Sunday, March 8, 2015

Giving more information and control to instructors

Today we made two changes to CrowdGrader that we hope will give instructors better control of the grades, and students more assurance that their work is properly evaluated.

Average grade, and grade difference, for every submission

First, the list of all submissions now includes both the average grade received by a submission, and the delta between the highest and lowest grades.  You can sort on either column, thus having quick access to the students who got the lowest average grade, and those who got the most widely differing grades.

This is how this looks on a sample assignment, and on a real assignment (with student identities hidden for privacy):

Average grade and grade difference
for a sample assignment.
Average grade and grade difference
for a real assignment.

Instructors can now also grade submissions

The other change, which goes hand in hand with the previous one, is that instructors can now grade submissions, if they wish.  The option is offered when viewing the details of a submission, and here is how it looks: 


If instructors select Review and grade this submission, they are led to a page where they enter a review, and a grade, for the submission, using an interface that is the same as the one used by regular reviewers.  When the assignment grades are computed, CrowdGrader gives precedence to the teacher grade over the student-assigned grades. 

The result: more information, and more control

These two changes together give more information and more control to instructors.  Instructors can easily see which submissions received the most widely different (and thus, more unreliable) grades, and they can easily re-grade a subset of the submissions, ensuring that they receive proper grades. 

Furthermore, in case where students give widely exaggerated grades, the instructors can easily intervene and assign proper grades to the submissions. 

Since instructor-assigned grades take precedence over student-assigned grades, this also creates a powerful incentive for students to be fair in their grading: students whose grade differs markedly from the instructor-assigned grade will receive a low review grade