tag:blogger.com,1999:blog-3623899854668090474.post7942965309086161433..comments2023-05-28T06:36:18.771-07:00Comments on The Dataphiles: The popular get richerUnknownnoreply@blogger.comBlogger5125tag:blogger.com,1999:blog-3623899854668090474.post-66578552023999698022010-02-15T11:58:48.834-08:002010-02-15T11:58:48.834-08:00This comment has been removed by a blog administrator.Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-3623899854668090474.post-47147763186219883362010-02-13T11:07:17.608-08:002010-02-13T11:07:17.608-08:00This comment has been removed by a blog administrator.Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-3623899854668090474.post-41680302851707094442010-02-01T19:09:40.653-08:002010-02-01T19:09:40.653-08:00This comment has been removed by a blog administrator.Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-3623899854668090474.post-81262590813099528892008-09-23T15:20:00.000-07:002008-09-23T15:20:00.000-07:00I didn't think there were a huge number-- my impre...I didn't think there were a huge number-- my impression it was a relatively small percentage of committee-to-committee compared with committee-to-candidate. I think this information would be valuable for proximity work, but we haven't gotten there yet.<BR/><BR/>Also, at least in the FEC data, when individuals donated directly to a committee they usually have their own FEC ID, correct? In those cases we counted them as a "committee". It would be pretty easy to spot these in the dataset since they presumably aren't also collecting from other individuals.<BR/><BR/>For individual disambiguation we used last name + zip code-- we're aware it's messy so we've chose to focus mainly on the committee-candidate side.<BR/><BR/>I'm not a domain expert, so I'd welcome any other advice you have.Mary McGlohonhttps://www.blogger.com/profile/15192693557157094495noreply@blogger.comtag:blogger.com,1999:blog-3623899854668090474.post-34680291878813366092008-09-23T12:50:00.000-07:002008-09-23T12:50:00.000-07:00>The data, fully built, will form a tripartite,...>The data, fully built, will form a tripartite, directed graph.<BR/><BR/>There are also huge numbers of committee-to-committee transactions, which makes the graph not strictly tripartite. <BR/><BR/>> Individual donations are all filtered through committees (usually a candidate has one or several designated committees), <BR/><BR/>Not always, in some cases individuals make contributions directly to candidates, not to their committee (especially in-kind, expenses on behalf of)<BR/><BR/>Have you cleaned the dataset to standardize and disambiguate individuals names? Wouldn't impact general statistics about distribution of contribution sizes, but the names problem makes it very hard to match individuals to build an appropriate network.Anonymousnoreply@blogger.com