Using data to compare teachers' performance - a note of caution!

This really bothers me.  On so many levels. 

Practically though, I think the following is really important to consider whether you're a classroom teacher, HoD or SLT.

In an age of marksheets, spreadsheets and tracking when number crunching is so quick and data is plentiful, it's all too easy to pop in a formula, press and button and compare your teachers. 

But for subjects where students are taught in sets it's just not that simple - and more importantly is it fair? With pay possibly linked to 'performance' I think there may be a real problem with this.

The 'Critchlow-Rogers Effect' research highlights the issues:

I apologise for the reproduction - I only have an old paper copy to hand. Full article can be read at the end of this post.

Is there a fairer way of comparing class teachers' data? 

 

Perhaps.

 

When comparing the 'performance' of teaching sets I've also looked at a percentage increase.

 

Here's a really basic example:

 

A student achieving a level 3 at KS2 who then achieved a Grade D (Approx level 6) would have made 3 levels progress but actually made 100% progress from their starting point.

A student with a level 5 at KS2 who then achieved a Grade B (Approx Level 7) would have also have made 3 levels of progress from KS2 but 60% progress from their starting point.

 

OK, so it's really simplistic example but it can be tweaked with decimal grades and grade boundaries.

 

Fairer?  Perhaps.

 

The important thing for me to remember is that when we download our results in August we should give real consideration to which buttons we press and what we do with the numbers that come out. 

 



 Full article: