I have to express concern about showing groups on the same graph as we show the whole cohort. For courses where we have large cohorts I do not think that we can show the whole cohort and the group on the same graph with the same axes.

For an example let's assume a course with 1,000 students.

And the graph broken in 20 ranges each of 5%.

Then we might easily imagine that >100 students will be in one of the ranges.

Now our tutor groups only have 20 students.

Taking the same proportions as above that translates into 2 students in a range.

And showing 2 students on the same axis as we show >100 will render the former as almost 'not there'.

So I don't want this left as it is.

Which raises the question of what we should do. The points are labelled only so that I can refer back to them below.

[option 1] Let's be clear that the main purpose of this graph is for course teams to get a pictorial view of how the assessment has performed. That is we are mainly interested in 'Course staff' viewing the graph of the whole cohort.

[option 2] Once we have this it is quite OK, and I think helpful, to show that graph to tutors of groups..

If we can easily show tutors where their students sit then that is worthwhile but as I've pointed out we can't do it using the same scales.

[option 3a] So either there will need to be two y scales, one for the cohort and one for the group ( I don't know that I like this).

[option 3b] Or we should arrange a graph for the group below that for the cohort and positioned so that shapes of the graphs can be compared.

[option 4] We should not try to show all groups on one graph. A course population of 1,000 translates into 50 tutor groups. And we have much bigger courses.

I wish to see [option 1]. This is a requirement.

[option 2] follows readily from [option 1].

[option 3a/b] would probably be good. But I'd like to see it with reasonable numbers.

[option 4] should be dismissed. Course staff can already see how tutor groups perform through the Gradebook. If we wish to reproduce this in the quiz reports we should use the same approach i.e. allow course staff to select a group. But this is not a requirement

I'm wondering if the graph should be for all quiz attempts or depending on scoring the highest graded / first / last attempt ??