Week 5: Statistical inference
Chi-Square (χ²) Tests
1. Introduction
The Chi-Square (χ²) test is one of the most important non-parametric tests used in statistics to
analyze data expressed in frequencies or counts rather than measurements.
It helps to test hypotheses about categorical variables that is, variables that can be divided into
different categories such as gender, customer preference, education level, etc.
2. Purpose of the Chi-Square Test
Chi-square tests are generally used for two main purposes:
1. Goodness of Fit Test – to determine whether the sample data fits a particular theoretical
distribution.
2. Test of Independence (or Association) – to check whether two categorical variables are
related or independent of each other.
3. Characteristics of Chi-Square Test
• It is a non-parametric test (does not assume normal distribution).
• It is based on frequency data (observed and expected frequencies).
• It can only be used with categorical (nominal or ordinal) data.
• The test statistic is always non-negative.
• The sampling distribution of χ² is positively skewed but approaches normality as sample
size increases.
4. The Formula
The general formula for the Chi-Square test statistic is:
The larger the value of χ², the greater the difference between observed and expected frequencies,
suggesting a poor fit (or possible association).
5. Chi-Square Goodness of Fit Test
Purpose:
To test whether the sample data matches an expected distribution.
Hypotheses:
• Null hypothesis (H₀): The observed frequencies fit the expected distribution.
• Alternative hypothesis (H₁): The observed frequencies do not fit the expected distribution.
where k = number of categories.
Decision Rule:
Compare the calculated χ² value with the critical value from the chi-square distribution table at a
given level of significance (α, usually 0.05).
• If χ² calculated > χ² critical → Reject H₀.
• If χ² calculated ≤ χ² critical → Fail to reject H₀.
Example 1: (Business Example)
A company claims that its market share in four regions (North, South, East, West) is equal.
To test this, a survey of 200 customers was conducted, yielding the following results:
Region Observed (O)
North 60
South 50
East 40
West 50
Hence, the observed distribution fits the expected one — the company’s claim holds true.
6. Chi-Square Test of Independence
Purpose:
To determine whether two categorical variables are independent or associated.
Hypotheses:
• H₀: The two variables are independent (no relationship).
• H₁: The two variables are not independent (there is a relationship).
Example 2: (Business/Marketing Example)
A marketing manager wants to know if gender is associated with preference for product type.
A survey of 100 people gives the following data:
Prefer Product A Prefer Product B Total
Male 30 20 50
Female 10 40 50
Total 40 60 100
Step 2: Apply the formula
7. Chi-Square Test of Association in Economics
Economists often use chi-square tests to analyze qualitative relationships, such as:
• Relationship between education level and employment status.
• Association between income group and spending habits.
• Independence between region and access to financial services.
• Relationship between occupation and loan approval.
8. Assumptions of Chi-Square Test
1. Data must be in the form of frequencies (counts), not percentages or ratios.
2. Observations must be independent of each other.
3. Expected frequency in each cell should generally be ≥ 5 (for validity).
4. The sample must be randomly selected.
9. Limitations
• Not suitable for small samples.
• Does not indicate the strength or direction of association—only whether one exists.
• Sensitive to sample size (a large sample can produce significant results even for small
differences).
• Applicable only to categorical (not numerical) data.
10. Summary Table
Degrees of
Type of Test Purpose Hypothesis Example
Freedom
Tests if observed
Market share
Goodness of Fit frequencies match H₀: Fit is good k–1
distribution
expected
Independence Tests if two variables H₀: Variables are Gender vs Product
(r – 1)(c – 1)
Test are related independent preference
11. Conclusion
The Chi-Square Test is a powerful tool in business, economics, and social sciences for analyzing
categorical data.
It helps decision-makers:
• Evaluate whether their assumptions about distributions are valid, and
• Detect relationships between qualitative variables, such as consumer demographics and
behavior.
In short, Chi-Square Tests provide the foundation for analyzing patterns and associations when
numerical analysis (like t-tests or regression) is not applicable.