Publications by Aparajita Sur
Longitudinal HW 1 - Aparajita Sur
Problem 2 Simulations library(mvtnorm) library(gee) library(lattice) #Initial parameters set.seed(1996) n = 10 #observations per subject m = 25 #number of subjects beta = c(10,0) #Beta vector t = 1:10 #times tij Mean.model = beta[1] + beta[2] * t ### Functions to generate data ##Random Intercepts RI.data = function(sigma, tau, n=1...
4616 sym R (20449 sym/86 pcs) 38 img
Missing data and descriptive analysis (Part 1)
Baseline data and summary of missing responses at baseline Among those who have scam data at baseline, summary statistics and the # missing responses for the other variables are presented below. Overall(N=1272) Fraud Yes 85 (6.7%) No 1180 (92.8%) 8 7 (0.6%) 9 0 (0%) Negative Life Events Mean (SD) 2.53 (1.87) Median [Min, Max] 2.00 [0, 8....
5300 sym R (1822 sym/6 pcs) 1 img 5 tbl
Sensitivity Analysis plots
Sensitivity Analysis plots Method 1: Add sensitivity params to MI predictive mean matching imputations SA1 - MNAR outcomes SA2 - MNAR outcomes and MNAR tailoring variable Method 2: Add sensitivity params to Bayesian posterior means SA1 - MNAR outcomes SA2 - MNAR outcomes and MNAR tailoring variable Method 3: Add sensitivity params to PANS ...
482 sym
Weighted per-protocol
ITT vs Weighted Per-protocol API Comparisons Compared below is the ITT vs. Weighted Per-protocol API Comparisons. Ratio of ratios for all outcomes This analysis was a weighted and replicated GEE with treatment weights, stage 1 and stage 2 weights. So a heavy drinker who participated in both stages had three weights multipled (treatment, stage 1...
3121 sym 1 img 2 tbl
What factors are associated with participating in M-bridge?
Engagement in Stage 1 The table below presents the demographics of students randomized to an API, divied up by whether they “participated” in the first stage. We define “participation” as completing at least one self-monitoring survey and viewing at least one PNF page. NOTE: Some students clicked on the link but did not view any PNF pages...
3846 sym R (10621 sym/11 pcs) 4 img 4 tbl
ITT vs. Per-protocol AIM2 M-bridge Analysis
ITT vs Per-protocol API Comparisons The per-protocol analysis includes any student who participated in stage 1 (viewing at least one PNF page and doing at least one self-monitoring survey). All heavy drinkers who participated in Stage 1, also engaged in stage 2 (just visited the site, partial or full completion). There are 344 students in the per...
615 sym 1 img 3 tbl
Mbridge - Risk factors for not engaging
Demographic variables by participation Note: the missing data for those don’t have pnf data means they did not click the link to view the pnf (n=152). There were also students who clicked the link but then did not view any of the pages (n=31). Some students did not complete SM1 but did complete later surveys - specifically 32, 22 and 20 student...
2730 sym R (16046 sym/13 pcs) 3 tbl
HTML
No(N=100) Yes(Partial or complete)(N=58) Overall(N=158) Age (years) Mean (SD) 18.1 (0.219) 18.0 (0.131) 18.0 (0.192) Sex Male 41 (41.0%) 17 (29.3%) 58 (36.7%) Female 59 (59.0%) 41 (70.7%) 100 (63.3%) Race/ethnicity Asian 6 (6.0%) 3 (5.2%) 9 (5.7%) Black 0 (0%) 0 (0%) 0 (0%) Hispanic/Latinx 4 (4.0%) 1 (1.7%) 5 (3.2%) White 87 (87.0%) 50 ...
5 sym 1 tbl
HTML
None(N=57) Just visited(N=43) Partial(N=26) Complete(N=32) Overall(N=158) Age (years) Mean (SD) 18.1 (0.225) 18.0 (0.213) 18.0 (0.196) 18.0 (0) 18.0 (0.192) Sex Male 26 (45.6%) 15 (34.9%) 12 (46.2%) 5 (15.6%) 58 (36.7%) Female 31 (54.4%) 28 (65.1%) 14 (53.8%) 27 (84.4%) 100 (63.3%) Race/ethnicity Asian 2 (3.5%) 4 (9.3%) 1 (3.8%) 2 (6.3%) 9...
5 sym 1 tbl
HTML
No(N=61) Yes(N=376) Overall(N=437) Age (years) Mean (SD) 18.1 (0.402) 18.1 (0.235) 18.1 (0.266) Sex Male 20 (32.8%) 135 (35.9%) 155 (35.5%) Female 41 (67.2%) 241 (64.1%) 282 (64.5%) Race/ethnicity Asian 4 (6.6%) 40 (10.6%) 44 (10.1%) Black 1 (1.6%) 15 (4.0%) 16 (3.7%) Hispanic/Latinx 3 (4.9%) 18 (4.8%) 21 (4.8%) White 50 (82.0%) 282 (75...
5 sym 1 tbl