Unfortunately, this level of detail is not available for the colleges listed above (Duke, Northwestern, and Vanderbilt). However, it is available in the Harvard lawsuit dataset. I added the overall admit rate to the numbers from my earlier post as listed below. When Harvard early action applicants had an an overall 6-7x greater admit rate than RD applicants, the analysis found a ~3.5x expected higher admit rate for SCEA applicannts over RD applicants with similar full controls, including similar hooks and strength of applicants. The remaining ~1.9x related to other differences between the applicant pools.
The analysis and differences in admit rates suggests ~3.5x increased SCEA admit rate over RD for direct SCEA boost, ~1.5x increased SCEA admit rate for differences in rate of hooks between SCEA and RD pools, and ~1.3x increased admit rate for other differences in applicant pools, including differences in strength of applicant pools. I’m sure Northwestern, Duke, and Vanderbilt have their own system that differs from Harvard and also have different applicant pools, but I’d be surprised if a 4x difference in admit rate is entirely due to differences in rate of hooks and/or applicant pool differences.
Harvard Class of 2018 Admit Rates
Early Action: Non-ALDC = 15.9%, ALDC = 63.7%, Overall = 24.5%
Regular Decision: Non-ALDC = 3.1%, ALDC = 18.3%, Overall = 3.5%
Harvard Class of 2019 Admit Rates
Early Action: Non-ALDC = 12.4%, ALDC = 61.9%, Overall = 19.8%
Regular Decision: Non-ALDC = 2.9%, ALDC = 16.1%, Overall = 3.2%