setwd("~/Documents/Dropbox/Research/Adrian")
a<-read.csv ("AI_conservatism_new_intervention_study.csv", header=T, sep=",")
names(a)
## [1] "v1" "V2" "V3"
## [4] "V4" "V5" "V6"
## [7] "V7" "V8" "V9"
## [10] "V10" "Q26" "Q28"
## [13] "Q30" "Q34" "control"
## [16] "risk" "downwards" "upwards"
## [19] "car_risk_1" "car_trust_1" "age"
## [22] "gender" "inc" "edu"
## [25] "poli_1" "poli_2" "poli_3"
## [28] "LocationLatitude" "LocationLongitude" "LocationAccuracy"
a$cond[a$control==1]<-"control"
a$cond[a$risk==1]<-"risk"
a$cond[a$downwards==1]<-"down"
a$cond[a$upwards==1]<-"up"
main effects of condition
summary(lm(car_risk_1 ~ cond, data=a))
##
## Call:
## lm(formula = car_risk_1 ~ cond, data = a)
##
## Residuals:
## Min 1Q Median 3Q Max
## -65.464 -21.550 6.301 20.263 48.450
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 63.699 1.760 36.192 < 2e-16 ***
## conddown 1.765 2.671 0.661 0.509
## condrisk -12.149 2.568 -4.731 2.65e-06 ***
## condup 1.038 2.697 0.385 0.700
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 27.04 on 797 degrees of freedom
## Multiple R-squared: 0.04337, Adjusted R-squared: 0.03977
## F-statistic: 12.04 on 3 and 797 DF, p-value: 1.023e-07
summary(lm(car_trust_1 ~ cond, data=a))
##
## Call:
## lm(formula = car_trust_1 ~ cond, data = a)
##
## Residuals:
## Min 1Q Median 3Q Max
## -49.895 -24.558 -2.558 26.105 57.442
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 43.0847 1.9236 22.398 <2e-16 ***
## conddown -0.5267 2.9197 -0.180 0.8569
## condrisk 6.8100 2.8068 2.426 0.0155 *
## condup -0.1819 2.9479 -0.062 0.9508
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 29.55 on 797 degrees of freedom
## Multiple R-squared: 0.01087, Adjusted R-squared: 0.007151
## F-statistic: 2.921 on 3 and 797 DF, p-value: 0.03326
main effects of conservatism
summary(lm(car_risk_1 ~ poli_2, data=a))
##
## Call:
## lm(formula = car_risk_1 ~ poli_2, data = a)
##
## Residuals:
## Min 1Q Median 3Q Max
## -74.553 -21.208 6.299 20.618 46.299
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 74.5527 2.3007 32.404 < 2e-16 ***
## poli_2 -0.2085 0.0326 -6.396 2.71e-10 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 26.93 on 799 degrees of freedom
## Multiple R-squared: 0.0487, Adjusted R-squared: 0.04751
## F-statistic: 40.91 on 1 and 799 DF, p-value: 2.715e-10
summary(lm(car_trust_1 ~ poli_2, data=a))
##
## Call:
## lm(formula = car_trust_1 ~ poli_2, data = a)
##
## Residuals:
## Min 1Q Median 3Q Max
## -53.598 -23.644 -3.404 25.858 69.794
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 28.71321 2.45806 11.681 < 2e-16 ***
## poli_2 0.24885 0.03483 7.145 2.03e-12 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 28.77 on 799 degrees of freedom
## Multiple R-squared: 0.06005, Adjusted R-squared: 0.05887
## F-statistic: 51.05 on 1 and 799 DF, p-value: 2.035e-12
interactions between condition and conservatism - upwards only significant one
summary(lm(car_risk_1 ~ poli_2 * cond, data=a))
##
## Call:
## lm(formula = car_risk_1 ~ poli_2 * cond, data = a)
##
## Residuals:
## Min 1Q Median 3Q Max
## -71.714 -18.747 4.656 19.490 60.058
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 77.24220 4.29658 17.978 < 2e-16 ***
## poli_2 -0.20634 0.06009 -3.434 0.000626 ***
## conddown 6.62627 6.37547 1.039 0.298964
## condrisk -5.52827 6.09298 -0.907 0.364515
## condup -10.88118 6.30627 -1.725 0.084835 .
## poli_2:conddown -0.08228 0.09020 -0.912 0.361930
## poli_2:condrisk -0.11138 0.08620 -1.292 0.196681
## poli_2:condup 0.18090 0.08876 2.038 0.041866 *
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 26.19 on 793 degrees of freedom
## Multiple R-squared: 0.1073, Adjusted R-squared: 0.09941
## F-statistic: 13.62 on 7 and 793 DF, p-value: < 2.2e-16
summary(lm(car_trust_1 ~ poli_2* cond, data=a))
##
## Call:
## lm(formula = car_trust_1 ~ poli_2 * cond, data = a)
##
## Residuals:
## Min 1Q Median 3Q Max
## -58.819 -23.296 -1.917 24.150 68.616
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 24.49893 4.68238 5.232 2.15e-07 ***
