setwd("~/Dropbox/Research/Adrian")
i<-read.csv ("AI_conservatism_intervention_study_v2.csv", header=T, sep=",")
#creating time 1 and time 2 DVs, plus their "difference scores"
i$easy1<-(i$easy_1 + i$predictable_1 + i$understand_1)/3
i$easy2<-(i$easy2_1 + i$predictable2_1 + i$understand2_1)/3
i$easydiff<-(i$easy2 - i$easy1)
i$help1<-(i$feeling_1 + i$needs_1 + i$comfort_1 + i$bestint_1 + i$lookout_1)/5
i$help2<-(i$feeling2_1 + i$anticipate2_1 + i$comfort2_1 + i$interests2_1 + i$lookout2_1)/5
i$helpdiff<-(i$help2 - i$help1)
i$risk1<- (i$risk_1 + i$safety.risk_1 + i$job.risk_1)/3
i$risk2<- (i$Q83_1 + i$Q84_1 + i$Q85_1)/3
i$riskdiff<- (i$risk2 - i$risk1)
i$trust1<-(i$trust_1 + i$hospital_1)/2
i$trust2<- (i$Q86_1 + i$Q87_1)/2
i$trustdiff<-(i$trust2 - i$trust1)
#creating condition variable
i$cond[i$risk==1]<-"risk"
i$cond[i$downwards==1]<-"down"
i$cond[i$upwards==1]<-"up"
i$cond[i$explain==1]<-"explain"
i$edu<-as.factor(i$edu)
i$inc<-as.factor(i$inc)
#no interaction for the "difference scores" (t2 measures - t1 measures)
summary(aov(trustdiff~cond * poli_1 + age + gender + inc + edu, i))## Df Sum Sq Mean Sq F value Pr(>F)
## cond 3 742 247.4 1.586 0.192
## poli_1 1 18 18.3 0.117 0.732
## age 1 26 26.0 0.167 0.683
## gender 1 376 375.6 2.407 0.122
## inc 6 990 165.1 1.058 0.387
## edu 5 1361 272.1 1.744 0.124
## cond:poli_1 3 637 212.4 1.361 0.254
## Residuals 380 59293 156.0summary(aov(riskdiff~cond * poli_1 + age + gender + inc + edu, i))## Df Sum Sq Mean Sq F value Pr(>F)
## cond 3 527 175.60 1.214 0.304
## poli_1 1 46 45.62 0.316 0.575
## age 1 20 19.89 0.138 0.711
## gender 1 25 24.82 0.172 0.679
## inc 6 664 110.62 0.765 0.598
## edu 5 926 185.21 1.281 0.271
## cond:poli_1 3 509 169.72 1.174 0.320
## Residuals 380 54951 144.61#interaction for just the t2 measure, but only significant for trust DV (not risk DV)
summary(aov(trust2~cond * poli_1 + age + gender + inc + edu, i))## Df Sum Sq Mean Sq F value Pr(>F)
## cond 3 350 117 0.202 0.8949
## poli_1 1 14562 14562 25.260 7.73e-07 ***
## age 1 4 4 0.007 0.9356
## gender 1 9761 9761 16.932 4.75e-05 ***
## inc 6 1247 208 0.360 0.9036
## edu 5 2198 440 0.762 0.5773
## cond:poli_1 3 4140 1380 2.394 0.0681 .
## Residuals 380 219066 576
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1summary(aov(risk2~cond * poli_1 + age + gender + inc + edu, i))## Df Sum Sq Mean Sq F value Pr(>F)
## cond 3 1697 565.8 1.122 0.339998
## poli_1 1 1002 1002.1 1.987 0.159478
## age 1 1 1.2 0.002 0.960934
## gender 1 489 489.0 0.970 0.325425
## inc 6 3760 626.6 1.242 0.283632
## edu 5 10917 2183.5 4.330 0.000763 ***
## cond:poli_1 3 1865 621.8 1.233 0.297460
## Residuals 380 191642 504.3
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1#breaking down the interaction between risk cond and conservatism
summary(lm(trust1 ~ poli_1, i))##
## Call:
## lm(formula = trust1 ~ poli_1, data = i)
##
## Residuals:
## Min 1Q Median 3Q Max
## -68.658 -14.175 3.661 18.325 51.169
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 48.83129 2.89028 16.895 < 2e-16 ***
## poli_1 0.19826 0.04214 4.705 3.51e-06 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 24.74 on 399 degrees of freedom
## Multiple R-squared: 0.05256, Adjusted R-squared: 0.05019
## F-statistic: 22.13 on 1 and 399 DF, p-value: 3.509e-06down<-subset(i, cond=="down")
up<-subset(i, cond=="up")
explain<-subset(i, cond=="explain")
risk<-subset(i, cond=="risk")
summary(lm(trust2 ~ poli_1, down))##
## Call:
## lm(formula = trust2 ~ poli_1, data = down)
##
## Residuals:
## Min 1Q Median 3Q Max
## -62.594 -12.165 3.118 19.653 41.588
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 54.48270 5.76562 9.450 2.07e-15 ***
## poli_1 0.17865 0.08794 2.032 0.0449 *
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 25.7 on 97 degrees of freedom
