A2 – Edward Zhang

1. Observations

I had the opportunity to observe people before classes of varying sizes – before a small seminar (COS598), a medium-sized class (COS435), a large lecture (COS436), and a huge lecture (MUS103). All of my observations were in the classrooms or directly outside of them before classes.

1.1 General Observations

  • In general, students are either having conversations with other students or are doing something on their laptop/phone.
  • Before the seminar class, most students are chatting with each other. Topics of conversation generally revolved around the papers to be discussed in that lecture. Essentially the entire class knows each other, since all of them are members of the graphics lab, but in previous seminar classes people generally gain the same familiarity to be able to strike up casual or academic conversations with most others.
  • In the medium-sized class, very few people are talking with others. Most are on their laptops (the desk space and convenient outlets in the Friend classrooms make this very convenient).
  • In HCI, many people come to class with their friend groups and often chat with them on the way in and after sitting down. It appears that the popularity of the class means that most people are friends with some number of others in COS436.
  • In the huge class, it appears to be difficult to chat because of the acoustics of the room – people seem to stick in groups of two or three instead. There appear to be a lot more untalkative “singles” compared to HCI; many have their phones out while others have paper and pen out for notes (I assume these are mostly studious freshmen…). There are fewer laptops than I expect in a class of that size, perhaps due to fear of the professor.

1.2 Individual Observations

  1. COS435 (medium-sized class): I observed an individual who I was not acquainted with. He entered the classroom and went to the same seat that he usually sat in (IIRC), conveniently in front of me. After opening up his laptop, he made sure to plug it into an outlet (this reminded me of the tiny desks and lack of power in lecture halls – the small auditorium and the Friend classrooms are very convenient in comparison). This individual checked their email (apparently Princeton Gmail), glanced at Facebook, and then opened up what appeared to be code for an assignment. I also noticed his neighbor take out a physical mouse and start playing Starcraft right after the professor arrived; his game continued into the lecture (?!).
  2. MUS103 (Huge class in McCosh): A girl carrying a salad in a plastic container (from Frist?) came in (alone) and sat down in front of where I was sitting (and left one seat between her and the next person over; apparently the movie theater law of “at least one seat between separate parties” holds in class too). She checked her phone while eating the salad; I couldn’t tell what she was doing, although based on her hands I assume she sent at least one text. Right before the lecture started, she got up and retrieved one of the handouts at the front of the classroom (did she forget it or just have her hands full?)
  3. COS598 (seminar-sized): Although most interactions before this seminar were as described above, there was one interesting case. I observed a female grad student who was new to the department who took the time before class to find someone not already engaged in a conversation and introduce herself (I was one of the ones she accosted). She asked whether I was a graduate student, what I thought of the class so far, and we discussed some of the technicalities of the readings. Through our short conversation she mentioned that she was a grad student who had just switched from ELE to the COS PIXL group and wanted to get to know people. I then watched her go off to introduce herself to another graduate student (she did this twice more the next class, but after that she stopped – I assume she had met everyone by then). This special effort to get to know everyone personally made a huge impact on me, as I had never encountered someone who explicitly went to every person in the class for introductions. However, I think that since it was a seminar class, she probably would have gotten to know everyone eventually. I imagine that her main goal was to meet everyone in the research group rather than the class, but it still inspired my many ideas involving giving people contexts to meet others in a more practical way for a larger class.

Brainstorming

Shared brainstorming with Connie Wan (cwan)

  1. Traffic Light Crossing Planning Aid – Tracks status of the crossings at Washington and in front of Forbes.
  2. Say hi to the camera – Like security cameras at store entrances, let people wave at them to amuse themselves.
  3. Paperwork area – Get handouts, sign in, vote on class polls, etc. in one location
  4. Music Areas – Let people plug in their computers/iPods into communal (directional?) speakers
  5. Phone game XL – Put large screens somewhere in the classroom to make phone games a social activity.
  6. Phone silencer – Deactivate/silence phones during class automatically
  7. Bike rental – Have stations around campus with bikes that you can sign out to get to classes faster.
  8. Make-a-friend phone game – Adapt social mobile games to look for people in the same classroom.
  9. Classwide opt-in games – Classwide trivia game that everyone in the class can join into (like the Delta in-flight trivia game).
  10. Crowd Traffic Analysis – Tell people which seats/entrances are crowded so they can choose which entrance to go into.
  11. Food Smell Diffuser/Eliminator – Show off your delicious food by wafting the smell over the entire classroom. Or, eliminate the smell of obnoxiously pungent food.
  12. Informal Discussion Organizer – Share your informal conversation topics so that people can find you and join in
  13. Student(s) of the day – Introduce the entire class to each other by randomly selecting people to record and play a short 20 second video of themselves.
  14. Whiteboard/Graffiti area – Put whiteboards on some of (all) the walls for graffiti, psets, etc.
  15. Electronic device charging – map all the locations of the nearest outlets in the classroom and whether they’re in use or not
  16. Announcement/spam board – Post event announcements, cool links, lost & founds, restricted to the ten minutes before class
  17. Fun floor area – Make interactive floor tiles or simple wall touch gadgets for amusement on the way to class (inspired by Disneyworld lineup areas)
  18. Cooperative puzzle – Have the entire class work together on some multipart challenge (e.g. sporcle).

