I’ve been thinking.
Following a comment on my last post about statistical significance I’m putting my hands up and saying I don’t know enough about the maths behind “significance”… there are much better equipped people out there that I am sure can inform me. All I know is that in the section where Ian Stockford thanks people for taking part he mentions:
- 3865 students. I have always “bandied” the figure of 500,000 year 9 students around, when in fact there are more than this as I’ve shown below (based on the Jan 2014 data from the government – the Jan 15 data didn’t load for me!!) so is 3865 out of, let’s call it 522,925 (so I’m excluding special schools and independents) this equates to 0.739%. If we look at this as % of ALL students that are currently in year 9 it equates to 0.672%.
- The Ofqual blog also mentions “two classes of students” … so at most this is 64, so I’ll be generous and call it 70 students. Again, excluding special schools and independent, this equates to 0.0133% and of ALL students this equates to 0.0121%.
To be honest I really don’t give a “monkeys” about “significance” … I just know that each and every one of these students are real people. This is their future and I’m not convinced I would want to be making changes that have seen this much controversy by looking at data that involves such a small number of students. Why am I looking at year 9’s? They are the cohort – real students that will be taking the first GCSEs in Maths. They are the students that are subject to the massive shift in content down the key stages. They are the students that we are teaching the new GCSE despite not yet having SAMs to work with. They are the students against which we will be measured for the following year. They are the students that will be leaving school with GCSEs with both “Levels” and Grades. They are the students that will have to explain to employers what this means … enough validation?
Some would argue that the strand that the “two classes” were being used for is quite insignificant. Regardless of how big or small, this is a research programme and there is no way that the whole ability range can have been involved: what if both these classes were “top sets” or had a particularly great teacher? There are just too many variables involved for my liking – at least not for a cynical person like me. There is no getting away from it, the results ARE going to be skewed.
I genuinely hope that the complete data set is released along with the schools that took part. For this to be genuinely representative of the school population I would hope that the school demographics reflect those of all the schools in England. And that Ladies & Gents is how I’ll be spending weeks of my summer holidays if they do release them!!
SO MY PARTING NOTE IS … 574,545
