Dummy Variables

Load Packages library(fastDummies) library(tidyverse) library(psych) Create a DataSet # Create a vector of race scores race <- c("White", "Black", "Asian", "Hispanic", "Other") # Generate random income values for each race (100 cases) set.seed(123) # for reproducibility income <- round(runif(100, min = 20000, max = 100000), digits = 2) # Repeat each race 20 times to get 100 cases race <- rep(race, each = 20) # Combine race and income into a data frame data <- data.frame(race, income) # Print the first few rows of the dataset print(head(data)) ## race income ## 1 White 43006.20 ## 2 White 83064.41 ## 3 White 52718.15 ## 4 White 90641.39 ## 5 White 95237.38 ## 6 White 23644.52 Create Dummy Variables data<-data %>% dummy_cols(select_columns = "race") Regress Income on Race (African Americans as the Reference Category) fit<-lm(income ~ race_Asian + race_Hispanic + race_Other + race_White, data=data) summary(fit) ## ## Call: ## lm(formula = income ~ race_Asian + race_Hispanic + race_Other + ## race_White, data = data) ## ## Residuals: ## Min 1Q Median 3Q Max ## -44169 -19531 -1137 18010 40481 ## ## Coefficients: ## Estimate Std. Error t value Pr(>|t|) ## (Intercept) 66138 5066 13.055 <2e-16 *** ## race_Asian -15015 7165 -2.096 0.0388 * ## race_Hispanic -7004 7165 -0.977 0.3308 ## race_Other -7173 7165 -1.001 0.3193 ## race_White -2073 7165 -0.289 0.7730 ## --- ## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1 ## ## Residual standard error: 22660 on 95 degrees of freedom ## Multiple R-squared: 0.05237, Adjusted R-squared: 0.01247 ## F-statistic: 1.313 on 4 and 95 DF, p-value: 0.2709

Feb 15, 2024