Friday, 23 March 2012

Ofqual report on AQA French

Ofqual has just published monitoring reports on three new GCSEs, including AQA French. I wonder why they chose this examination. Was it connected with all the complaints they may have picked up from teachers, particularly about the marking of controlled assessments?

You can read the pdf document here:

http://www.ofqual.gov.uk/news-and-announcements/130/868

It isn't terribly long and there are some quite specific recommendations on how AQA could improve its examinations and advice to centres on controlled assessments. Here is an example of a comment on speaking CAs:

"Also for AQA French speaking, AQA’s requirement for candidates to cover a series of bullet points was judged to have constrained opportunities for natural conversations to take place for all candidates."

Er... yes. But then the format of the exam rewards candidates who pre-learn everything in detail, so what sensible teacher would encourage "natural conversations". Ofqual would have been better advised to point out that the format is just flawed.

Their advice to AQA?

 "For speaking controlled assessments, continue to reinforce the advice given to centres, in that best practice aims to achieve a natural conversation (from January 2012)"

Hmm... the mark scheme does not reward naturalness of conversation. To get 30/30 on a task candidates should memorise their best possible answers and avoid spontaneity.

Here is a comment on controlled assessment of writing:

"For AQA French writing, although AQA had instructed examiners to consider candidates’ work as a whole, in several cases examiners were crediting the number of characteristics – tenses, opinions and justifications – shown in the responses at the expense of the overall quality. This was particularly evident from the sample of work reviewed for mid-ability candidates, who in a number of cases were overly credited for the work that they had produced."

Interesting point, but then examiners can only work with the mark scheme they are given. My experience was that work of very good quality was not highly rewarded enough and certainly was graded lower than under the previous coursework format. There were a lot of very unhappy teachers this year.

One piece of advice Ofqual offer AQA:

"For writing controlled assessments, provide exemplar marked candidates’ work to include examples of tasks that successfully meet the criteria for assessment".

In fact, AQA already do this and there are ample examples of candidate exemplar work on the AQA site. This is not the problem. The problem is making sure that examiners are consistent and grades accurate. There is plenty of anecdotal evidence to show this was not the case last year.

They also suggest:

"For writing controlled assessments, provide further guidance to examiners so that they reward the overall quality of candidates’ work (from June 2012 marking period)."

Good idea.

Lastly, Ofqual found faults in the setting of reading comprehension papers (which they call, curiously, written papers). These included pictures which could cause confusion and inconsistencies between specimen papers and actual papers.

Their advice to AQA:

  • ensure that for future examination series (from 2013 onwards), images are made clear to all candidates and there is no possibility of confusion in questions
  • ensure that for future examination series (from 2013 onwards), as far as possible, each question in all its parts is appropriately demanding for the target candidates, including in this careful consideration of particular task types for questions aimed at key grades
  • ensure that for future examination series (from 2013 onwards), questions make clear to candidates the information required by the mark scheme
  • ensure that for future examination series (from 2013 onwards), written papers provide a varied range of text types.
Why didn't AQA get this right first time?

Perhaps they should do what the people at Asset Languages do and pre-test questions properly, rejecting ones which do not produce the right outcomes or which produce unnecessary confusion.

No comments:

Post a Comment