Hi all,

We did this for five years on the first unit of our PgCert, which was 100% assessed by self and peer assessment of participation in blog-based activities (responses to reading, conversation tasks, reflection on practice, etc). Participants were put into groups of four and required to post, and read and respond to one another's posts, on a fortnightly basis. It worked really well.

It wasn't pass/fail, participants used our standard postgraduate marking matrix. I had them 'translate' the generic band descriptors though, through the activity that I've pictured (I still do this with them, for other tasks); a blank matrix where they generate band descriptors specific to the task at hand. Having generated the task-specific band descriptors, they then try them out through an open formative peer and self assessment activity. They consider their current participation/attainment for said activity and lay down coloured cards on the matrix in accordance with which band they think they're in (each person uses a different colour). The self-assessor goes first, followed by the peer assessors. They then have an open discussion, explaining their choices, and try to reach a consensus.

I found this to be a nice way to prepare them for the summative peer and self assessment, which – although it was online and anonymous – generated sensitive, detailed and constructive feedback. I have other data around variation in peer assessment grades, difference between self and peer grading, etc, which tell interesting stories :) Happy to discuss further.

NB Our cohort size was around 75 at this time.

Lindsay