Completing the Anonymity Feature in Discussions Redesign

stimme
Community Coach
Community Coach

Recent posts in other Community groups this week drew my attention to the anonymous posting feature available only in the redesign. I suggest that there are at least four changes needed to consider this feature completely & properly implemented. 

  1. Anonymous discussions are an inherent new feature of Discussions/Announcements Redesign. It should be an optional feature at the account level, because many or most institutions do not want to allow anonymous posts in Canvas. While it is not an optional feature, institutions who do not want anonymous posts have reason to de-anonymize posts. Working around anonymity features is not ethical, but administrators may (correctly or not) deem the greater risk to be with true anonymity in Canvas.
  2. The Discussion Topics API enacts anonymity by pretending objects don't exist. This is not great, but it's better than ignoring anonymity. Ideally the API would have a field like the one in the GUI indicating whether the topic is fully, partially, or in no way anonymous. The API response for the topic as well as the entries should exclude the identifying fields for anonymous entries in the same way that the GUI does.
  3. Canvas Data 2's discussion topic and entry data sets have no indication of the anonymity level for topics or entries, and the IDs of users who author anonymous posts are currently included. User IDs should not be included in CD2 where they are anonymized in the GUI. At the very least, fields should be added to indicate which entries are anonymous and the level of anonymity for the topic so that reports based on the data can be anonymized to the same extent as the GUI.
  4. The Course Access Report & New Analytics currently show participation counts by students in fully or partially anonymous discussions. For conformity with the GUI and API's attempts to enforce anonymity, these participations should not be shown in the reports that instructors and admins can access.