Lecture 4

Question 1

Factor analysis

Answer 1

Idenitfy hidden patterns or factors in data. It simplifies complex data.

Question 2

Why do we do factor analysis?

Answer 2

1-Wether we find the expected structure in the item responses.
2-Investigate the dimensionality.
3-Provides evidence for validity (does the test measure what it has to measure)

Question 3

What are 2 kinds of factor analysis?

Answer 3

1-Exploratory
2-Confirmatory

Question 4

Common factor model

Answer 4

1-Do my items measure my intended latent variable?
2-Are my items good indicator of this latent variable?
3-Item correlations should be 0

Question 5

In the common factor model what if the item correlation is not 0?

Answer 5

There might be a different reason that corrletaes items.

Question 6

What’s the use of common factor model?

Answer 6

1- Way to express correlations between items.
2- Assumes individual differences in the item scores are caused by one single latent variable.
3-Uses linear regression

Question 7

Unidimensional

Answer 7

a concept, test, or model measures or reflects only one underlying factor or trait.

Question 8

What is the correlation between residuals under comon factor model?

Answer 8

The correlation should be 0 because the model is unidimensional and residuals are uncorrelated with each other.