What is the formula for calculating expected frequencies ($e_{ij}$) in a contingency table?

(Row Total * Column Total) / Sample Size

Chi-Square Tests: Independence & Goodness of Fit

This chapter covers chi-square tests for comparing multiple proportions and assessing the independence of categorical variables. These tests are crucial for analyzing categorical data and drawing conclusions about population proportions and relationships between variables.

Core Principles

Chi-square distribution is used for the test statistic.
Data must be categorical for these tests.
Hypothesis testing framework (null and alternative hypotheses) is applied.
Observed frequencies (f_ij) are compared to expected frequencies (e_ij).
The test statistic measures the discrepancy between observed and expected values.
P-value is used to determine statistical significance.
The conclusion involves either rejecting or failing to reject the null hypothesis.

Action Steps

1. State the null ($H_0$) and alternative ($H_a$) hypotheses.
2. Set the significance level ($\alpha$).
3. Compute observed frequencies ($f_{ij}$) from sample data.
4. Compute expected frequencies ($e_{ij}$) for each cell.
5. Calculate the chi-square test statistic.
6. Determine the degrees of freedom (df).
7. Compute the p-value using the chi-square distribution.
8. Compare the p-value to $\alpha$ and make a conclusion.

Formulas

$ \chi^2 = \sum_{i} \sum_{j} \frac{(f_{ij} - e_{ij})^2}{e_{ij}} $
$ e_{ij} = \frac{(\text{Row } i \text{ Total})(\text{Column } j \text{ Total})}{\text{Sample Size}} $
$ df = (n-1)(m-1) $

Key Terms

Categorical Data: Data that can be divided into groups or categories, such as 'yes/no', 'male/female', or 'color'.
Chi-Square Distribution: A probability distribution that is used for tests involving categorical data, particularly in chi-square tests.
Contingency Table: A table used in statistics to display the frequency distribution of variables, often used for testing independence.
Nominal Variable: A variable that has no numerical value and is used for naming or labeling categories, e.g., gender, occupation.
Observed Frequency ($f_{ij}$): The actual count of data points in a specific cell of a contingency table.
Expected Frequency ($e_{ij}$): The count of data points that would be expected in a specific cell if the null hypothesis were true.
Test Statistic: A value calculated from sample data used to test a hypothesis; for chi-square tests, it's the chi-square value.
P-value: The probability of observing a test statistic as extreme as, or more extreme than, the one calculated from the sample data, assuming the null hypothesis is true.
Degrees of Freedom (df): A parameter that determines the shape of the chi-square distribution, calculated based on the number of rows and columns in the contingency table.

Real World Examples

Analyzing airline flight data to see if the proportion of late flights is the same across different airlines.: Testing equality of population proportions.
Investigating if there is a relationship between the type of airline ticket purchased (First, Business, Economy) and the type of flight (Domestic, International).: Test of Independence.
Determining if gender (Male, Female) is independent of handedness (Right-handed, Left-handed).: Test of Independence.

Timeline

March 2012: Sample data collected for airline flight delays (Delta, United, US Airways).
N/A: Bloomberg Businessweek subscriber study conducted on travel habits.
N/A: Hypotheses formulated for testing equality of proportions.
N/A: Chi-square test statistic calculated for flight delay data.
N/A: P-value determined for flight delay data.
N/A: Conclusion drawn regarding flight delay proportions.
N/A: Contingency table constructed for ticket type vs. flight type.
N/A: Chi-square test statistic calculated for ticket type vs. flight type.
N/A: P-value determined for ticket type vs. flight type.
N/A: Conclusion drawn regarding independence of ticket type and flight type.

People

Anderson, Sweeney, Williams, Camm, Cochran: Authors of the 'Statistics for Business & Economics' textbook, from which this chapter content is derived.

Quiz

What type of data is required for Chi-Square tests of independence and goodness of fit?: Categorical data
The null hypothesis for a test of independence states that two variables are:: Independent
What is the formula for calculating expected frequencies ($e_{ij}$) in a contingency table?: (Row Total * Column Total) / Sample Size