Detailed Notes on Hypothesis Testing

Basics of Hypothesis Testing

  • Hypothesis testing involves evaluating claims about population parameters.
  • Procedures include:
    • P-value Method
    • Critical Value Method
    • Confidence Intervals

Definitions

  • Hypothesis: A claim about a population characteristic.
  • Hypothesis Testing: A process for assessing the validity of such claims.

Example of a Claim

  • Claim: "More than half of Internet users use two-factor authentication."
  • Using symbol p for the proportion, the claim translates to: p > 0.5.

Evaluating Results

  • Data from 926 users shows:
    • 464 users (50.1%) claimed they use two-factor authentication.
    • 925 users (99.9%) clearly support the claim.
  • A result of 510 users (55.1%) requires hypothesis testing to determine significance.

Key Concepts

  • Significance: Evaluating when sample results are significantly low or high.

Null and Alternative Hypotheses

  • Null Hypothesis (H0): The value of a parameter is equal to a specified value (e.g., H0: p = 0.5).
  • Alternative Hypothesis (H1): Indicates that the parameter value differs (e.g., H1: p > 0.5).

Testing a Population Proportion

Normal Approximation Method

  • Objective: Conduct a hypothesis test regarding a population proportion p.
  • Notation:
    • n: sample size
    • p: population proportion (in H0)
    • p̂ : sample proportion
    • q: 1 - p

Requirements

  1. Simple random sample.
  2. Satisfy binomial conditions:
    • Fixed number of trials.
    • Independent trials.
    • Two categories of success and failure.
    • Constant probability.
  3. Success and failure conditions: np ≥ 5 and nq ≥ 5.

Example of Testing a Claim

  • Sample of 926 users shows 482 (52%) yes responses. Test that p > 0.5.
  • Significant level set at α = 0.05.
  • Sample statistical approaches:
    • Test statistic (z) calculated from sample values.

P-Value Method

  • Calculate test statistic and corresponding P-value.
  • Compare P-value to α to determine whether to reject H0.

Decision and Conclusion

  • Fail to reject H0 indicates insufficient evidence to support alternative claims.
  • Use clear language to report findings, emphasizing the inability to provide evidence for claims.

Type I and Type II Errors

  • Type I Error: Rejecting a true null hypothesis (probability α).
  • Type II Error: Failing to reject a false null hypothesis (probability β).

Power of a Hypothesis Test

  • Defined as 1 - β, indicating the probability of rejecting a false null hypothesis.
  • Example: Power may vary with the actual population parameter.

Determining Sample Size for Desired Power

  • Illustrates how researchers determine sample size to achieve high statistical power in hypothesis tests.