Proposal feedback

[bensoltoff]

• A bit less theory-centric. What is the importance of clustering and predicting these users and outcomes? You don’t have to have a full causal theory or model, but it isn’t clear yet why we should care
• So will this be something like a network analysis paper?
• Lots of advanced methods. How large is the data source? Will you do this on a personal computer or do you need to use a computing cluster?
• How will you label the dataset for supervised learning?
• You did not make clear in your presentation the social science value of your proposed study. I think there is value here, but I expect to see more of an emphasis on this in your literature review