Piazza statistics
Introduction
As we’re nearing the end of the quarter and I’m about to gather the data from our class forum (I’ve been using Piazza for the past four years in my classes), I was wondering how class participation has evolved between offerings of the same courses, or between different courses.
So here is all the (interesting) data I was able to collect.
Note that the statistics on Piazza have a specific format. When students ask a new question, that’s a post. When an instructor answers or when other students answer, that’s an answer. Now, on Piazza, there’s also the possibility to have follow-up discussions as part of each post. These discussions are not counted separately. Instead, Piazza reports a number of contributions which just includes everything (posts + answers + follow conversations).
Graphs
Contributions vs class size
The first thing I wanted to gauge was the Piazza activity per class. In order to compare apples with apples, let’s graph the number total contributions divided by the number of students for each class.
Some interesting observations here:
ECS 36C - SQ19
was by far the most active class. The number of contributions per students is almost double the mean for all the other classes.- The activity for
ECS 150
was higher in the first couple quarters (e.g., SQ17 and WQ18) and then stabilized itself for about 5 quarters. I’m guessing that after a couple offerings, and once the class contents became more stable, past students really started sharing their knowledge about the class with new students and there is a transmission of skills happening outside of the forum. My guess comes from the fact that the activity forECS 150
surged again after we switched to online learning (FQ20, WQ21). Since students could not get knowledge from past students, since there is very little outside of class, in-person interactions anymore, the official class forum took a significant role again. - What is weird is that when
ECS 36C
switched to online last year (SQ20), the activity was the lowest observed for that class, compared to the previous, in-person offerings. I am guessing here that the switch to online learning was unsurprisingly confusing and painful.
Students vs staff
The second thing I wanted to measure was the activity of the staff (i.e., myself and sometimes the TAs) compared to the activity of the students.
In some classes, where the contributions from students directly was very active
and of very high quality, then the staff contribution could be as low as ~15%
(e.g., ECS 36C - SQ19
). In some other classes, especially new offerings of a
class, the staff contribution could be as high as ~30% (e.g., ECS 150 - WQ17
or ECS 158 - FQ18
).
On average, about 20% of all contributions in each class came directly from me. (My wife says I shouldn’t be on Piazza so much 8-|)
Posts vs answers
Next, let’s measure the number of answers per post. When a student asks a new question on Piazza, the question can be answer by other students (collaboratively) or by the instructor or both.
As we can see, students are pretty active to answer other students’ questions. On average, almost 1.4 students collaborate to answer each new question.
Additionally, I (or TAs) also answer about 3/4 of all the questions on average.
Response time vs class size
Finally, let’s see the average response time, again versus the class size in order to fairly compare these numbers. Since the average class size is 200 students, the response time for each class was adjusted against that baseline.
As we can see, students can usually expect to receive an answer to their question within 8 minutes. In some classes, it’s been as low as 3 minutes! In some other classes, it’s been as high as 15 minutes…
Detailed data by course
You’ll find below the data I was able to collect from Piazza, and that was used to build the graphs above. If you play some more with the data and find interesting trends, please let me know!
ECS 30
#students | Total posts | Staff responses | Student responses | Total contributions | Prof contributions | TAs contributions | Avg response time | |
WQ18 | 393 | 936 | 534 | 1703 | 4718 | 793 | 50 | 4 min |
SQ18 | 352 | 698 | 619 | 1062 | 3386 | 437 | 410 | 6 min |
ECS 36C
#students | Total posts | Staff responses | Student responses | Total contributions | Prof contributions | TAs contributions | Avg response time | |
FQ18 | 179 | 674 | 539 | 1079 | 3746 | 701 | 190 | 5 min |
SQ19 | 160 | 822 | 603 | 1373 | 4779 | 505 | 202 | 4 min |
SQ20 | 155 | 373 | 311 | 387 | 1682 | 452 | 0 | 16 min |
ECS 50
#students | Total posts | Staff responses | Student responses | Total contributions | Prof contributions | TAs contributions | Avg response time | |
FQ19 | 149 | 486 | 308 | 715 | 2490 | 462 | 6 | 6 min |
ECS 150
#students | Total posts | Staff responses | Student responses | Total contributions | Prof contributions | TAs contributions | Avg response time | |
WQ17 | 142 | 568 | 505 | 472 | 2444 | 526 | 252 | 21 min |
SQ17 | 206 | 993 | 647 | 1266 | 4975 | 884 | 170 | 8 min |
WQ18 | 201 | 848 | 631 | 1061 | 4321 | 765 | 220 | 6 min |
SQ18 | 204 | 566 | 431 | 817 | 2826 | 491 | 200 | 8 min |
WQ19 | 183 | 531 | 329 | 754 | 2632 | 534 | 9 | 8 min |
SQ19 | 200 | 741 | 573 | 837 | 3288 | 452 | 336 | 9 min |
FQ19 | 192 | 617 | 386 | 809 | 2903 | 603 | 42 | 11 min |
WQ20 | 196 | 524 | 353 | 686 | 2246 | 468 | 0 | 15 min |
FQ20 | 198 | 864 | 600 | 1297 | 4531 | 863 | 0 | 10 min |
WQ21 | 196 | 746 | 531 | 1082 | 3849 | 788 | 0 | 10 min |
ECS 158
#students | Total posts | Staff responses | Student responses | Total contributions | Prof contributions | TAs contributions | Avg response time | |
FQ18 | 51 | 119 | 82 | 105 | 440 | 140 | 0 | 16 min |