jgy4
commented
2 years ago Hi all, looking forward to meeting tomorrow!
In regards to defining the 'Colleges' Sample do you think it would be sufficient to say "Campuses With Demographic Data Compiled by The Students Learn Students Vote Coalition from IPEDS (The Integrated Post-Secondary Education System) & OPE (Office of Post-Secondary Education)"?
Additionally, in regards to states I've compiled a list of state-by-state early voting laws/sources here:
- States in Tan are states that Ballot Ready includes in their early voting data.
- States in Red are states we believe do not have early voting or only allow it at a local clerks office
- There are some states in bright yellow that we were on the fence about - they either had conflicting information online or a large number of polling places that weren't only clerk offices.
Based on this we intend to exclude Alabama, Hawaii, New Hampshire, North Dakota, Rhode Island, South Carolina, Vermont, Washington, & Wyoming from our early voting statistics.
Looking forward to discussing this more!
nickeubank
dapoade
adrianefresh
Hey All,
For meeting Monday, the thing that Adriane and I are most interested in seeing I think are clear articulations of the exact population underlying all of the statistics we use. I think you are all doing a terrific job of measuring distances using different metrics, in the speed with which you have picked up the nuances of geospatial analysis are terrific! But I think the thing that we don't have a really good grip on is the exact population (sample) for which we are calculating our statistics
In particular:
Colleges
I understand we're using the 2,500 colleges and universities that we were able to merge between the HIFLD and the +1 demographic data, but what population does that correspond to? In our press release, when we say "only 30% of {define sample} colleges and universities have a polling place within a mile of campus", what goes in the
{define sample}spot? It isn't ALL colleges and universities in the US, so what is it?HIFLD as a public dataset and so it's fine to refer to that, but nobody else has this +1 spreadsheet, so we have to be able to define the population of colleges and universities that were included in the data. #84
States
We also want to make sure we're clear on the states we're including. For example, for early voting, we're using the Ballot Ready data (Yay!), and particularly the (number?) states with "meaningful" early voting (meaning we drop [list of states] that have no early voting, and [list of states] that only allow early voting at the government office.
For election day, we're using CPI for [list of states], and [other source? for which states?] for our other states.