In the version we are using, 188.8.131.52, there is the following comment in the PETs algorithm class:
// @todo: if there's a non-submission, fake the scores using avg / q
// score = ( old-score / (total_marks_awarded_in_peeronly / num_members - 1) ) * peer_only_avg_per_question
Has anyone implemented and tested it to see if this resolves the peer only issue described below?
From: WebPA [mailto:[log in to unmask]] On Behalf Of Stephen P Vickers
Sent: 04 September 2014 17:29
To: [log in to unmask]
Subject: Re: Can you change the Scoring algorithms?
The PETS option was removed a while ago (c. 2 years) and I am not aware of it being put back yet. The comment inserted into the code reads:
> // @todo : There appear to be grading problems with the PETS algorithm in peer-only mode.
> // Until we have conducted a full investigation and nailed down what's
> happening // this algorithm is disabled.
Hopefully someone else can provide an update on the status of this algorithm.
On 04/09/2014 16:58, MURRAY M.R. wrote:
> Apologies for this "newbie" question - we are just trying to get our
> head around WebPA at Durham, being on day 2 of our production service
> There is a section in the Academic Tutor manual available at:
> in the 'Add/Edit Criteria for Your Form' part (just after the Split
> 100) paragraph that reads:
> Both criteria will work with either the WebPA Algorithm or with the
> PETS algorithm.
> One of my academic colleagues has interpreted this to mean that it is
> possible to select/switch algorithms, but I can't see any way of doing
> this via the interface, or any mention of this in the documentation.
> Can anyone shed some light on this matter?
> Thanks in advance,
> Dr Malcolm Murray FRGS FHEA CMALT PG Cert
> eLearning Manager - Computing and Information Services | Honorary
> Fellow - School of Education Durham University | South Road |
> Durham | DH1 3LE | UK | http://tinyurl.com/durltt |
> @learntechdurham <https://twitter.com/learntechdurham>