```r # start n: 4557 condition treatment start_n <char> <char> <int> 1: persuadeHarris llm 1218 2: persuadeHarris video 1000 3: persuadeTrump llm 1259 4: persuadeTrump video 1080 # final n: 3127 (those who already voted don't get counted here) condition treatment final_n <char> <char> <num> 1: persuadeHarris llm 715 2: persuadeHarris video 769 3: persuadeTrump llm 801 4: persuadeTrump video 842 # n over time checkpoint question prepost has_response has_response_perc <num> <fctr> <char> <num> <num> 1: 0 q00_voted pre-treat 4557 100.00000 2: 1 q01_writescreen pre-treat 3444 75.57604 # 25% of people voted 3: 2 q02_av pre-treat 3297 72.35023 4: 3 q03_age pre-treat 3279 71.95523 5: 4 q04_income pre-treat 3275 71.86746 6: 5 q05_conserv pre-treat 3274 71.84551 7: 6 q06_policy_imp pre-treat 3257 71.47246 8: 7 q07_text1 pre-treat 3241 71.12135 9: 8 q08_text2 pre-treat 3233 70.94580 10: 9 q09_posttreat post-treat 3133 68.75137 11: 10 q10_screener post-treat 3127 68.61971 12: 11 q11_end post-treat 3127 68.61971 ``` ![[1729792757.png]] # how many people already voted? ```r condition treatment question n_voted start_n perc_voted <char> <char> <fctr> <int> <num> <num> 1: persuadeHarris llm q00_voted 255 1218 20.93596 2: persuadeTrump llm q00_voted 256 1259 20.33360 3: persuadeHarris video q00_voted 227 1000 22.70000 4: persuadeTrump video q00_voted 231 1080 21.38889 ``` # among those who made it to the question immediately before treatment n = 3233 made it to this point. ```r condition treatment N <char> <char> <int> 1: persuadeHarris llm 772 2: persuadeHarris video 773 3: persuadeTrump llm 839 4: persuadeTrump video 849 ``` differential attrition rate. how many attrited by the end of study ```r condition treatment attrit_n pre_treat_n attrit_perc <char> <char> <num> <int> <num> 1: persuadeHarris llm 57 772 7.3834197 2: persuadeHarris video 4 773 0.5174644 3: persuadeTrump llm 38 839 4.5292014 4: persuadeTrump video 7 849 0.8244994 > summ(lm(attrit ~ condition + treatment, data = d11)) term result <char> <char> 1: (Intercept) b = 0.07 [0.05, 0.08], p < .001 2: conditionpersuadeTrump b = -0.01 [-0.02, 0.00], p = .041 3: treatmentvideo b = -0.05 [-0.06, -0.04], p < .001 > summ(lm(attrit ~ (condition + treatment) * lean_bidentrump_1, data = d11)) term result <char> <char> 1: (Intercept) b = 0.06 [0.04, 0.08], p < .001 2: conditionpersuadeTrump b = -0.01 [-0.03, 0.01], p = .203 3: treatmentvideo b = -0.05 [-0.07, -0.03], p < .001 4: lean_bidentrump_1 b = 0.00 [0.00, 0.00], p = .306 5: conditionpersuadeTrump × lean_bidentrump_1 b = 0.00 [0.00, 0.00], p = .854 6: treatmentvideo × lean_bidentrump_1 b = 0.00 [0.00, 0.00], p = .851 ``` ## updated plots/results for these people ![[1729794325.png]] ![[1729794326.png]]