Yacapaca

Imagine having a detailed profile of the progression of every student, in every subject, not just by grade or even by topic, but right down to the level of concepts known, understood and evaluated.

Who would use it? I can see value to classroom teachers, department heads, senior teachers, parents and most of all, the students themselves.

How would you use it? Would you use it to plan lessons to reinforce weak concepts? Assign individual work? Make personalised interventions? Or something else?

But what would this cost to achieve? If students could be organised into tutor groups of about six, I think the tutors could remember that degree of detail about each student. Currently, classes of 30 are achieved with an overall student:teacher ratio of 20:1, so to achieve classes of 6 we’d need a ratio of 4:1 – five times more teachers than we have now. Hmm….

How do we make this realistically affordable? The answer comes in two parts:

1. gather the raw data passively, as a by-product of other activities. Don’t expend teachers’ or students’ time on gathering this data as a separate exercise.
2. store, collate, analyse and disseminate the information from a central database that is designed for just this function.

It became obvious to me two years ago that Yacapaca is well-placed to do this, but traditionally we had only recorded or reported grades (too big) or individual question responses (too small). So we have spent the last couple of years adding keywords to questions so that we can collate the answers into a meaningful concept profile of each student.

It is trickier than you may imagine. We have to persuade authors and teachers to enter keywords, and then we have to filter out all those keywords that make sense to the author and no-one else. There is still a lot of ‘noise’ in the system, but we are making good progress. There are a multiple hidden data-gathering algorithms within the system that help us to sort and prioritise the keywords.

We have also made great progress in using this data, again often invisibly to the user. The most visible example is the Results page (Assignments tab -> Results -> Results (beta)). This provides a key concept report for each student, for a given quiz. The Analyse popup does the same thing for the whole class.

It is the hidden uses that I am most proud of. In particular, Revision. This uses the keyword profile of each individual student to select the concepts that most need revision, and then selects questions – often new questions the student has not seen before – that match the keywords.

There is a long way to go before we have really learned to extract the maximum value from the data we gather. We are going in the right direction and I am happy to see that already the more adventurous teachers in our community are making use of the existing features to drive their lesson plans.

The next step is to  see what happens when this principle is applied consistently across a whole school. I shall be working with a few schools this year to at least move towards that, and I’ll report progress as I go.