1. INFO 7470 Session 12 Updates John M. Abowd and Lars Vilhuber April 25, 2016

2. Lab for Session 8 Peer assessment of proposal – Due April 25, 23:00 UTC (=7PM EDT) April 11, 2016 © John M. Abowd and Lars Vilhuber 2016, all rights reserved 2

3. Lab for Session 10 Posted on Friday, April 15 First part (programming assignment) – due April 22, 23:00 UTC (=7PM EDT) – EXTENDED to April 29, 23:00 UTC Second part (peer assessment) – due April 29, 23:00 UTC (=7PM EDT) – EXTENDED May 6, 23:00 UTC Brief exercise follows April 11, 2016 © John M. Abowd and Lars Vilhuber 2016, all rights reserved 3

4. Last flipped class: Session 13 GIS and Geographic Analysis – Guest lecturer Nicholas Nagle (University of Tennessee – Knoxville and NCRN node) – Posted 4PM EDT 2016-04-25 in edX – Please view before May 2, 2015 class. April 11, 2016 © John M. Abowd and Lars Vilhuber 2016, all rights reserved 4

5. IMPUTATION – A TOY EXAMPLE April 11, 2016 © John M. Abowd and Lars Vilhuber 2016, all rights reserved 5

6. Consider the following problem You have a dataset (say, for marketing) which only has – County of residence – Tulip purchases You are asked to give an estimate of the tulip consumption by race, because your boss believes that matters for marketing efforts April 11, 2016 © John M. Abowd and Lars Vilhuber 2016, all rights reserved 6

7. Your data PersonID County of residence RaceTulip orders 1 Baldwin County, Alabama missing3 2 Baldwin County, Alabama missing2 3 Baldwin County, Alabama missing6 4 Winnebago County, Wisconsin missing2 5 Winnebago County, Wisconsin missing7 6 Winnebago County, Wisconsin missing3 7 Winnebago County, Wisconsin missing1 April 11, 2016 © John M. Abowd and Lars Vilhuber 2016, all rights reserved 7

8. Where to get race information? You download the 2014 ACS 1-year estimates – We will ignore for now uncertainty in the estimates April 11, 2016 © John M. Abowd and Lars Vilhuber 2016, all rights reserved 8

9. April 11, 2016 © John M. Abowd and Lars Vilhuber 2016, all rights reserved 9

10. Data CountyRaceEstimateMOEPercentPercent MOE Baldwin County, AlabamaWhite175327+/-1,82887.6%+/-0.9 Baldwin County, Alabama Black or African American20062+/-50310.0%+/-0.3 Baldwin County, Alabama American Indian and Alaska Native2965+/-7801.5%+/-0.4 Baldwin County, AlabamaAsian2388+/-1091.2%+/-0.1 Baldwin County, Alabama Native Hawaiian and Other Pacific IslanderNNNN Baldwin County, AlabamaSome other race3223+/-1,7591.6% Winnebago County, WisconsinWhite158847+/-1,50193.7%+/-0.9 Winnebago County, Wisconsin Black or African American4514+/-2882.7%+/-0.2 April 11, 2016 © John M. Abowd and Lars Vilhuber 2016, all rights reserved 10

11. April 11, 2016 © John M. Abowd and Lars Vilhuber 2016, all rights reserved 11

12. April 11, 2016 © John M. Abowd and Lars Vilhuber 2016, all rights reserved 12

13. April 11, 2016 © John M. Abowd and Lars Vilhuber 2016, all rights reserved 13

14. April 11, 2016 © John M. Abowd and Lars Vilhuber 2016, all rights reserved 14

15. April 11, 2016 © John M. Abowd and Lars Vilhuber 2016, all rights reserved 15

16. April 11, 2016 © John M. Abowd and Lars Vilhuber 2016, all rights reserved 16 White

17. April 11, 2016 © John M. Abowd and Lars Vilhuber 2016, all rights reserved 17

18. Five imputations Repeat the exercise another 4 times – With different seed! – Keep track of imputation (if stacking) Compute statistics according to formulae Toy example spreadsheet available on website April 11, 2016 © John M. Abowd and Lars Vilhuber 2016, all rights reserved 18