Hypothesis Testing Procedure

Hypothesis Testing Procedure

Overview

Hypothesis testing is a systematic process for making statistical inferences about population parameters based on sample data. This guide outlines the step-by-step procedure for conducting hypothesis tests.

Step-by-Step Procedure

1. State the Hypotheses

Null Hypothesis (H0)

Alternative Hypothesis (Ha)

2. Choose Significance Level (α)

3. Select Appropriate Test

Consider:

Common tests:

4. Calculate Test Statistic

5. Determine Critical Value(s)

6. Make Decision

Using Critical Value Method:

Using P-value Method:

7. State Conclusion

Common Pitfalls to Avoid

  1. Multiple Testing

    • Performing multiple tests increases Type I error
    • Use appropriate corrections (Bonferroni, etc.)
  2. Data Dredging

    • Testing multiple hypotheses without pre-specification
    • Increases false positive rate
  3. Misinterpreting P-values

    • P-value is not the probability that H0 is true
    • Small p-value doesn't guarantee practical significance
  4. Ignoring Assumptions

    • Each test has specific assumptions
    • Violating assumptions can invalidate results

Example: One-Sample t-test

Scenario

Testing if a new teaching method improves test scores (known population mean = 75)

Steps

  1. Hypotheses

    • H0:μ=75
    • Ha:μ>75
  2. Significance Level

    • α=0.05
  3. Test Selection

    • One-sample t-test (comparing mean to known value)
  4. Test Statistic

    • Sample mean = 78
    • Sample SD = 5
    • Sample size = 25
    • t=78755/25=3
  5. Critical Value

    • t0.05,24=1.711 (one-tailed)
  6. Decision

    • t=3>1.711
    • Reject H0
  7. Conclusion

    • "The new teaching method significantly improves test scores (p < 0.05)"