- see also [[220705_154903 propensity scores|propensity scores]] - 275 subjs ```r condition N 1: accuracy 183 2: interest 92 ``` covariate balance checks - linear regression for continuous variables (`cov ~ condition`) - chi-square/fisher test for categorical variables (contingency tables) results - 7 covariates differ between conditions: age, time/days between day1 & day2 survey completion, region, country, domain, AOT, covid concern - 4 covariates don't differ between conditions: tenure days, gender, role, conspiracy beliefs ```r # age differ between condition > summaryh(lm(age ~ condition, d1)) term results 1: (Intercept) b = 27.40, SE = 0.43, t(273) = 63.16, p < .001, r = 0.97 2: conditioninterest b = 1.61, SE = 0.75, t(273) = 2.15, p = .033, r = 0.13 # tenure days ns > summaryh(lm(tenuredays ~ condition, d1)) term results 1: (Intercept) b = 704.02, SE = 33.58, t(273) = 20.96, p < .001, r = 0.78 2: conditioninterest b = −88.67, SE = 58.06, t(273) = −1.53, p = .128, r = −0.09 # time/days between day 1 and day 2 completion significant > summaryh(lm(gap_days ~ condition, d1)) term results 1: (Intercept) b = 4.63, SE = 0.17, t(273) = 26.98, p < .001, r = 0.85 2: conditioninterest b = −1.16, SE = 0.30, t(273) = −3.92, p < .001, r = −0.23 # gender ns Female Male accuracy 83 100 interest 49 43 p-value = 0.2469 # region significant (more indians - condition 6) EU IN LATIN SEA accuracy 8 109 3 63 interest 13 24 6 49 p-value = 0.0004998 # country significant (more indians - condition 6) CO GR IN PH TW accuracy 3 8 109 59 4 interest 6 13 24 44 5 p-value = 0.0004998 # domain significant (more sm [social media]) game hitech media_ent prof sm support accuracy 15 24 15 9 119 1 interest 7 21 13 9 42 0 p-value = 0.02249 # role ns som tl tm accuracy 1 5 177 interest 0 5 87 p-value = 0.5402 # actively open-minded thinking significant > summaryh(lm(aot ~ condition, d1)) term results 1: (Intercept) b = 3.36, SE = 0.09, t(273) = 36.02, p < .001, r = 0.91 2: conditioninterest b = 0.36, SE = 0.16, t(273) = 2.24, p = .026, r = 0.14 # covid concern significant > summaryh(lm(covid_concern ~ condition, d1)) term results 1: (Intercept) b = 5.28, SE = 0.09, t(273) = 57.34, p < .001, r = 0.96 2: conditioninterest b = −0.56, SE = 0.16, t(273) = −3.53, p < .001, r = −0.21 # conspiracy beliefs ns > summaryh(lm(conspiracy ~ condition, d1)) term results 1: (Intercept) b = 7.54, SE = 0.11, t(273) = 68.52, p < .001, r = 0.97 2: conditioninterest b = 0.15, SE = 0.19, t(273) = 0.77, p = .444, r = 0.05 ```