Skip to main content
. Author manuscript; available in PMC: 2016 Jan 1.
Published in final edited form as: J Res Educ Eff. 2014 Jul 24;8(3):419–450. doi: 10.1080/19345747.2014.921259

Appendix Table 1.

Attrition in the New York City School Choice Scholarships Program

Math scores Reading scores
Year 1 Year 2 Year 3 Year 1 Year 2 Year 3
Panel A: Full sample, T and C
Number randomized 2666 2666 2666 2666 2666 2666
Number with missing test score 586 912 865 586 912 865
Number with valid test score 1977 1626 1734 1995 1697 1743
Number with invalid 99 test score 103 128 67 85 57 58
Panel B: Control group, no offer of scholarship
Number randomized 1292 1292 1292 1292 1292 1292
Number with missing test score 319 504 445 319 504 445
Number with valid test score 904 739 811 929 766 830
Number with invalid 99 test score 69 49 36 44 22 17
Panel C: Treatment group, received offer of scholarship
Number randomized 1374 1374 1374 1374 1374 1374
Number with missing test score 267 408 420 267 408 420
Number with valid test score 1073 887 923 1066 931 913
Number with invalid 99 test score 34 79 31 41 35 41

Notes: Table reports number of observations with valid scores or invalid scores of the total number randomized for whom data are present in the Mathematica restricted use data. Top panel presents total N in data, along with counts of those missing test scores, with valid test scores, and with invalid (99) test scores. Data from the New York City School Choice Scholarships Program evaluation conducted by Mathematica Policy Research.