"Rapleaf aggregates consumer data from data providers, cleanses it and maps it to emails, and ultimately makes it accesible through our easy to use Append portal and API. We partner with dozens of large and small data companies to aggregate data that we ultimately anonymize and tie to email addresses. We source it from only legitimate data bureaus who adhere to the highest consumer privacy standards -- sources that give consumers appropriate notice and choice about sharing their information and have opted in to make their data accesible."
What? Are they claiming to know the salary of all the github users used in this study? Were the users filtered down to those who had income data available? Because if so, that's a really weird (bad) selection technique. If not, what is being done with users whose income data is not available? Also where the hell is rapleaf getting their data?? This whole thing confuses the heck out of me, someone please help.
Rapleaf's API provides income given an email address, when available. I ran the list of emails from the git commit log across that API, which only gives me data points for the users who have income data available.
Is there really any way to correct for the fact that not all the data is going to be available? It certainly doesn't exclude a sampling bias, which I did mention in the post, possibly not aggressively enough.
http://www.rapleaf.com/about-us/faq/#where
"Rapleaf aggregates consumer data from data providers, cleanses it and maps it to emails, and ultimately makes it accesible through our easy to use Append portal and API. We partner with dozens of large and small data companies to aggregate data that we ultimately anonymize and tie to email addresses. We source it from only legitimate data bureaus who adhere to the highest consumer privacy standards -- sources that give consumers appropriate notice and choice about sharing their information and have opted in to make their data accesible."
What? Are they claiming to know the salary of all the github users used in this study? Were the users filtered down to those who had income data available? Because if so, that's a really weird (bad) selection technique. If not, what is being done with users whose income data is not available? Also where the hell is rapleaf getting their data?? This whole thing confuses the heck out of me, someone please help.