Generated by Bing Image Creator: Phot of beautiful nature scene, that includes blue sky, while clouds, yellow flowers, green fields.
This post is following of the above post.
In this post I use l_pc_gdp and country for explanatory variable.
Let's start.
The observed slope for l_pc_gdp is 11.3.
Next, let's generate bootstrap distribuution.
set.seed(260) boot_dist_mult2 <- df %>% specify(trust ~ l_pc_gdp + country) %>% generate(reps = 10000, type = "bootstrap") %>% fit()
boot_dist_mult2
Since I am interested in l_pc_gdp only, I only visualize for it.
boot_dist_mult2 %>% filter(term == "l_pc_gdp") %>% ggplot(aes(x = estimate)) + geom_histogram(color = "white") + labs(title = "Bootstrap Distribution for l_pc_gdp slope")
Above histogram does not include 0.
Next, let's calculate 95% confidence intervals.
multi_ci2 <- get_confidence_interval(boot_dist_mult2, level = 0.95, type = "percentile", point_estimate = obs_slope2)
multi_ci2 %>% filter(term == "l_pc_gdp")
95% confidence interval for l_pc_gdp is between 6.92 and 15.7, which does not include 0.
Let's add confidence interval and observed slope into above histogram.
boot_dist_mult2 %>% filter(term == "l_pc_gdp") %>% ggplot(aes(x = estimate)) + geom_histogram(bins = 30, color = "white") + geom_vline(xintercept = 6.92, color = "green", linewidth = 2) + geom_vline(xintercept = 15.7, color = "green", linewidth = 2) + geom_vline(xintercept = 11.3, color = "red", linewidth = 2) + labs(title = "Boostrap l_pc_gdp distribution and CI and observed slope")
So far,
I tries regression analysis to see whether trust in government is related to per capita GDP.
All four confidence interval don't include 0. So, I am very confident that per capita GDP is related to trust in government.
That's it! Thank you!
To read the first post,
