Context:
Women’s Voices. Women Vote. seeks to mobilize unmarried women to register and vote, but consumer data lacks marital status for almost one-third of women and is not highly accurate for identifying unmarried women.
Our Work:
MSHC Partners developed data on the characteristics of each household comprising at least one woman, then conducted a large sample telephone survey of marital status. We combined the survey results and the new household characteristics data using advanced statistical analysis and data mining procedures to estimate the probability that each woman was unmarried.
Results:
MSHC Partners successfully created a model that was more accurate than available consumer data, allowing Women’s Voices. Women Vote. to target 2 million unmarried women for registration in 8 states and 2.1 million unmarried women for GOTV in 11 states. |
