Final Grading Details

in Posts
Page content

Here is some information on where the final grades came from…

DE13 Self-Evaluations

Thanks to everyone who did this (almost everyone). They were pretty uniformly insightful and thoughtful. They gave me lots of ideas. I will fight the urge to say “but it was that way for a reason” - and instead say “OK, need to work on that in the future”. And there were some great specific things that I can do differently.

I did read them all carefully. Rather than “grading” them, I took lots of notes on what to think about when the class is offered again, and considered that everyone did a good job with it when we did other grading. The students who didn’t do it (very few): we assessed a penalty (5% of assignment), but in all cases, your final class grades didn’t change (you were far enough from a border that).

DE13 Grading

Expect to get an email (from Keaton) with DE13 grades. I will create short feedback notes that Keaton will send by email.

Basically, both of us looked at all the assignments individually, made notes and initial assessments. Then we met in person and in a marathon session went through each project and came up with a grade. And then we made another pass to make sure that the grades were appropriately consistent.

We did grade an an A/AB/B scale. (nothing is below a B). Within each grade bucket, we added extra categories (A-, AB+, AB-) to distinguish projects within a bucket.

We did all this with manual processes (spreadsheets) with two eyes on it to catch mistakes (doing the manual data join to share grades among partners took several tried to get right).

And because I am sure you’re curious… 1/3 of the projects got As. The distribution worked out reasonably - we did not tweak it. The mean was 3.5 (middle of the ABs). There were good (and bad) projects in all categories.

After the final grades were done, we did look back - borderline project grades were usually people whose ultimate grade wasn’t going to be negatively effected.

Interpreting the DE13 Email

All team members get the same grade. Not doing a self-eval was done after these scores.

Note that the notes are notes that we made for ourselves during grading. No promises that they are useful for anyone else. We are not including Keaton’s notes (as these really were meant for internal use).

You will get:

  • A grade (B, AB, or A) that may have a +/- after it. Think “AB-” as a low AB, higher than a high B but lower than a regular AB
  • A “summary” comment - this is a quick note to remind myself why you got the grade you did
  • My internal notes - this is very tied to the process by which we did grading. These were meant as notes to myself, so they might not make sense to anyone else. I was not consistent in writing everything down (some categories got more or less use at different points in going through things). There were many other columns (including several iterations of scoring), but these are the ones that might be interpretable to someone other than me:
    1. notes from docs - I always read the document first; these are the notes I made when reading the document. Often, things were clarified when I looked at other parts of the assignment
    2. notes from other parts* - these are notes I made while looking at the artifacts, demos, videos, …
    3. writeup and visualization content - I wrote these two comments as I was coming up with a grade to justify it to myself
    4. video - this is one of the columns I tried to put “codes” in - the codes aren’t consistent but they should correlate with the quality of the video. Videos only helped.
    5. adjustment - these are factors I wanted to remember (usually why the initial impression from the document why things may not be the same as the initial score)

Final Grading

We ended up going with something simple:

  • We created an “online” discussion score by combining End-of-week, Online Discussion (initial posts) and Seek and Find (initial posts). We gave the top 1/3 of the class an A, the negative outliers a B, and everyone else got something based on the scores in between.

  • We converted the 3 “major assignments” to letter grades. We weighted these based on the number of weeks (3,3,4) to get a single average.

We computed a score that was half of each. We then assessed penalties for attendance problems (missing many classes without telling us), and bonuses for online participation, but these were small and rarely moved people between boundaries.

In a sense the “everyone generally did OK at all the other stuff” is built in (which is why the scales basically start at B).

Only then, did we go back to check if this all lead to a reasonable grade distribution (in line with what I expect for a grad class). (A little more than 1/3 As, about 1/2 ABs and 1/6 Bs) - and it came out exactly matching my expectations. So I decided just to go with the simple thing.