I’m going to try to make this a no waffle post … so let’s get on with it … Part III

I want to look at the fact that for strand 1 of the research modular papers were used to compare the so-called “recent” questions. Basically for those of you that don’t know modular exams are no longer modular in the sense of being sat at various times during KS4 but are now all sat as terminal exams by some students and many people now call them “unitised”

In this post I am arguing that using modular papers wasn’t the best move for this strand as they aren’t really representative of what the vast majority of students sit. Basically what I’m trying to say is that using modular papers doesn’t give a good comparison to what is going to be a linear specification in the new GCSE. It’s like comparing apples and oranges … yes they’re both fruits but totally different types. Some of the exam boards actually “examine” and limit the more elementary parts of the specification within certain papers and so questions will almost certainly look easier to someone trying to assess difficulty. It just doesn’t make sense are there are perfectly good (and more recent) LINEAR papers available.

That aside – my biggest gripe is the fact that modular papers are sat by such a small percentage of the total entries and there are different outcomes from each board and each type of exam when compared with linear. I have used the publicly available, provisional data (from Summer 2014) published by JCQ and each of the exam boards showing grades and also entries for the two types of exam – this is also available historically (here you can find links to data from AQA 2014, OCR 2014, Pearson/ Edexcel 2014and the JCQ summary document is here-> GCSE Results 2014). I will give a health warning on this – there are still too many factors that affect some of these percentages for example OCR have such a small number entered for modular that their data may be skewed; I have only collated the data for 3 boards and the “missing” data may show a different picture; The tier of entry may also have an impact. This is just me being “quick and dirty” with the data.

Anyway, my “back of a fag packet” calculations show: 707,360 total entries by the 3 main boards (against a total from JCQ of 736,403 – bear in mind I’m only looking at linear & modular quals) of which, at least 634,961 were linear entries and 72,669 were unitised/modular entries. So given that less than 10% of the total entries were on modular/unitised papers why, oh why would they be used in the research project? Even if I’ve got my numbers slightly wrong … even if it was 20% (and IT ISNT!!) of the total entries it would in my opinion still be wrong. My concern stems from politicians and people in power reading the report on face value when they don’t understand the complexities of the different types of exams.

Looking at the data in more detail for you can see that there is a difference between the % of students that “passed” on the different models of papers and between each board – by “Pass” I mean gained a grade.

In the case of all three boards their unitised papers outperformed their linear papers – with AQA and OCR at the top – and when looking at the linear you can see that Edexcel now has the lowest pass rate with AQA and OCR almost identical, but both higher than Edexcel and the UK average.

At the top end looking at A and A* (note OCRs are probably skewed due to the small number entered for their modular/unitised exam) the modular specifications all have higher percentages of A*s and As than any linear specification. Interestingly all 3 linear specifications are below the JCQ UK averages for A and A*. Suggesting that modular really is easier!

As with all the stuff I publish … I ask you to form your own opinions and draw your own conclusions and I could go comparing and contrasting but below is the table of data and so you can do some number crunching yourself. I do know some kind soul will be along who may well have done more detailed work … if so drop a link in the comments below.

I’m still only going through notes from strand 1 … next time I’ll be looking at the language used within the report when making comments about each of the different boards. I’ll be back … off to have a life for a few hours. But before I go my last thought: I am genuinely wondering why these papers were used in the research project when our new GCSE is linear.