The problem with a best-fit approach

Scattergraph
Here’s a handy graph, just ready for a best fit line to be drawn through it

It is not often that I read a 50-page government-commissioned document and get wildly excited. But that is what happened when I read the “Final Report on the Commission on Assessment without levels” (September 2015).

I found pages 12-18 particularly inspiring as the commission explains the rationale behind why levels are bad and ‘not-levels’ are good.

This statement was pertinent:

Levels also used a ‘best fit’ model, which meant that a pupil could have serious gaps in their knowledge and understanding, but still be placed within the level.

And this one:

Levels were used to measure both end of phase achievement and lesson-by-lesson formative progress, but they had not been designed to fulfil the latter purpose, with the result that formative assessment was often distorted.

Just as in yesterday’s post, when I tried to explain a character flaw that had held back my practice, here was a practical example of something I was doing day-to-day which was not actually any good.

As a key stage 2 teacher I had experienced this over many years. Children came up to my phase ostensibly as ‘a level 2 child’, but the best fit approach meant that while they could add a pair of 2-digit numbers with the support of a 1-100 number square, they had no idea of any other strategies for doing the same thing, nor really any sense of the the size of the numbers they were dealing with, nor the purpose of the equals sign.

And likewise I must have been doing the same to my secondary colleagues for years: children from key stage 2 going up to key stage 3 being able to scrape enough marks to get a level 4 in a SATs test, but not with the necessary number or problem solving skills to really go any further. In fact at my school, analysis of SATS shows me that number skills have always been high, but problem solving really low – the children have been able to scrape through with good routine number skills, but lack the fluency to really excel in maths.

What is even more insidious about the ‘best-fit’ approach is what it does to ability groups within a class, particularly low ability children. But that’s tomorrow’s topic.

Image courtesry of: https://upload.wikimedia.org/wikipedia/commons/a/a2/Scatterplot_r%3D.24.png

 

Remember that you are dust, and to dust you shall return.

3 thoughts on “The problem with a best-fit approach”

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.