5 ms·
But how do you measure progress, then? If you are already on the backfoot, don't you want some kind of concrete measurement of how you improve? Admitted, this
by choeger 5y ago
But how do you measure progress, then? If you are already on the backfoot, don't you want some kind of concrete measurement of how you improve?
Admitted, this would require to use grades slightly differently (not as aggregation towards an end result) and it would require the teachers to repeat tests multiple times. But shouldn't that be how learning is organized?
- etataetaet 5y agoI think most people can measure progress without grades! For me at least, getting a 90% on a test is usually the same as a 100%. That 10% difference is almost always just silly math errors or simple misunderstanding and not a fundamental issue.
- choeger 5y agoI don't think that most people can do that. In fact, I think only the ones that already understand a topic can realistically give an estimate of the percentage of their understanding. Someone who "got" 25% or 45% of some material will usually not see their progress.
- etataetaet 5y agoFair enough! If you don't understand the material it makes sense you wouldn't be able to properly judge how well you're doing. Thanks for the thoughtful response
- MikeTheGreat 5y agoI'm going to do the obligatory thing & cite the Dunning-Kruger Effect here :) Also, from personal experience as someone who's taught computer programming (US college 1st & 2nd year courses - listed as "computer science" but really mostly computer programming): I've seen a _lot_ of people tell me some variation on "My program won't compile, I'm totally stuck, but those are just (implied minor) details - I really do understand what's going on!". Bonus points for when their code indicates that they've missed huge concepts at both the language level ("What do you mean I can't do arrayName = arrayName + 1 in Java?") and/or algorithmic level ("The variable defined by the outer loop of the bubble sort was never used so clearly that loop was useless so I removed it"). Ok, I'll stop ranting :) My point being that you may be able to measure your own progress but not everyone is. Plus, certification / credentializing is one of the services that education provides to the the rest of society.
- etataetaet 5y agoYeah I see what you're saying! However, for me specifically, I know the mistakes were not fundamental ( Just gotta trust me on this one xD ). The other comment pointed out that if you fully understand the material its a lot easier to judge your own understanding. Now that someone said that it seems really obvious, but It wasn't something I ever thought about! Thanks for the insight
- solarmist 5y agoI think something like a moving average would be much better. Historical data without too much emphasis on the past. Then you can also adjust the way you grade.
- afiori 5y agoGrades are averaged mostly because it is an obvious operation that sort of make sense, but the core issue to me is that what a grade _is_, how it is computed, and how it is then used are all nebulous concepts beyond "higher is better". unless you are in Germany
- jimbokun 5y agoShe said she gave lots of detailed feedback on all the assignments, and opportunities for revision.
- shalmanese 5y agoIt’s not clear from the article but it seems like students do get a final course grade which is done by students suggesting what grade they feel they deserve and the instructor rarely modifying it if they feel it’s inaccurate. They mentioned they adjust grades up as much as down.