#Making dummy codes; Dummy coded regression

#First, we need to convert our nominal variable into a labelled character factor
#so we'll better be able to understand our output

Wk10nm$Group2 = factor(Wk10nm$Group, 
                               levels = c(1,2,3,4,5), 
                               labels = c("Grp1","Grp2","Grp3","Grp4","Grp5"))


#install.packages('fastDummies')
library(fastDummies)

# Dummy code MARSTAT2
Wk10nm2 <- dummy_cols(Wk10nm, select_columns = c("Group2"))


###Regression of DV1 on FOUR Group2 dummies,
# CHOOSING to leave one out ("reference category"), Group2_Grp1

fit5 <- lm(DV1 ~ Group2_Grp2+Group2_Grp3+Group2_Grp4+Group2_Grp5, data=Wk10nm2)
summary(fit5) # show results

## 
## Call:
## lm(formula = DV1 ~ Group2_Grp2 + Group2_Grp3 + Group2_Grp4 + 
##     Group2_Grp5, data = Wk10nm2)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -36.988 -11.219  -0.262   9.909  39.930 
## 
## Coefficients:
##             Estimate Std. Error t value Pr(>|t|)    
## (Intercept)   38.333      2.885  13.285  < 2e-16 ***
## Group2_Grp2   13.321      3.587   3.714 0.000267 ***
## Group2_Grp3    8.258      3.515   2.350 0.019798 *  
## Group2_Grp4   15.070      3.860   3.904 0.000130 ***
## Group2_Grp5   27.600      4.998   5.523 1.06e-07 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 15.8 on 195 degrees of freedom
## Multiple R-squared:  0.1582, Adjusted R-squared:  0.1409 
## F-statistic: 9.159 on 4 and 195 DF,  p-value: 8.414e-07

confint(fit5, level=0.95) # CIs for model parameters 

##                 2.5 %   97.5 %
## (Intercept) 32.642787 44.02388
## Group2_Grp2  6.246934 20.39549
## Group2_Grp3  1.326168 15.18996
## Group2_Grp4  7.457862 22.68249
## Group2_Grp5 17.743684 37.45632

lm.beta(fit5)

## Group2_Grp2 Group2_Grp3 Group2_Grp4 Group2_Grp5 
##   0.3497256   0.2245598   0.3476050   0.4274241