Paper Prototypes

My first choice for paper prototyping is the Informal Discussion Organizer (#12). I think it would be the most concretely useful, especially in classes like HCI where informal, creative discussion can introduce you to ideas, viewpoints, and systems that you might not have heard of otherwise.

Home screen for the “Conversation Finder”, aka Informal Discussion Organizer. Major features:

  • The current class and classroom (detected via your schedule OR through which access point you’re accessing wifi from)
  • A list of current conversation topics
  • A button allowing you to add a discussion/topic

When you click on one of the topics, it lists where the discussion is taking place and who started the conversation (and optionally who else is taking part).


When the “Add Topic” button is clicked on the main screen, you get brought to a simple form that lets you enter a subject, choose from a list of locations, and optionally indicate who else is discussing (your name would be auto-filled upon posting). You can post it by clicking the green button, or cancel by clicking the red one.

My second choice for paper prototyping is the Student of the Day. Every Princeton student is a fascinating person, and I think giving people the opportunity to get to know people in their classes is valuable both for social and academic reasons.

The Student of the Day (currently a generically named individual) displays the name and a short (under 30 seconds) video clip introducing themselves with whatever they can fit in the allotted time. Their netid is also displayed in case you want to contact them later. Underneath, a history of previous videos can be viewed based on date. (Names are random generic names)

At some point during the course, the app will select you as the Student of the Day. This message will show up prompting you to record your short clip.

Clicking the button will bring you to a camera recording interface, with a timer underneath showing you how long your video is.

User Testing

  1. Teodor Georgiev: This participant, after looking over the home screen of the app, decided to browse the current topics and decided to “run off and join the conversation”. This was great because they did not spend long on the app and got the relevant information for them to run off and start talking. When I had an accomplice start a conversation with him (and hint that it might be of interest to other people) he tried out the Add Topic page. Unfortunately, he felt that the process was a little bit tedious. Furthermore, he said afterwards probably wouldn’t have thought to open it up without the obvious prompting, and even if he did it was a little inconvenient to think up and type in the topic.

    This user spent a bit of time looking at and analyzing every part of the interface, but after examining everything and getting “back in character” he decided to “go find one of the conversations” right after viewing it.

  2. Brenda Hiller: Like the first participant, the user was able to use the main interface to find a conversation of interest very rapidly. In contrast to the first participant, this user was very excited to be able to add their topic and was very interested in advertising their conversation to the class. She did not think that the fluidness of the workflow was a significant downside, although when questioned further she did admit having it more streamlined would help a lot.

    This user checked a topic of interest, and went off to find the conversation right after taking the first image. The second image shows a conversation an accomplice struck up with the user, and the user added the conversation with little prompting.

  3. David Bieber: This user tried to click everything on the app, which caused a lot of confusion when things did not work (either not implemented in the prototype or not intended to be interactive). However after exhausting the possible click locations, the user decided to go and join a conversation. Upon looking at the options for adding a topic, he was confused by the options presented for locations because he had misconceptualized the use of the app – he had thought it was a general conversation finder, rather than only for a single class, at a specific time, and even after an explanation he gave up since he was in a rush.

    The user is vigorously swiping and pressing buttons on the left, and disappointed when only a few locations actually do something. On the right, they are confronted with a list of location options that they did not understand.

I think this study may be a little bit skewed, since it was obvious through the app that I intended the user to participate in conversations. However, it seems likely for at least one of the users tested that they probably would have just stayed on their laptop alone, rather than actively seek out conversations.

Insights

Having a list of conversations going on in a handy place is very valuable to people – my participants rapidly pulled out the app, found what was going on, and went off to talk. Like Google, we want people to spend as little time in our app as possible, and it was successful in this regard. However, it is clear that having people populate the list of conversation topics will be very difficult, especially given the already short waiting time before class. This app would definitely have to make the topic creation workflow nearly instant, or even automatic (picking up group conversations, extracting a topic, and posting them automatically). One interesting thing I noted was that people would only open up conversation info for topics they were interested in – they would click a topic of interest, determine where it was, and go find the people. Nobody opened the info for a conversation they were not interested in. This suggests that the topic description must be chosen extra carefully, since, if a larger list of topics was in the app, people would want to filter conversations more carefully.