Skip to content ↓

COVID-19

Learn more about how we are responding to COVID-19 in this blog post from our Dean and dedicated FAQs.

MIT student blogger Kathleen E. '22

Classes, by the Group Chats by Kathleen E. '23

and their potential replacement, Interstellar!

At the start of each semester, I typically join Facebook Messenger group chats for most of my classes. These group chats are somewhat helpful for:

  1. Resolving confusion about class policies.
  2. Getting hints about pset problems!
  3. Spreading the word about announcements made in lecture (like pset extensions!) before they’re emailed out.
  4. Memeing.

This semester, an exciting new resource called Interstellar was born! Interstellar seems to me like a much-needed upgrade from Facebook Messenger group chats. Students can use it to share the classes they’re taking and see the classes their friends are taking, work on psets together in virtual Zoom lounges, share resources, ask questions (and post memes) in a forum, and keep track of their assignments. I’ll be taking a gap year this year, so I unfortunately won’t be using Interstellar, but I look forward to using it next fall!

Interstellar was created by four students in the class of 2022: Daniel S., Akshaj K., Vivek B., and Guang C. I’m guessing the name is inspired by “Stellar,” which is MIT’s platform for learning management. On Stellar, each class has a page with things like assignment files and grades. Over the summer, many classes were moved to a different system called Canvas, so Stellar might be in the procss of getting phased out. Regardless, I like the name “Interstellar” a lot. So many of the things we use at MIT have cool names: Kerberos, Athena, and now Interstellar!

landing page

Though limited in function, the Facebook Messenger group chats were pretty fun. A few highlights of things sent in the chats:

  1. The time our 18.03 (Differential Equations) professor accidentally emailed us way too many times about a relevant IAP course:way too many emails
  2. The time our 6.006 (Introduction to Algorithms) professor surprised us by jumping onto the chalk ledge:jason ku agile
  3. This meme about how it feels to run out of extensions for 6.009 (Fundamentals of Programming) labs:6.009 meme

I downloaded the message logs from six of my Facebook Messenger group chats to do a little analysis and see if I could get any insights. The classes were:

  1. 18.03 Differential Equations
  2. 18.06 Linear Algebra
  3. 6.009 Fundamentals of Programming
  4. 6.034 Artificial Intelligence
  5. 6.036 Introduction to Machine Learning
  6. 6.042 Mathematics for Computer Science

Honestly, it wasn’t that interesting, but here’s what I found:

I. Comparing class and group chat size

The size of the class01 Taken as the average class size listed on firehose.guide wasn’t necessarily indicative of how big the group chat would be. Among my six chats, there was a low of 26% of the class (6.034) and a high of 69% of the class (6.036) present in the group chat.

bar chart showing both class and group chat sizeII. Comparing class difficulty and group chat inclusion

There slight trend between the difficulty of the class (as measured by the weekly hours on Firehose) and the percent of the class in the chat! This makes sense to me; the easier the class, the less of a need there is for a group chat. The harder the class, the more helpful a group chat is in getting through the semester.

scatter plot

III. Reaccs

I didn’t send that many messages in most of these chats. However, I did use the react feature a lot. I wanted to see if my reacts differed much between chats.

bar chart showing reacts

I was a lot more active in some chats than others. However, the proportion of positive to negative reacts didn’t reflect how I felt about the class. For example, I had a ton of negative reacts in 6.034 but found it very enjoyable, while the opposite can be said of 6.006.

IV. Harvard sentiment

In the 6.036 (Introduction to Machine Learning) group chat, the message with the highest positive sentiment02 as measured by Vader (Valence Aware Dictionary for sEntiment Reasoning) was a sarcastic message to a Harvard student in the chat. In response to a Harvard student calling MIT their engineering school, an MIT student said, “Ok sweetie keep cross regging for your engineering classes. Hope the commute gets better xoxo.”

Why were Harvard students in the chat? Harvard and MIT students can cross-register for classes at the other school, and MIT computer science classes are a pretty popular pick. Anyway, funnily enough, when I made a word cloud out of the messages with the most negative sentiment in the chat, “Harvard” appeared (in the top left corner). Sorry, Harvard friends! Please keep cross-regging (not sarcastic).

6.036 negative message word cloud

word cloud of messages in the 6.036 group chat with <-0.5 compound sentiment from VADER

  1. Taken as the average class size listed on firehose.guide back to text
  2. as measured by Vader (Valence Aware Dictionary for sEntiment Reasoning) back to text