of the test.  As the years of school choice advance, the high SES families from multiple schools flock to just a few schools, creating a few higher decile schools with very little SES diversity among the students.  Remember that decile ratings are created by ranking the current schools.  So, as mid to high-SES schools become less SES diverse, they don’t necessarily become a higher decile rated school because they still are ranked with other schools.  When schools were more SES-diverse, the chances were more equal to selecting low-SES students or high-SES students.  Now, with very low SES diversity the chances are not as equal.  High decile schools have a very high chance of selecting a high SES student to participate in PISA.  Low decile schools have a very high chance of selecting a low SES student.  My hypothesis 1 states that using the decile system to select schools is not reliable since the very definition of the decile levels changes with each census due to school choice and gentrification of neighborhoods.  The decile system is increasing the chances of selecting low-SES students, leading to lower average PISA scores for New Zealand.  A better method might to to randomly select students from the entire population based on family SES.  This would ensure that an equal distribution of students from different SES households.

Hypothesis 2

Hypothesis 2 aligns with the same theme of school choice.  In this hypothesis, though, I look longitudinally from 1989 (the year of major school reform in NZ) to 2012.  Once 1990 rolled around, parents had full choice of schools to send their children to.  Schools began to understand the problems of school choice.  A few started hiring PR firms to keep and/or attract families.  As the years progressed, more schools started marketing themselves against the other schools.  Teachers stopped collaborating between schools because they felt pressure to do whatever was necessary to keep their students enrolled in their schools.  This led to extreme isolation of teachers.  Teachers still collaborated within their school, but not outside of their schools.  They still attended conferences because that was a national crowd.  But very few teachers collaborated with teachers from schools down the road.  Universities found a niche for providing PLD (professional learning and development) after a few years into the school choice reform.

Schools, and every  organization with shared leadership models, are slow to adapt.  So, while school choice started in 1989, the full effects of the problems did not appear for many years because schools still had not felt the effects of low teacher collaboration until years into the reform.

The first group of students to take the PISA (in 2000) entered the NZ school system around 1990.  Those students still had early educational environments that benefitted from pre-1989 teacher collaboration.  As the years progressed, pedagogy became stale due to low teacher collaboration.

Hypothesis 2 states that the PISA2000 students had the early parts of their education immersed in fresh pedagogy and as they moved closer to high school, the pedagogy slowly became stale.  The PISA2012 students never felt the effects of pre-1989 teacher collaboration because they entered the NZ school system around 1997.  Hypothesis 2 incorporates the idea that a major component to fresh and relevant pedagogy is open collaboration between teachers.

I have a third hypothesis brewing, but I’m not ready to write about it.  I need to observe more struggling schools before I can solidify my thoughts.  More to come on that at a later date.

References

May, S., Cowles, S., & Lamy, M.  (2013).  PISA2012 New Zealand summary report.  Ministry of Education.  Retrieved from https://www.educationcounts.govt.nz/__data/assets/pdf_file/0008/144872/1015_PISA-Summary_2012.pdf