## poli_2 0.28317 0.06549 4.324 1.73e-05 ***
## conddown -6.79081 6.94794 -0.977 0.3287
## condrisk 9.89365 6.64009 1.490 0.1366
## condup 12.77179 6.87253 1.858 0.0635 .
## poli_2:conddown 0.10653 0.09829 1.084 0.2788
## poli_2:condrisk -0.03890 0.09394 -0.414 0.6789
## poli_2:condup -0.19494 0.09673 -2.015 0.0442 *
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 28.54 on 793 degrees of freedom
## Multiple R-squared: 0.08222, Adjusted R-squared: 0.07412
## F-statistic: 10.15 on 7 and 793 DF, p-value: 3.327e-12
control<-subset(a, cond=="control")
up<-subset(a, cond=="up")
down<-subset(a, cond=="down")
risk<-subset(a, cond=="risk")
summary(lm(car_risk_1 ~ poli_2, data=control))
##
## Call:
## lm(formula = car_risk_1 ~ poli_2, data = control)
##
## Residuals:
## Min 1Q Median 3Q Max
## -62.799 -18.136 5.963 20.761 43.391
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 77.24220 4.35374 17.742 < 2e-16 ***
## poli_2 -0.20634 0.06089 -3.389 0.000824 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 26.53 on 234 degrees of freedom
## Multiple R-squared: 0.04678, Adjusted R-squared: 0.04271
## F-statistic: 11.48 on 1 and 234 DF, p-value: 0.0008238
no effect of conservatism in upwards condition only
summary(lm(car_risk_1 ~ poli_2, data=up))
##
## Call:
## lm(formula = car_risk_1 ~ poli_2, data = up)
##
## Residuals:
## Min 1Q Median 3Q Max
## -65.089 -18.924 5.402 20.369 36.183
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 66.36102 4.77863 13.887 <2e-16 ***
## poli_2 -0.02544 0.06762 -0.376 0.707
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 27.11 on 173 degrees of freedom
## Multiple R-squared: 0.0008172, Adjusted R-squared: -0.004958
## F-statistic: 0.1415 on 1 and 173 DF, p-value: 0.7073
summary(lm(car_risk_1 ~ poli_2, data=down))
##
## Call:
## lm(formula = car_risk_1 ~ poli_2, data = down)
##
## Residuals:
## Min 1Q Median 3Q Max
## -70.405 -12.169 3.965 17.395 41.530
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 83.86847 4.42156 18.968 < 2e-16 ***
## poli_2 -0.28861 0.06314 -4.571 9.03e-06 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 24.58 on 179 degrees of freedom
## Multiple R-squared: 0.1045, Adjusted R-squared: 0.09952
## F-statistic: 20.89 on 1 and 179 DF, p-value: 9.03e-06
summary(lm(car_risk_1 ~ poli_2, data=risk))
##
## Call:
## lm(formula = car_risk_1 ~ poli_2, data = risk)
##
## Residuals:
## Min 1Q Median 3Q Max
## -71.71 -20.94 0.49 19.49 60.06
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 71.71393 4.34724 16.496 < 2e-16 ***
## poli_2 -0.31772 0.06219 -5.109 7.34e-07 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 26.35 on 207 degrees of freedom
## Multiple R-squared: 0.112, Adjusted R-squared: 0.1077
## F-statistic: 26.1 on 1 and 207 DF, p-value: 7.34e-07
summary(lm(car_trust_1 ~ poli_2, data=control))
##
## Call:
## lm(formula = car_trust_1 ~ poli_2, data = control)
##
## Residuals:
## Min 1Q Median 3Q Max
## -52.82 -24.62 -2.86 26.00 60.19
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 24.49893 4.75659 5.151 5.51e-07 ***
## poli_2 0.28317 0.06652 4.257 3.00e-05 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 28.99 on 234 degrees of freedom
## Multiple R-squared: 0.07187, Adjusted R-squared: 0.0679
## F-statistic: 18.12 on 1 and 234 DF, p-value: 3.004e-05
summary(lm(car_trust_1 ~ poli_2, data=up))
##
## Call:
## lm(formula = car_trust_1 ~ poli_2, data = up)
##
## Residuals:
## Min 1Q Median 3Q Max
## -46.09 -25.99 -3.77 22.44 62.20
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 37.27072 5.27072 7.071 3.64e-11 ***
## poli_2 0.08822 0.07459 1.183 0.238
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 29.9 on 173 degrees of freedom
## Multiple R-squared: 0.008022, Adjusted R-squared: 0.002288
## F-statistic: 1.399 on 1 and 173 DF, p-value: 0.2385
summary(lm(car_trust_1 ~ poli_2, data=down))
##
## Call:
## lm(formula = car_trust_1 ~ poli_2, data = down)
##
## Residuals:
## Min 1Q Median 3Q Max
## -54.729 -18.164 -3.789 20.859 68.616
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 17.70812 4.68079 3.783 0.000211 ***
