\(\color{grey}{\text{Note: Code not displayed as it is too long}}\)
tab_model(cd_list1, show.ci = FALSE, show.p=FALSE, show.intercept = FALSE, show.se = TRUE,
dv.labels = c("All deaths (Age)", "All deaths (Controls)", "Motor-vehicle accidents (Age)",
"Motor-vehicle accidents (Controls)", "Suicide (Age)", "Suicide (Controls)"),
rm.terms=c("age_c", "age_squared"), title = "Sharp RD estimates of MLDA effects on mortality")
 | All deaths (Age) | All deaths (Controls) | Motor-vehicle accidents (Age) | Motor-vehicle accidents (Controls) | Suicide (Age) | Suicide (Controls) | ||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|
Predictors | Estimates | std. Error | Estimates | std. Error | Estimates | std. Error | Estimates | std. Error | Estimates | std. Error | Estimates | std. Error |
Da [1] | 7.66 | 1.44 | 9.55 | 2.00 | 4.53 | 0.77 | 4.66 | 1.14 | 1.79 | 0.45 | 1.81 | 0.69 |
Da [1] * age_squared | 2.90 | 2.30 | 0.20 | 1.31 | 0.03 | 0.80 | ||||||
Observations | 48 | 48 | 48 | 48 | 48 | 48 | ||||||
R2 / R2 adjusted | 0.595 / 0.577 | 0.669 / 0.639 | 0.703 / 0.689 | 0.722 / 0.696 | 0.475 / 0.452 | 0.487 / 0.439 |
\(\color{grey}{\text{Note: (Age) represents regression ran with only age as control and (Controls) represent regression ran with age, age2 and age2 interacted with being older than 21.}}\)
tab_model(cd_list2, show.ci = FALSE, show.p=FALSE, show.intercept = FALSE, show.se = TRUE,
dv.labels = c("All deaths (Age)", "All deaths (Controls)", "Motor-vehicle accidents (Age)",
"Motor-vehicle accidents (Controls)", "Suicide (Age)", "Suicide (Controls)"),
rm.terms=c("age_c", "age_squared"), title = "Sharp RD estimates of MLDA effects on mortality")
 | All deaths (Age) | All deaths (Controls) | Motor-vehicle accidents (Age) | Motor-vehicle accidents (Controls) | Suicide (Age) | Suicide (Controls) | ||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|
Predictors | Estimates | std. Error | Estimates | std. Error | Estimates | std. Error | Estimates | std. Error | Estimates | std. Error | Estimates | std. Error |
Da [1] | 9.75 | 1.93 | 9.61 | 3.06 | 4.76 | 1.10 | 5.89 | 1.69 | 1.72 | 0.69 | 1.30 | 1.10 |
Da [1] * age_squared | -0.87 | 14.13 | 6.97 | 7.81 | -2.63 | 5.06 | ||||||
Observations | 24 | 24 | 24 | 24 | 24 | 24 | ||||||
R2 / R2 adjusted | 0.703 / 0.675 | 0.707 / 0.646 | 0.474 / 0.423 | 0.509 / 0.405 | 0.541 / 0.497 | 0.548 / 0.453 |
\(\color{grey}{\text{Note: (Age) represents regression ran with only age as control and (Controls) represent regression ran with age, age2 and age2 interacted with being older than 21.}}\)
cd_plot1 <- RDestimate(all ~ agecell, data=cd_new, cutpoint = 21)
plot(cd_plot1)
title(main="Effect of Alcohol Consumption on Mortality", xlab="Age (Years)", ylab="Mortality Rate (per 100,000)")
cd_plot2 <- RDestimate(mva ~ agecell, data=cd_new, cutpoint = 21)
plot(cd_plot2)
title(main="Effect of Alcohol Consumption on Mortality", xlab="Age (Years)", ylab="Motor-Vehicle Mortality Rate (per 100,000)")
cd_plot3 <- RDestimate(suicide ~ agecell, data=cd_new, cutpoint = 21)
plot(cd_plot3)
title(main="Effect of Alcohol Consumption on Mortality", xlab="Age (Years)", ylab="Suicide Mortality Rate (per 100,000)")
cd_new %>%
ggplot(aes(x=agecell, y=mva, color=Da)) +
geom_point(col="black") +
geom_vline(xintercept=21, linetype="longdash") +
stat_smooth(method = lm, show.legend = FALSE) +
labs(y="Motor-Vehicle Mortality Rates (per 100,000)", x="Age (Years)") +
theme_minimal() +
ggtitle("Effect of Alcohol Consumption on Motor-Vehicle Mortality")
This figure is depicting the effect of alcohol consumption on motor-vehicle mortality rates. Considering the legal drinking age (in the US) is 21, that is where the cutoff point is placed in this model. The graph shows that individuals just below 21 years have lower motor-vehicle mortality rates than those just above 21. We see a jump in the mortality rate at the cutoff point 21, where the rate is around 30 (per 100,000) for those below the cutoff, and around 34 (per 100,000) for those just above 21.