The perils of streaming

I managed the streams at my previous school for a number of years and significantly reduced the amount of issues raised by teachers.  It was a difficult process to manage as the stream was one size fits all - move in maths, you move in all the MESH subjects.  To get around this required three of the four to suggest a stream or one humanities/english, one math/science to be moved.  At it's best, students worked with teachers over at least a term to get promotion.  By the end the move was valued and deserved.  Streaming was done twice a year for all students and reviewed each term for students that had entered the school during the last semester.

The problem with streaming is that it lacks a research basis for putting it in place in an average school.  In a school where the level of teaching required is 4+ years within the same room, there is a basis for it as it is difficult to teach students that far apart.  In my new school, after removing a small number that require IEPs, it appears that the remaining students fall within a 2-3 year syllabus bracket in each class.

I am struggling with the existing streaming model I am faced with as I come to grips with the benefits and deficits of it.  Three streams have been devised with top, middle, bottom courses.  Top, middle, bottom have two classes in each, also streamed (creating six fully streamed classes) - the best students in each stream in one class and the remainder in another.  The streams appear to align with NAPLAN (with some obvious exceptions being rectified) indicating that for the most part they have been ranked correctly. One teacher in each stream sets all assessments.  All assessments are written from scratch in each stream.

Issues appear when the performance of the streams are analysed.  Out of the 60 students in the top stream, a significant proportion of the second class are achieving less than 50%.  The gap between the mean score in the top class and the second class in each stream increases significantly as students progress through school. My top class outperformed the second top class and were highly motivated and enthused, but students in the second class universally wanted to be in my class (not theirs) but were unlikely to be promoted as they significantly performed lower and demonstrated lower levels of motivation.

Having a highly motivated class may be seen as positive in isolation, but having half the students undermotivated in the top course made me question current practices.  In each stream, one teacher sets the assessment with little communication with the other teacher.  This leads to the second class being given an assessment that is pitched at the wrong level for the second class resulting in lower levels of engagement.  Thus high performing (top pathway but in the second class) students are left believing that they are not solid high performing mathematics students.  It also has the potential for friction between the two teachers (at the level of the assessment set and/or the students of the underperforming second class (blaming the teacher for the underperformance).

As a trial I rebalanced the year 8 top stream to have two classes of equal ability and challenged each class to raise their performance.  To date I have seen little difference in student performance in my rebalanced (lower ability than before the rebalancing) class and hope to see the other class rise to the challenge. We now have two classes of equal ability, we should achieve median test scores approximately the same. With both classes being exposed to high performing students they should have an increased potential for success.

The second issue that I see is that streaming in Year 7 occurs too early.  This separates kids into haves and have nots very early in their high school career, prior to them engaging with specialist mathematics teachers and gaining a love of mathematics.  The issues with this is seen with undermotivated middle tier classes in later years and students that struggle to fully engage with Mathematics as they lack proper role models, particularly boys who generally develop later than girls.

A quick analysis today indicated that there is a clear difference in teaching with respect to results in Year 7, as four classes with similar composition (same NAPLAN mean) with the most at risk removed into a remediation class, performed significantly differently on the same assessment, with one class clearly outperforming the others.  There are some factors that can be attributed to class differences (the spread of ability was different in two of the classes), but if the differences in programme and pedagogy can be identified then improvement can be made in subsequent delivery across all classes.  This was also after one assessment, this may be a strength of one teacher in one topic and may vary in future topics or in students adjusting to a different style of teaching.

The third issue identified is with the performance of upper school classes - there is always room for improvement.  Although small Methods (10-15 students) and Specialist (5-7 students) classes consistently produce 55+ course scores, it is not as consistent in Applications courses (30-40 students) with between one third and one half of students achieving a 55+ course score. In a school of so many students in Band 9 and 10 students in NAPLAN9,  there appears potential for higher numbers of Methods students out of the 60 students typically in the top stream.  To produce a higher number of Methods students requires aligning the Year 10 course with Methods rather than the current programme of 10/10A aligned to the Specialist course and only producing 6-7 Specialist students.  We need to make Mathematics a subject of choice for students at the school and fall in their top two subjects ranked by ATAR course score.  This means the Year 10 teacher would need to extend potential Specialist students through differentiation in class (rather than through the general programme) and assess predominantly on the year 10 (not 10A) curriculum.  If the current course (with a 10A focus) produced more students in Methods with the current programme, it could be seen as workable, but they are not choosing Methods, underperforming in it with many withdrawing from the course in Semester 1, Year 11 (thanks SCSA for an overloaded Semester 1 course) - typically seeing the Applications course as an easier path through Mathematics; resulting in few students using Mathematics as their first or second score in Year 12.

The fourth issue comes with creating unteachable classes.  If students that are disengaged are lumped into a class together, there is the risk of creating a difficult class to engage. Rather than working with difficult students, the first option appears to be to use the streaming process to remove them to another class - reasoning that needs to be constantly challenged.  My prediction when setting up streams at my prior school was that middle classes would prove the most difficult to teach never eventuated there, but I can see it at the new school.  The middle classes appear to lack a clear pathway to keep them motivated, lack positive role models and this may be a good area to explore for further improvement.

Food for thought going forward.  Now to get teachers to see the streams holistically and derive positive changes that can be measured for the benefit of students.