Differentiating differences

Exam paperFollowing Michael Gove’s latest proposals on changes to GCSEs, Warwick Mansell recently wondered about the difference between tiering and extension papers:

There would be a single paper, Mr Gove confirmed last week, because tiered papers provided a “cap on ambition” by giving foundation paper pupils the chance only of a C grade. He did, though, muddy the waters by suggesting that “extension papers” might also be available to challenge higher achievers, leaving me, and others I think, wondering how this system differed from tiering.

So, just how does tiering differ from a system of ‘extension’ papers? And how do those systems differ from the old two-level system of CSEs and O-levels?

Same idea, different approach

CSE/O-levels
O-levels were aimed at the most able students, the top 20 per cent or so. CSEs were introduced in 1965 as a school leaving certificate, so that those who did not sit O-levels could still leave school with a qualification.
GCSE tiers
Some GCSEs are split into two ‘tiers’– a higher tier, in which candidates can achieve grades A* to E, and a foundation tier, where they can attain grades C to G. Higher tier papers assume candidates have grasped material covered in the lower tier.
Extension papers
All candidates take a general paper, with higher grades only accessible by also taking an extension paper. Candidates are not assumed to have mastered the content of the general paper, and are awarded the higher of the two grades they achieve.

All three systems noted above are ways of differentiating students of different ability, hence the likely confusion, but the implications of each system are quite far reaching.

‘Filtering’ students in this way is believed to be necessary in assessment as, if you don’t differentiate between groups, you run two major risks. For the weakest candidates, achievement could be based on ‘random success with half understood ideas’1 rather than evidence of a coherent set of skills. For the most able, achievement may come as a result of mastering the easier content rather than the most challenging.

Such was the experience of the O-level: the weakest drilled in concepts they failed to conquer, the most able drilled in the simplest concepts they had long since grasped (this is explored in the 1982 Cockcroft report, Mathematics Counts).

Tiering splits the assessment so that the most able are assumed to have grasped the easier content, and the weakest are not put through the dispiriting experience of being unable to answer questions on concepts they haven’t understood.

Unfortunately there are two side effects of tiering. The first is low pass marks. If you assume that a certain amount of content has been mastered, you can focus on challenging questions; failure to answer these questions still implies, however, that the candidate has mastered easier content and is still worthy of a reasonable grade. Low pass marks can sit uneasily with the public.

The second side effect of tiering is that choosing to enter a student for a higher or lower tier can be a gamble. Underestimating ability and opting for the foundation tier may cap a student’s achievement at a grade C, when they were capable of more. But overestimating ability and choosing the higher tier risks a student attaining so few marks that they fail to gain a key grade. A great deal of the responsibility in a tiered system rests with teachers choosing the right tier for their students2. The weight of responsibility has increased since the key overlapping grade C has become known as a ‘pass’ and has high stakes for schools in their performance measures.

The side-effects of tiering are, however, tiny in comparison to those of the old O-level and Certificate of Secondary Education (CSE) system. With a tiered GCSE you can receive the same qualification at the same grade, regardless of the tier you enter. A grade C in a GCSE gives you access to further study and employment in a way that a level 1 CSE never did (students’ GCSE certificates don’t state which tier they entered, removing the stigma that CSE students may have experienced).

Secondly, tiering is only used in around of half of GCSEs, with the rest relying on providing appropriate content for all levels of ability (differentiating by candidates’ performance after taking a common exam, rather than preemptively filtering to different exam papers). And in those subjects where it is still used, the incidence of candidates being capped – being entered inappropriately for the lower tier – is tiny, no more than 1 or 2 per cent (see Wheadon & Beguin, 2010). Of course, within schools students have long been split into sets by ability, so any ‘cap’ on aspiration comes earlier than the exam itself. It is possible that tiering has encouraged more schools to divide their students in this way, but the practice has been in place for far longer than tiering has been in existence.

Extending old ideas?

So how could new general and extension papers differ from tiering? Some GCSEs used a mixture of general and extension papers and tiering in the early days of the late 80s and 90s, so we actually know a great deal about how the systems differ. Under the general and extension paper system, all candidates had to take the general paper, but access to the highest grades required that a candidate also took the extension paper. Following extensive trials, it was decided that candidates who took the extension paper would be awarded the higher of their two grades3.

Two differences from tiering are clear. Firstly, those candidates taking the extension paper are not assumed to have mastered the content of the general paper. Secondly, candidates unsure of their ability do not have to make the difficult decision over which tier they enter. In the model used in the 1990s, there was no risk attached to the least able entering the extension paper; if they performed poorly on it they would receive the grade they had achieved on the general paper anyway.

 

While the general and extension paper system allowed the decision regarding which examination papers a candidate would take to be left open until a late stage, there were significant disadvantages. Firstly, it was felt that a great proportion of candidates were wasting their time answering questions in the general paper on content they could safely have been assumed to have grasped. This was a waste of their time, and a costly exercise in unnecessary marking. In NEAB GCSE French in 1997, between 65 and 85 per cent of candidates (depending on the combination of options candidates selected) entered the optional extension paper, and between 21 and 50 per cent scored 0 marks2.

Secondly, there was an inevitable drift to everyone entering for both the general and extension paper, at which point the distinction between a two level and a two paper exam becomes redundant (see graph).

So there are differences between tiering and a general and extension paper: both are preferable to the discredited O-level and CSE system. While no assessment system is ever perfect, this is an opportunity to consider what we have learned over the last 25 years to ensure that we have the best possible assessment system for 2015.

Chris Wheadon

References

  1. Secondary Examinations Council (1985). Differentiated Assessment in GCSE, London.
  2. Baird, J., Fearnley, A., Fowles, D., Jones, B., Morfidi, E. & White, D. (2001). Tiering in the GCSE: A study undertaken by AQA on behalf of the Joint Council for General Qualifications. Joint Council for General Qualifications.
  3. Good, F.J., and Cresswell, M.J. (1988). Grading the GCSE. London: Secondary Examinations Council.
  4. Cockroft, W. H. (1982). Mathematics counts: Report of the committee of inquiry into the teaching of mathematics in schools. London: Her Majesty’s Stationery Office.

Share this page