Hypothesis Testing Essentials
Hypothesis testing is a statistical procedure that uses sample evidence to determine if a stated hypothesis about a population is reasonable or should be rejected.
Core Principles
- Null Hypothesis (H0): A statement about a population parameter assumed to be true, always containing an equality.
- Alternative Hypothesis (Ha): A claim that contradicts the null hypothesis, indicating what we suspect might be true.
- Type I Error: Rejecting a true null hypothesis (false positive).
- Type II Error: Failing to reject a false null hypothesis (false negative).
- Significance Level (α): The maximum allowable probability of a Type I error, typically set at 10% or less.
- P-value: The probability of observing sample data as extreme as, or more extreme than, the observed data, assuming the null hypothesis is true.
Action Steps
- 1. State the Null (H0) and Alternative (Ha) hypotheses.
- 2. Define the test procedure and the significance level (α).
- 3. Calculate the Test Statistic (TS) and the P-value.
- 4. Compare the P-value to α: Reject H0 if P-value < α; otherwise, do not reject H0.
- 5. Interpret the decision in the context of the problem.
Formulas
- One-sample z-test statistic: $$z = \frac{\bar{x} - \mu_0}{\sigma/\sqrt{n}}$$
- One-sample t-test statistic: $$t = \frac{\bar{x} - \mu_0}{s/\sqrt{n}}$$
- Two-sample t-test statistic: $$t = \frac{(\bar{x}_1 - \bar{x}_2) - (\mu_1 - \mu_2)}{\sqrt{\frac{s_1^2}{n_1} + \frac{s_2^2}{n_2}}}$$
- Paired t-test statistic: $$t = \frac{k_d}{\bar{s}_d/\sqrt{n}}$$
- One-sample z-test statistic (prop): $$z = \frac{\hat{p} - p_0}{\sqrt{\frac{p_0(1-p_0)}{n}}}$$
- Two-sample z-test statistic (prop): $$z = \frac{(\hat{p}_1 - \hat{p}_2) - (p_1 - p_2)}{\sqrt{\hat{p}(1-\hat{p})(\frac{1}{n_1} + \frac{1}{n_2})}}$$
- P-value for Ha: μ > μ0: $$P = P(Z > z_{stat})$$ (right-tail)
- P-value for Ha: μ < μ0: $$P = P(Z < z_{stat})$$ (left-tail)
- P-value for Ha: μ ≠ μ0: $$P = 2 \times \text{one-tailed P-value}$$
Key Terms
- Hypothesis: An assumption or statement about a population parameter.
- Parameter: A population mean or proportion.
- Null Hypothesis (H0): A statement about the population that will be assumed to be true.
- Alternative Hypothesis (Ha): A claim about the population that is contradictory to H0.
- Type I Error: Rejecting a true null hypothesis (false positive).
- Type II Error: Failing to reject a false null hypothesis (false negative).
- Significance Level (α): The maximum allowable probability of a Type I error.
- P-value: The probability that an outcome of the data will happen purely by chance when the null hypothesis is true.
- Test Statistic: A standardized value calculated from sample data used to determine whether to reject H0.
- One-tailed Test: A test that checks for deviation in only one direction (greater than or less than).
- Two-tailed Test: A test that checks for deviation in both directions (greater or less than).
Pro Tips
- The null hypothesis (H0) must always contain equality (=, ≤, or ≥).
- The alternative hypothesis (Ha) always contradicts the null hypothesis.
- A small p-value (typically ≤ 0.05) indicates strong evidence against H0, leading to its rejection.
- The Power of the Test (1-β) represents the probability of correctly rejecting a false null hypothesis.
Pitfalls to Avoid
- Failing to identify the parameter before analysis leads to incorrect hypothesis formulation.
- Confusing Type I and Type II errors results in misinterpreting test outcomes.
- Using a large p-value (> 0.05) to reject H0 leads to incorrect conclusions.
Myth vs Reality
- Hypothesis testing aims to prove the null hypothesis is true.: Hypothesis testing aims to determine if there is enough evidence to reject the null hypothesis; it does not prove it true.
- A Type I error is more serious than a Type II error.: The relative seriousness of Type I and Type II errors depends on the specific context of the problem.
Real World Examples
- Testing if college students take less than five years to graduate.: H0: μ ≥ 5, Ha: μ < 5 (left-tailed test).
- Testing if 29% of high school seniors get drunk each month.: H0: p = 0.29, Ha: p ≠ 0.29 (two-tailed test).
- A supplement company claims its Vitamin D pills raise blood D levels above the national average of 20 ng/mL.: This is a right-tailed test to see if the mean is greater than 20 ng/mL.
Statistics
- Maximum allowable probability of a Type I error (α): Set at most 10%
- Typical threshold for small p-value: ≤ 0.05
Quiz
- What is the primary purpose of hypothesis testing?: To determine if sample evidence supports rejecting a population hypothesis.
- Which of the following statements correctly defines a Type I error?: Rejecting a true null hypothesis.
- If a p-value is 0.03 and the significance level (α) is 0.05, what is the decision regarding the null hypothesis?: Reject H0.
Related cheat sheets
More like this