Due: Tuesday, April 23 at 11:59pm. Submit on Canvas.
It's challenging to design social computing systems that support positive norms and interactions. Our goal with Assignment 2 is to draw on the class's discussions—of online norms, of bustling spaces, and of prototyping—to fashion a pro-social discussion environment.
In brief: you will design a set of interactions for a small team brainstorm that aims to promote high levels of psychological safety amongst the team members. Let's unpack this piece by piece.
You will recruit and gather small teams of three people online who work together in a pseudonymous chat environment to brainstorm. Each team's goal will be to brainstorm creative titles for a GoFundMe campaign, with the aim of helping that campaign attract donations. (These titles will not actually be launched, so the result they produce is of less importance than the social environment you create for them.) For example, for this GoFundMe campaign, teams might brainstorm a title "Save A Future: No Teen Suicide".
The teams should convene online in a private online chatroom to brainstorm. Team members should not know each other (well) prior to the task. Teams have a 15 minute time limit to meet each other, brainstorm, and then select a favorite title to submit. Software such as Slack is typically a reasonable pick, but you can use any chat software you'd like where people use pseudonyms rather than real names so they don't know who each other are. We mean something that might look like this:
Example chat interface
You will design a written interaction scaffold that these teams should follow when brainstorming. These people have never worked together before, and you want them to help them have a fun and effective brainstorm. You will author a scaffold that will be displayed to the team at the start of their brainstorm, suggesting how they should interact with each other. For example, you might write something that looks like:
These are not intended to be good scaffolds, per se, but to give you a sense of the kinds of degrees of freedom you have in designing yours. There is no specific length constraint here, but given that the teams have 15 minutes start to finish, and that members haven't met previously, we recommend keeping it tight. You are not allowed to intervene while the team is working: you can't enforce that the team follows your plan. You may create automated systems (e.g., bots) that participate if it's an important part of the scaffold, but this is neither required nor expected.
Your goal in crafting the plan is to create high levels of psychological safety in the team. Psychological safety refers to a shared belief that it's safe to take risks and make mistakes on a team. In a massive data-driven study of its own teams, Google found that psychological safety was the most important predictor of happy and effective teams). After the fifteen-minute brainstorm, in addition to submitting their recommended title, team members should individually and privately self-report levels of psychological safety with this validated survey scale:
On a 1 (strongly disagree) – 7 (strongly agree) scale, indicate how strongly you agree with each of these statements:
To calculate the level of psychological safety that your plan produced, sum up the 7-point responses (reversing #1, #3, and #5, since they're reverse-coded) per person, and average across people on the team. From looking at this scale, one thing to note when you're designing is that if the team doesn't feel like they they are able to take risks, or come up with silly or useless ideas, they are unlikely to feel much psychological safety.
So, in brief: you will create an interaction scaffold for small teams to use when brainstorming catchy titles for GoFundMe campaigns. Assuming a successful brainstorm, you will measure your success by how much psychological safety the team felt when following your interaction plan.
How do you gain confidence that your approach will work? Reflect on the prototyping discussions in class. Start prototyping with smaller teams, maybe just a team of two (a partner and you), and then iterate! Find ways to convene small teams — swapping with classmates, borrowing friends or dormmates — and test out your ideas on them. It's extremely difficult to create a sense of psychological safety in just 15 minutes, so repeated brainstorming and prototyping are your best bet. Plan carefully so that you don't burn all your social capital: your friends are busy too. We will help facilitate prototyping meetups using Piazza.
Finally, once you're ready, convene three teams of three people and test your final plan using a chat platform of your choice and a 15-minute time limit per team. The three people on each team should not know each other, or if they do, should not be aware of each others' identities. You may recruit any volunteers who you can find conveniently: friends at Stanford, classmates, or friends or family from home, for example. (Reach out to the TAs if you are unsure about anything or have more questions about this.) All three should brainstorm for the same GoFundMe campaign; you can choose any campaign that you'd like. Keep the chat transcripts.
Submit one PDF that contains two pieces of content.
First, include your final scaffold, in plain text — we will be copy-pasting it into a discussion environment for the extra credit challenge.
Please also submit a PDF of no more than 750 words containing (1) a description of your process to generate the plan, (2) a report of what team interactions you observed arising from your plan when you performed your final tests, and (3) reflections on what worked and what didn't as you prototyped and iterated on earlier versions.
You will be graded primarily on two factors: (1) whether the process that you followed to prototype, and the designs that you tried, were grounded in lessons from the class; and (2) your analysis of how the final design impacted the team interactions.
Extra credit challenge: 15% extra credit. Outside of your tests, the staff will be launching all of your suggestions with teams of paid workers who we recruit on Amazon Mechanical Turk. We will follow the same method described above, and record the psychological safety levels reported by teams using your scaffolds. Note that we can only include plain text in the scaffolds for this deployment, so images and videos will need to be included as text links in your submission, and bots etc. will not be possible to test. We anticipate awarding the extra credit to approximately the top 5% of submissions by average psychological safety level of team members. Looking forward to seeing your designs fly!