-
Notifications
You must be signed in to change notification settings - Fork 20
Description
Is your feature request related to a problem? Please describe.
Currently, self-reviews are considered as rubrics attached to an existing assignment. This means self-reviews are visible at all times once entered. This can influence designs in negative ways and hurts some learning objectives in OOD. and PDP. See the Additional context section for an explanation.
Describe the solution you'd like
An option to enable self-reviews to become visible at a certain time. When enabled, we can enter a date and time. At that date and time, the rubric items in the self-review become visible (by changing visibility from "never" to "always").
Describe alternatives you've considered
That means when I want to students to answer questions about their design and testing without the questions themselves influencing that design, I have to make sure the self-review is available for 48 hours instead of 24 and only present the rubric into the assignment after the late due date.
This workaround leaves self-reviews as empty but submittable, which allows a student to erroneously "complete the review" with no way for the instructor to re-enable it.
Another workaround would be to write the rubric, but make the visibility "never". Then at the time to release, we manually change the review items to "always". I am unsure if required but invisible checks prevent a student from completing a review without revealing the checks themselves.
There is no way presently for me to make a new assignment that downloads the latest submission of a previous assignment for each group and is staggered in the event of a late token use as another workaround and even if so, that is way too complex for something an additional option would easily expose.
Additional context
Most of this comes from how OOD and PDP use self-reviews: as a way to ask them if they tested and designed according to the highly detailed specifications given to them. This is a skill that both courses teach and test. Telling them the questions ahead of time influences their design for better AND worse.
For example, if you ask a question like "Does you card encode a position?", students will assume they need to. However, if the answer is no and the specification implies this as well by defining cards, then the review has influenced a design negatively. If you instead decide to only ask questions with a default answer of "yes", then you have all but told them a certain design, which means they do not practice how to get information from specifications and apply their own design choices but instead learn "the self-review questions tell me my design" which limits their usefulness and hurts them when they are tasked with making their own design on examinations or on co-op as software developers. Finally, this removes possible questions where a student can not only state they encoded a position, but have a very convincing argument as to why they did this. This freedom in design is key to both courses and their learning objectives.
I think the option to enable this allows for both this type of self-review and the self-review envisioned by Pawtograder to co-exist. I do know about the code walk feature, but only graders can fill out a code walk, which makes sense as a design. Finally, while OOD is to be replaced by PDI-2 next semester, PDP is still unchanged so they benefit from this option.