Anti-religious speech

Author

DL Oberski

Exercises

Using poLCA, fit a two-class LCA to these data.
Create a profile plot.
How would you label the classes?
What can you say about the estimated class sizes? What does this mean for the prevalence of the attitudes you labeled under (2)?
Model fit
1. How many parameters are there?
2. How many unique data patterns are there (fixing the sample size \(n=1713\))?
3. Can you explain the number of degrees of freedom?
4. Can you explain the value of the G^2 (\(G^2\)) and X^2 (\(\chi^2\)) statistics?

Data

library(tidyverse)
library(broom)
library(poLCA)

Read the data from the General Social Survey 1987. It’s not old, it’s a classic!

antireli <- read.csv("https://daob.nl/files/lca/antireli_data.csv")

head(antireli)

Show the data as pattern frequencies.

table(antireli) |> knitr::kable()

Y1	Y2	Y3	Freq
1	1	1	696
2	1	1	34
1	2	1	275
2	2	1	125
1	1	2	68
2	1	2	19
1	2	2	130
2	2	2	366

---
title: "Anti-religious speech"
author: "DL Oberski"
format: 
  html:
    theme: zephyr
    toc: true
    code-tools: true
    code-link: true
editor: visual
execute:
  cache: false
---

### Exercises

(@) Using `poLCA`, fit a two-class LCA to these data.
(@) Create a profile plot.
(@) How would you label the classes?
(@) What can you say about the estimated class sizes? What does this mean for the prevalence of the attitudes you labeled under (2)?
(@) Model fit
    a. How many parameters are there? 
    b. How many _unique_ data patterns are there (fixing the sample size $n=1713$)?
    c. Can you explain the number of degrees of freedom?
    d. Can you explain the value of  the `G^2` ($G^2$) and `X^2` ($\chi^2$) statistics?



### Data

```{r}
#| message: false
library(tidyverse)
library(broom)
library(poLCA)
```

Read the data from the General Social Survey 1987. It's not old, it's a classic!

```{r}
antireli <- read.csv("https://daob.nl/files/lca/antireli_data.csv")

head(antireli)
```

Show the data as pattern frequencies.

```{r}
table(antireli) |> knitr::kable()
```