How to Perform Chi-Square Test in R: Syntax and Example

To perform a chi-square test in R, use the chisq.test() function with a contingency table or vector of observed counts. This function tests if there is a significant association between categorical variables or if observed frequencies differ from expected frequencies.

Syntax

The basic syntax of the chi-square test in R is:

• chisq.test(x, y = NULL, correct = TRUE)

Where:

• x is a contingency table or a vector of observed counts.
• y is an optional vector of expected counts or a second factor for two-sample tests.
• correct applies Yates' continuity correction for 2x2 tables by default.

r

chisq.test(x, y = NULL, correct = TRUE)

Example

This example shows how to perform a chi-square test on a 2x2 contingency table to check if two categorical variables are independent.

r

data <- matrix(c(30, 10, 20, 40), nrow = 2, byrow = TRUE)
rownames(data) <- c("Group1", "Group2")
colnames(data) <- c("Success", "Failure")

result <- chisq.test(data)
print(result)

Output

Pearson's Chi-squared test with Yates' continuity correction

X-squared = 8.3333, df = 1, p-value = 0.0039

Common Pitfalls

• Passing raw counts without converting to a table or matrix can cause errors.
• Using correct = TRUE applies Yates' correction which may be unnecessary for larger samples.
• Expected counts should not be too small; otherwise, the test may be invalid.
• Confusing the order of rows and columns can lead to misinterpretation.

r

## Wrong: passing a vector without table structure
chisq.test(c(30, 10, 20, 40))

## Right: convert to matrix/table first
chisq.test(matrix(c(30, 10, 20, 40), nrow=2, byrow=TRUE))

Quick Reference

Parameter	Description
x	Observed counts as table or matrix
y	Optional expected counts or second vector
correct	Apply Yates' continuity correction (default TRUE)
p-value	Probability value to decide significance
X-squared	Chi-square test statistic value

Key Takeaways

Use chisq.test() with a contingency table or matrix of counts to perform the chi-square test in R.

Yates' continuity correction is applied by default for 2x2 tables but can be disabled with correct = FALSE.

Ensure your data is in the correct table or matrix format before running the test.

Check that expected counts are not too small to keep the test valid.

Interpret the p-value to decide if there is a significant association between variables.