Modeling the adoption of innovations in the presence of geographic and media influences: Supporting Information

Jameson L. Toole[1]∗, Meeyoung Cha[2], Marta C. Gonzalez[3] 
1 Engineering Systems Division, Massachusetts Institute of Technology, Cambridge, MA, USA 
2 Graduate School of Culture Technology, Korea Advanced Institute of Science and Technology, Korea 
3 Department of Civil and Environmental Engineering, Massachusetts Institute of Technology, Cambridge, MA, USA 
∗ E-mail: Corresponding jltoole@mit.edu

*** DATASET S1 README ***
We have provided to the community a dataset containing empirical data concerning Twitter’s adoption. This data was to calibrate and test our model of social contagion. We have included information pertaining to the 408 US cities modeled in a Microsoft Excel workbook with multiple sheets. The data was printed such that the ordering of each sheet (with noted exceptions) is consistent. The first entry in each sheet corresponds to the first city, the second entry to the second city, and so on. The sheet labeled time series week contains a 180 by 408 matrix where the rows correspond to weeks (time) and the columns to cities. A list of all sheets and their descriptions follows:

city_lat_lon : contains latitude and longitudinal coordinates for each city. 

city_names : the names and state for each city. 3. city type composition - the measured fraction of a cities population who were labeled early adopters. 

crit_mass_ach_times : the week at which each city achieved a critical mass (13.5%)

google_news : weekly news volume has measured from Google Trends (keyword twitter). There are 180 data points, one for each week. Values are normalized so that the maximum over the interval is 100.

google_search : weekly search volume as provided by Google Trends (keyword twitter). There are 180 data points, one for each week. Values are normalized so that the maximum over the interval is 100.

time_series_week : A 180 x 408 matrix where the (i,j)th element corresponds to the number of new uses who signed up for twitter in week i at location j.

total_users_per_city : the total number of users that signed up for twitter from March 2006 through August 2009 in each city.