News and Geolocated Social Media Accurately Measure Protest Size Variation

@Joo_Jungseock_UCLA @Zachary_Steinert-Threlkeld_UCLA and @Ryan_Fox_Squire_SafeGraph, can you confirm that the Safegraph data used in is the pre-2017 Safegraph data that has a very different form than the current data, providing geohash-7 destinations (@Susan_Athey1 I recall you being the only one at Stanford to have used this original data)? And that such data was acquired back in 2017 and is not available from Safegraph for more recent years?

I’m sure I speak for many in saying that geohash-7 destinations could unlock useful insight for COVID-19 research, especially if an unknown but potentially significant amount of community spread may be happening in non-POI locations like in private social events or in public space.

@Daniel_Ho_Stanford_University @Simone_Speizer_Stanford @Cansu_Stanford @Konstantinos_Pelechrinis_University_of_Pittsburgh @David_Van_Dijcke_University_of_Michigan

@Derek_Ouyang_Stanford I can confirm that this sort of data is not available from SafeGraph. However it is possible to obtain this sort of data from SafeGraph partners like

@Derek_Ouyang_Stanford Here is an example from @Austin_L_Wright_Univ_of_Chicago using this version of the data, though it relies on POI locations.

@Derek_Ouyang_Stanford Also that paper is now published: