Repression Study 5 pilot analysis
Siwar Hasan-Aslih
2024-07-24
#means
S5 %>%
group_by(condition) %>%
summarise_at(.vars = c("ViolentPolice_sc", "ViolentProtester_sc"), .funs = mean)## # A tibble: 4 × 3
## condition ViolentPolice_sc ViolentProtester_sc
## <fct> <dbl> <dbl>
## 1 control 8.01 8.44
## 2 police_violence 84.4 18.3
## 3 protester_violence 13.5 89.2
## 4 mutual_violence 64.8 73.9S5 %>%
group_by(condition) %>%
summarise_at(.vars = c("MoralPolice_sc", "MoralProtester_sc"), .funs = mean)## # A tibble: 4 × 3
## condition MoralPolice_sc MoralProtester_sc
## <fct> <dbl> <dbl>
## 1 control 88.8 83.0
## 2 police_violence 15.4 71.3
## 3 protester_violence 75.2 10.9
## 4 mutual_violence 34.7 24.4S5 %>%
group_by(condition) %>%
summarise_at(.vars = c("HarmfulPolice_sc", "HarmfulProtester_sc"), .funs = mean)## # A tibble: 4 × 3
## condition HarmfulPolice_sc HarmfulProtester_sc
## <fct> <dbl> <dbl>
## 1 control 8.06 9.97
## 2 police_violence 84.5 20.8
## 3 protester_violence 15.0 89.5
## 4 mutual_violence 67.5 75.9S5 %>%
group_by(condition) %>%
summarise_at(.vars = c("TrustPolice_sc", "TrustGov_sc"), .funs = mean)## # A tibble: 4 × 3
## condition TrustPolice_sc TrustGov_sc
## <fct> <dbl> <dbl>
## 1 control 71.4 58.3
## 2 police_violence 24.4 40.1
## 3 protester_violence 67.6 49.9
## 4 mutual_violence 43.4 45.8#effects on trust
summary(model <- lm(TrustPolice_sc ~ condD1*condD2*condD3*ideology, S5, na.action=na.omit))##
## Call:
## lm(formula = TrustPolice_sc ~ condD1 * condD2 * condD3 * ideology,
## data = S5, na.action = na.omit)
##
## Residuals:
## Min 1Q Median 3Q Max
## -80.712 -18.644 0.333 17.342 72.783
##
## Coefficients: (8 not defined because of singularities)
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 49.99827 6.20183 8.062 1.39e-14 ***
## condD1 -48.24654 8.89643 -5.423 1.13e-07 ***
## condD2 -9.10310 9.73585 -0.935 0.350466
## condD3 -28.62519 8.59080 -3.332 0.000959 ***
## ideology 5.75203 1.48203 3.881 0.000125 ***
## condD1:condD2 NA NA NA NA
## condD1:condD3 NA NA NA NA
## condD2:condD3 NA NA NA NA
## condD1:ideology 0.40856 2.14739 0.190 0.849223
## condD2:ideology 0.88408 2.26528 0.390 0.696586
## condD3:ideology 0.09214 2.03506 0.045 0.963915
## condD1:condD2:condD3 NA NA NA NA
## condD1:condD2:ideology NA NA NA NA
## condD1:condD3:ideology NA NA NA NA
## condD2:condD3:ideology NA NA NA NA
## condD1:condD2:condD3:ideology NA NA NA NA
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 26.32 on 331 degrees of freedom
## Multiple R-squared: 0.4213, Adjusted R-squared: 0.4091
## F-statistic: 34.43 on 7 and 331 DF, p-value: < 2.2e-16summary(model <- lm(TrustGov_sc ~ condD1*condD2*condD3*ideology, S5, na.action=na.omit))##
## Call:
## lm(formula = TrustGov_sc ~ condD1 * condD2 * condD3 * ideology,
## data = S5, na.action = na.omit)
##
## Residuals:
## Min 1Q Median 3Q Max
## -54.419 -19.725 3.185 17.579 54.578
##
## Coefficients: (8 not defined because of singularities)
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 47.4533 6.0362 7.861 5.4e-14 ***
## condD1 -14.7323 8.6589 -1.701 0.0898 .
## condD2 6.5975 9.4759 0.696 0.4868
## condD3 -2.8467 8.3614 -0.340 0.7337
## ideology 2.9110 1.4425 2.018 0.0444 *
## condD1:condD2 NA NA NA NA
## condD1:condD3 NA NA NA NA
## condD2:condD3 NA NA NA NA
## condD1:ideology -0.8991 2.0901 -0.430 0.6674
## condD2:ideology -3.9503 2.2048 -1.792 0.0741 .
## condD3:ideology -2.5960 1.9807 -1.311 0.1909
## condD1:condD2:condD3 NA NA NA NA
## condD1:condD2:ideology NA NA NA NA
## condD1:condD3:ideology NA NA NA NA
## condD2:condD3:ideology NA NA NA NA
## condD1:condD2:condD3:ideology NA NA NA NA
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 25.61 on 331 degrees of freedom
## Multiple R-squared: 0.08103, Adjusted R-squared: 0.06159
## F-statistic: 4.169 on 7 and 331 DF, p-value: 0.0002042#plot trust in the police
S5_long <- S5 %>%
pivot_longer(cols = c(TrustGov_sc, TrustPolice_sc, MoralPolice_sc, HarmfulPolice_sc, PolicySupport),
names_to = "DV", values_to = "value")
S5_long$DV <- factor(S5_long$DV, levels = c("HarmfulPolice_sc", "MoralPolice_sc", "TrustPolice_sc", "TrustGov_sc", "PolicySupport"))
# Plotting
p_trustpol <- ggplot(S5_long, aes(x = DV, y = value, fill = condition)) +
stat_summary(fun = mean, geom = "bar", position = position_dodge(width = 0.8), width = 0.75) +
stat_summary(fun.data = mean_cl_boot, geom = "errorbar", colour = "grey",
position = position_dodge(width = 0.8), width = 0.2) +
scale_y_continuous(limits = c(0, 100)) +
theme_bw() +
scale_fill_manual(values = c("#ffffb3","#80b1d3", "#8dd3c7", "#fb8072")) +
labs(x = "Condition",
y = "Score",
title = "Effects on dependent variables") +
theme(
plot.title = element_text(hjust = 0.5, size = 14),
plot.subtitle = element_text(hjust = 0.5),
plot.caption = element_text(hjust = 0.5, face = "italic")
)
# Print the plot
print(p_trustpol)
#violin plot with ideology
S5 %>%
filter(!is.na(ideology.cat) & !is.na(TrustPolice_sc)) %>%
ggplot(aes(x = condition, y = TrustPolice_sc, fill = ideology.cat))+
geom_split_violin(trim = FALSE, alpha = .4)+
geom_boxplot(width = .2, alpha = .6, position = position_dodge(.25))+
scale_fill_viridis_d(option = "E") +
stat_summary(fun = "mean", geom = "point", position = position_dodge(width = 0.25)) +
stat_summary(fun.data = "mean_se", geom = "errorbar", width = .1, position = position_dodge(width = 0.25))+
scale_x_discrete(name = "Condition", labels = c("control", "police_violence", "protester_violence", "mutual_violence")) +
scale_y_continuous(name = "Trust in the police") +
coord_cartesian(ylim = c(0, 100)) + # Set the y-axis limits
theme_minimal()+
scale_fill_manual(values=c('tomato2',"steelblue2"), name = "Ideology")+
labs(title = "Effects of police violence on trust",
caption = "How much do you trust Philadelphia police department?")+
theme(
plot.title = element_text(hjust = 0.5, size = 16, face = 'bold', margin = margin(b = 20)),
plot.caption = element_text(hjust = 0.5, face = "italic", size = 12),
axis.text.x = element_text(size = 14, margin = margin(t = 10)),
axis.title.y = element_text(size = 16, face = 'bold', margin = margin(r = 15)),
axis.title.x = element_text(size = 14, face = 'bold', margin = margin(t = 15, b = 15)),
legend.text = element_text(size = 13),
legend.title = element_text(size = 14, face = 'bold'),
plot.margin = margin(t = 20, l = 20, b = 20)
)## Scale for fill is already present.
## Adding another scale for fill, which will replace the existing scale.