>> Also I have never worked out why exam results are "normalised" That is grade level
>> requirements are altered to meet a % of passes. This is no basis for improvement,
>> merely dumbing down the whole process.
Two ways of marking.
In one the examiner aims to the hurdles at the same height year/year. In that case it's possible that, like in athletics with those doing 100m in under 10 secs or a mile in under three minutes, numbers clearing the hurdles can increase over time. More kids get an A* in 2013 than did in 2003
Other argument says cohort/cohort kids should be roughly same. You aim to set the exam hurdles at same height but even if timings improve you still give A* to top 5% and A to next 15%.
Which is right?
|