## Multiple R-squared: 0.04081, Adjusted R-squared: 0.03093
## F-statistic: 4.127 on 1 and 97 DF, p-value: 0.04493#only condition with no effect of conservatism on trust is the "upwards" condition (parallels last study's findings - this is the emotion condition)
summary(lm(trust2 ~ poli_1, up)) #####
## Call:
## lm(formula = trust2 ~ poli_1, data = up)
##
## Residuals:
## Min 1Q Median 3Q Max
## -62.55 -14.48 2.92 17.45 41.46
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 58.54381 5.59762 10.459 <2e-16 ***
## poli_1 0.09806 0.07913 1.239 0.218
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 23.15 on 100 degrees of freedom
## Multiple R-squared: 0.01513, Adjusted R-squared: 0.005277
## F-statistic: 1.536 on 1 and 100 DF, p-value: 0.2181summary(lm(trust2 ~ poli_1, risk))##
## Call:
## lm(formula = trust2 ~ poli_1, data = risk)
##
## Residuals:
## Min 1Q Median 3Q Max
## -81.46 -11.12 6.35 15.15 57.60
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 42.39557 6.02412 7.038 2.79e-10 ***
## poli_1 0.39068 0.08635 4.524 1.72e-05 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 24.83 on 97 degrees of freedom
## Multiple R-squared: 0.1743, Adjusted R-squared: 0.1658
## F-statistic: 20.47 on 1 and 97 DF, p-value: 1.717e-05summary(lm(trust2 ~ poli_1, explain))##
## Call:
## lm(formula = trust2 ~ poli_1, data = explain)
##
## Residuals:
## Min 1Q Median 3Q Max
## -66.232 -13.456 1.889 15.778 44.295
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 55.70514 5.38037 10.353 <2e-16 ***
## poli_1 0.16502 0.07901 2.089 0.0393 *
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 23.54 on 99 degrees of freedom
## Multiple R-squared: 0.04221, Adjusted R-squared: 0.03253
## F-statistic: 4.363 on 1 and 99 DF, p-value: 0.0393mediators
names(i)## [1] "V1" "V2" "V3"
## [4] "V4" "V5" "V6"
## [7] "V7" "V8" "V9"
## [10] "V10" "Q26" "Q34"
## [13] "Q63_1" "Q63_2" "Q63_3"
## [16] "Q63_4" "control" "easy_1"
## [19] "predictable_1" "understand_1" "useful_1"
## [22] "feeling_1" "needs_1" "comfort_1"
## [25] "bestint_1" "lookout_1" "risk_1"
## [28] "safety.risk_1" "job.risk_1" "trust_1"
## [31] "hospital_1" "Q69_1" "Q69_2"
## [34] "Q69_3" "Q69_4" "risk"
## [37] "Q66_1" "Q66_2" "Q66_3"
## [40] "Q66_4" "downwards" "Q65_1"
## [43] "Q65_2" "Q65_3" "Q65_4"
## [46] "explain" "Q64_1" "Q64_2"
## [49] "Q64_3" "Q64_4" "upwards"
## [52] "easy2_1" "predictable2_1" "understand2_1"
## [55] "useful2_1" "attn100_1" "feeling2_1"
## [58] "anticipate2_1" "comfort2_1" "interests2_1"
## [61] "lookout2_1" "Q83_1" "Q84_1"
## [64] "Q85_1" "Q86_1" "Q87_1"
## [67] "age" "gender" "inc"
## [70] "edu" "poli_1" "poli_2"
## [73] "poli_3" "LocationLatitude" "LocationLongitude"
## [76] "LocationAccuracy" "easy1" "easy2"
## [79] "easydiff" "help1" "help2"
## [82] "helpdiff" "risk1" "risk2"
## [85] "riskdiff" "trust1" "trust2"
## [88] "trustdiff" "cond"summary(aov(easy2~cond * poli_1 + age + gender + inc + edu, i))## Df Sum Sq Mean Sq F value Pr(>F)
## cond 3 2502 834.0 1.615 0.1853
## poli_1 1 2191 2191.1 4.244 0.0401 *
## age 1 1 1.1 0.002 0.9632
## gender 1 0 0.0 0.000 0.9925
## inc 6 2128 354.7 0.687 0.6603
## edu 5 6215 1243.0 2.407 0.0362 *
## cond:poli_1 3 3149 1049.7 2.033 0.1088
## Residuals 380 196202 516.3
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1summary(aov(help2~cond * poli_1 + age + gender + inc + edu, i))## Df Sum Sq Mean Sq F value Pr(>F)
## cond 3 3170 1057 1.566 0.197141
## poli_1 1 8014 8014 11.877 0.000632 ***
## age 1 9 9 0.014 0.906120
## gender 1 8080 8080 11.974 0.000601 ***
## inc 6 4242 707 1.048 0.393842
## edu 5 6684 1337 1.981 0.080579 .
## cond:poli_1 3 1046 349 0.517 0.671122
## Residuals 380 256403 675
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1