Categorical Data: Chi-Square Flashcards Preview

Statistics & Data Analysis > Categorical Data: Chi-Square > Flashcards

Flashcards in Categorical Data: Chi-Square Deck (18)
Loading flashcards...
1

What time of data is Chi-square used for?

Nominal/categorical data.

2

What is the Chi-square Goodness of Fit Test?

Chi-square with one variable.

3

What two types of frequency does the Goodness of Fit Test use?

It compares the observed frequencies and expected frequencies.

4

How do you calculate expected frequency in the Goodness of Fit Test?

It is the number of participants / number of categories.

5

What are the two ways to enter data in the Goodness of Fit Test?

By participant or total frequency count.

6

What should you weight cases by?

Frequency.

7

What is a weakness of Chi-square?

Unable to tell which categories differ from which.

8

What is the Chi-square Test of Association?

Chi-square with two variables (also known as Pearson's Chi-square).

9

How do you calculate expected frequency in the Test of Association?

It is the (row total x column total) / grand total.

10

How do you calculate degrees of freedom for the Test of Association?

It is the (number of rows - 1) x (number of columns - 1).

11

How do you know if results are significant in the Chi-square tests?

The calculated statistic should be higher than the table value.

12

How would you calculate the Odds of females studying Psychology compared to Engineering?

Number of females studying Psychology / number of females studying Engineering.

13

How would you calculate the Odds Ratio of females studying Psychology compared to males?

Odds that females will study Psychology / Odds that males will study Psychology.

14

Name the two main assumptions of Chi-Square.

Observations must be independent (each participant should only contribute to one contingency cell).
There should be adequate expected frequencies in each cell.

15

What is the adequate expected frequencies in each cell?

No more than 20% of the expected frequencies should have a value of less than 5.

16

What is the solution for when expected frequencies exceed the 20% rule for contingency tables larger than 2x2?

Collapse variables together.

17

What is the solution for when expected frequencies exceed the 20% rule for 2x2 contingency tables?

Use Fisher's Exact Test statistic instead of Pearson's Chi-square statistic (this can only be done when sample size is small).

18

Why is categorical data seen as a weakness?

It gives us less information than data measured at other levels. We should always try and measure at higher levels.