Matching Confidence Intervals and Hypothesis Tests

Jul 14, 2025

Overview

This lecture reviews the relationship between confidence intervals and hypothesis tests when analyzing means, focusing on conditions for when both methods provide the same conclusion.

Conditions for Matching Results

  • Confidence intervals and hypothesis tests yield the same result if two conditions are met.
  • Condition 1: The alternative hypothesis (Ha) uses the "not equal to" (≠) symbol.
  • Condition 2: The confidence level (%) plus the significance level (α) adds up to 100%.

Practical Example

  • Hypothesis: Null (H₀): μ = 80; Alternative (Ha): μ ≠ 80; where μ is the average height of male college basketball players.
  • Significance level is set to 5% (α = 0.05).
  • The p-value calculated is 0.02.
  • Since p-value < significance level (0.02 < 0.05), reject the null hypothesis.
  • Since both conditions are satisfied (Ha uses ≠ and 95% + 5% = 100%), results of the hypothesis test and confidence interval will match.

Interpreting Results

  • Confidence interval defines plausible values for μ (the population mean).
  • If the null value (80) is rejected in the hypothesis test, it will not be contained in the 95% confidence interval.
  • Rejection of the null hypothesis corresponds to the null value not appearing in the confidence interval.

Key Terms & Definitions

  • Null Hypothesis (H₀) — The default claim that there is no effect or difference (e.g., μ = 80).
  • Alternative Hypothesis (Ha) — The claim we test for, often stating μ ≠ value.
  • Significance Level (α) — The threshold for rejecting H₀, typically 0.05 (5%).
  • Confidence Interval — A range of plausible values for the population parameter at a specific confidence level.
  • P-value — Probability of observing data at least as extreme as the sample, assuming H₀ is true.

Action Items / Next Steps

  • Review Sections 9.3 and 9.4 on confidence intervals and hypothesis testing.
  • Understand how to apply the two conditions for matching results.
  • Practice identifying whether a confidence interval will contain the null value based on test results.