Last time we discussed 4 settings: population density (log or no log) and exposure (use it or not). Did those results come in? I think the point was to establish that we get consistently better prediction without using exposure and, in addition to better prediction, social lag was significant. Are those results on the wiki?
For Friday April 1st meeting discussion.