## poli_2 0.38969 0.06684 5.830 2.54e-08 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 26.02 on 179 degrees of freedom
## Multiple R-squared: 0.1596, Adjusted R-squared: 0.1549
## F-statistic: 33.99 on 1 and 179 DF, p-value: 2.536e-08
summary(lm(car_trust_1 ~ poli_2, data=risk))
##
## Call:
## lm(formula = car_trust_1 ~ poli_2, data = risk)
##
## Residuals:
## Min 1Q Median 3Q Max
## -58.819 -24.393 3.486 25.607 51.661
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 34.39258 4.77330 7.205 1.06e-11 ***
## poli_2 0.24427 0.06828 3.577 0.000432 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 28.93 on 207 degrees of freedom
## Multiple R-squared: 0.05822, Adjusted R-squared: 0.05367
## F-statistic: 12.8 on 1 and 207 DF, p-value: 0.0004323
library(visreg)
all1<-lm(car_trust_1 ~ poli_2*cond, data=a)
contup<-subset(a, cond=="control" | cond=="up")
contup1<-lm(car_trust_1 ~ poli_2*cond, data=contup)
plotting just the control and upwards conditions
visreg(contup1, "poli_2", by="cond", overlay=TRUE, partial=FALSE, scale="response", xlab="cons", ylab="trust")
#plotting all conditions
visreg(all1, "poli_2", by="cond", overlay=TRUE, partial=FALSE, scale="response", xlab="cons", ylab="trust")
summary(a$poli_2)
## Min. 1st Qu. Median Mean 3rd Qu. Max.
## 0.00 50.00 68.00 64.25 90.00 100.00
cons<-subset(a, poli_2 <=60)
lib<-subset(a, poli_2>60)
looking at main effects of condition among conservatives and liberals
(defining this as above or below 50 would be most natural, but the conservatism sample is quite small in that case)
summary(lm(car_trust_1~ cond, data=cons))
##
## Call:
## lm(formula = car_trust_1 ~ cond, data = cons)
##
## Residuals:
## Min 1Q Median 3Q Max
## -43.357 -23.357 -5.451 25.293 63.560
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 34.440 2.856 12.060 <2e-16 ***
## conddown -4.978 4.283 -1.162 0.2460
## condrisk 8.917 4.059 2.197 0.0287 *
## condup 7.586 4.346 1.746 0.0817 .
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 28.56 on 350 degrees of freedom
## Multiple R-squared: 0.03718, Adjusted R-squared: 0.02893
## F-statistic: 4.505 on 3 and 350 DF, p-value: 0.00407
summary(lm(car_trust_1~ cond, data=lib))
##
## Call:
## lm(formula = car_trust_1 ~ cond, data = lib)
##
## Residuals:
## Min 1Q Median 3Q Max
## -55.667 -24.186 1.559 24.333 56.424
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 49.441 2.461 20.087 <2e-16 ***
## conddown 3.490 3.770 0.925 0.3552
## condrisk 6.225 3.672 1.696 0.0907 .
## condup -5.865 3.792 -1.547 0.1226
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 28.7 on 443 degrees of freedom
## Multiple R-squared: 0.02265, Adjusted R-squared: 0.01603
## F-statistic: 3.422 on 3 and 443 DF, p-value: 0.01725
summary(lm(car_risk_1~ cond, data=cons))
##
## Call:
## lm(formula = car_risk_1 ~ cond, data = cons)
##
## Residuals:
## Min 1Q Median 3Q Max
## -74.613 -10.418 5.171 18.333 38.582
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 70.310 2.519 27.913 <2e-16 ***
## conddown 4.303 3.778 1.139 0.2556
## condrisk -8.892 3.580 -2.483 0.0135 *
## condup -5.481 3.833 -1.430 0.1536
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 25.19 on 350 degrees of freedom
## Multiple R-squared: 0.03893, Adjusted R-squared: 0.0307
## F-statistic: 4.726 on 3 and 350 DF, p-value: 0.003017
summary(lm(car_risk_1~ cond, data=lib))
##
## Call:
## lm(formula = car_risk_1 ~ cond, data = lib)
##
## Residuals:
## Min 1Q Median 3Q Max
## -64.667 -22.252 4.782 21.248 57.162
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 58.8382 2.3156 25.410 < 2e-16 ***
## conddown -0.6204 3.5471 -0.175 0.861
## condrisk -16.0004 3.4542 -4.632 4.76e-06 ***
## condup 5.8284 3.5676 1.634 0.103
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 27 on 443 degrees of freedom
## Multiple R-squared: 0.08044, Adjusted R-squared: 0.07421
## F-statistic: 12.92 on 3 and 443 DF, p-value: 4.193e-08