The Calculation of the Final Grade
The table on this screen shows how the the final grades for the students
are calculated. The final grades are a weighted sum of up to five components.
- The teacher's grade for their submitted work. This is optional and will be
used if the teacher actually assesses the student's work. If the student submits
more than one piece of work the "best" grade is used. Here, best
means the piece of work with the highest weighted combination of teacher's
grade and peer grade...
- The average of the peer grades for their submitted work. Again if the student
submits more than one piece of work the "best" grade is used. The
peer grade can optionally
include the teacher's grade. This grade would be included if the number of
peer gradings is very low or it is thought that the peer gradings are suspect
either because of bias (usually on the high side) or for not being reliable.
If included the teacher's grade is treated in the same way as a peer grade in
the calculation of the average.
- The student's bias in grading peer work. This is measure of whether the
student grades work either too high or too low. It is not an absolute measure
as it is based on the difference between the student's grade and the peer
averages for each of the submissions they assessed. In general this component
should NOT be given a high weighting.
- The student's reliability in grading peer work. This is a measure on how well
a students grades follow the peer average for the pieces of work they
assessed. The measure discounts the student bias and averages the absolute
differences between their grades and the peer average grades. In theory if
the students gives high marks for good pieces of work and low marks for poor
pieces of work their reliability will be high. If it is suspected that the students in
general are poor assessors then the teacher's grades should be included into
the ppeer averages, this should make the reliability values more meaningful.
- The average grade given by the teacher for the student's assessments.
This includes both the preliminary assessments made by the student on the
example pieces of work and any grading the teacher makes on the asessments
produced during the peer assessement phase of the assignment. In general this
component is probably more important than both the Bias and Reliability
components and thus, if available, should be weighted higher.
These five components can be weighted as deemed appropriate for the
assignment. For example the teacher's grade might be weighted strongly
if the peer grading part of the assignment is only considered a minor part
of the assignment as a whole. Alternatively, if the teacher only grades a few
of the submissions these grades can be ignored by giving them a zero weighting.
If the assignment is all about the students as judges and the providing of feedback
then first two components may be set to zero (or low) and the students'
grading abilities will determine the final grades.
Note that this screen is used iteratively and the final grades are not normally
made available to the students until the final phase of the assignment. Once the
the teacher is happy with the final grades and their weightings then they can
be made available to the students.