StatisticsInferential Statistics

The Frequency Table Shows the Results of a Survey: Did Voter Preferences Change?

Use a chi-square test for homogeneity to compare voter preferences before and after an earthquake from a complete 2×3 frequency table.

Question

Both before and after a recent earthquake, surveys asked voters which of three candidates they planned to support in an upcoming city council election. The frequency table shows the results of the survey.

Survey timePerezChungStevens
Before167128135
After214197225

At the α = 0.05 significance level, has the distribution of voter preferences changed since the earthquake?

Use a chi-square test for homogeneity because the question compares the distribution of one categorical response across two survey times.

1. State the hypotheses

  • H₀: The distribution of voter preferences is the same before and after the earthquake.
  • Hₐ: The distribution of voter preferences is not the same before and after the earthquake.

2. Find totals and expected counts

The row totals are 430 before and 636 after. The candidate totals are 381 for Perez, 325 for Chung, and 360 for Stevens, with grand total 1066.

For each cell,

expected count = (row total × column total) / grand total.

Survey timePerezChungStevens
Before expected153.687131.098145.216
After expected227.313193.902214.784

Every expected count is at least 5, so the expected-count condition is satisfied. The inference also assumes the surveyed observations are independent.

3. Compute the test statistic

χ² = Σ((O − E)² / E) = 3.2603.

There are two rows and three columns, so

df = (2 − 1)(3 − 1) = 2.

For χ² = 3.2603 with df = 2, the p-value is approximately 0.1959.

4. Make the decision

Because 0.1959 > 0.05, fail to reject H₀.

At the 5% significance level, the surveys do not provide sufficient evidence that the distribution of voter preferences changed after the earthquake. This is not proof that the distributions are identical; it means the observed differences are not statistically significant at α = 0.05.

Evidence boundary

题面、观测频数、显著性水平和教材核对答案来自 OpenStax;期望频数、逐格贡献及舍入值是根据这些数据计算得出。结论只说明证据不足以认定分布改变,不声称证明“完全没有改变”。

Sources

Các nguồn này hỗ trợ những khái niệm và phương pháp được giải thích ở trên.