Are misspellings important to note in learning analytics?

[ggurdin]

1. It's possible to run spell check locally, more quickly than we can run correction on our backend.
2. Do they clutter analytics?

[wcjord]

Nice to have but not vital?

[isbell]

I don't know that misspellings are so critical; phone-based texting is pretty error prone (at least for me it is..) and a lot of minor mechanical issues might get flagged (e.g., accents and ~ for Spanish due to not having the right keyboard).

From a research perspective (and potentially/eventually for application), I wonder if detailed keystroke logging is possible. In research programs like Inputlog (https://www.inputlog.net/overview/) are used to investigate writing processes and increasingly this kind of tech is being used to generate learning analytics; it can tell you things about fluency and composing processes. Very fine-grained though, and may be too much to log/store?

[github-actions[bot]]

This issue is stale because it has been open for 120 days with no activity.