The blinding process is a core feature of the Exchange. It’s what makes it possible for hundreds of different progressive organizations, including candidate campaigns and independent expenditure groups, to share valuable data safely, securely, and in compliance with election laws.
In the Exchange, “blinded data” means that sensitive or identifying information about the data’s origin is obscured, while the core value of the data is preserved. This process has three main components:
1. Blinding the Data Source
When you receive data from the Exchange, you will not be able to see which specific organization collected it. The contributing organization is always anonymous.
2. Blinding the Date of Contact
The exact date and time of a voter contact is obscured. Instead of a specific timestamp, each contact is assigned a date that reflects a two-month collection interval.
3. Blinding the Survey Questions (Normalization)
To standardize data from hundreds of different member scripts, we use a process called normalization. DDx maps member survey questions into a structured taxonomy of standardized categories such as Candidate, Issue, Voting, Engagement, and others, each with defined sub-types. This allows for preserving the core intent of each question while obscuring the original proprietary language.
This process is what allows organizations across the ecosystem to share survey data without exposing their specific language, and it keeps the data consistent across the many variations in question wording, response scales, and field conventions that hundreds of member organizations submit.