🔬

Hypothesis Testing and Significance

Jul 12, 2025

Overview

This lecture explains the purpose of hypothesis testing, the meaning of statistical significance, and how increasing sample size impacts the likelihood of detecting significant results.

Purpose of Hypothesis Testing

  • Hypothesis testing aims to make decisions about populations by evaluating whether to reject the null hypothesis in favor of the alternative.
  • Statistical power is the probability of correctly rejecting the null when the alternative is true.
  • A statistically significant result indicates sufficient evidence against the null hypothesis.

Statistical Significance and P-values

  • Statistical significance means the result is unlikely due to random chance, based on a chosen significance level (usually 0.05).
  • If the P-value is less than the significance level, we reject the null hypothesis and declare significance.
  • If the P-value is greater, we fail to reject the null and state there is not enough evidence.

Coin Flip Example and Sample Size Impact

  • Null hypothesis: coin lands heads or tails 50/50; alternative: probability is higher for starting face (51%).
  • With 700 flips, P-value = 0.298; since 0.298 > 0.05, we fail to reject the null hypothesis, deeming the result not significant.
  • With 7,000 flips, P-value = 0.471; since 0.471 < 0.05, we reject the null hypothesis, result is significant (Note: This seems like a lecture slip; normally, 0.471 > 0.05, but the main lesson focuses on larger sample size leading to lower P-values and significance).
  • Increasing sample size decreases P-value, making it easier to detect statistical significance.

Importance of Sample Size

  • A larger sample size enhances the test's power and likelihood of finding significant results.
  • Regardless of method, using a large enough sample increases the chance of rejecting the null hypothesis when appropriate.

Key Terms & Definitions

  • Null Hypothesis (H₀) — The default assumption, e.g., no difference or effect (coin is 50/50).
  • Alternative Hypothesis (H₁) — The claim that challenges the null (coin is more likely to land on starting face).
  • P-value — Probability of obtaining results as extreme as observed, under the null hypothesis.
  • Significance Level (α) — Threshold to declare significance, commonly 0.05.
  • Statistical Significance — Determination that observed results are unlikely due to random variation.
  • Statistical Power — Probability of correctly rejecting a false null hypothesis.

Action Items / Next Steps

  • Practice calculating P-values and interpreting their relationship to significance level.
  • Review how sample size affects statistical power and significance in hypothesis tests.