8.4 Item-level Relative Fit measures (Cont’d)

Before we try to identify the appropriate model for the following data. In this case, class represents the latent class/ groups, and prob represents the probability of being correct.

##   class prob
## 1    00 0.20
## 2    10 0.90
## 3    01 0.80
## 4    11 0.83

Let us explore the probability distribution of different attribute class. Can you guess the most appropriate model for the data?

Code
library(ggplot2)
df <- data.frame(class=c("00","10","01","11"),prob=c(0.2,0.9,0.8,0.83))
ggplot(data=df, aes(x=class, y=prob)) +
  geom_bar(stat="identity", width=0.5)+xlab("latent groups")+ylab("probability of success")