Hypothesis Testing

Definition (#f7aeae)

Important (#edcae9)

Extra (#fffe9d)

Types of variables:

  1. Independent variable.

  2. Dependent variable.

  3. Control variable.

  4. Extraneous variable.

  5. Confounding variable.

Research approaches:

  • Correlational research:

    • Examines the relationship between two or more variables.

    • No manipulation of variables, only observation.

    • Correlation does not imply causation.

    • Surveys, observational studies, secondary data analysis.

  • Experimental research:

    • Investigates cause-and-effect relationships.

    • Manipulates an IV and measures its effect on a DV.

    • Uses random assignment to reduce bias.

    • Controlled environment to eliminate extraneous variables.

  • Between subject design:

    • Different groups of participants are assigned to different conditions.

    • Each participant experiences only one level of the IV.

    • Reduces the risk of practice or fatigue effects since participants only do one condition.

    • Requires more participants since each person is only tested in one condition.

  • Within subjects design:

    • The same participants go through all conditions of the experiment.

    • Reduces individual differences, as each person serves as their own control.

    • More statistically powerful because differences between individuals are minimized.

    • Risk of order effects (practice, fatigue), which can be controlled using counterbalancing.

Hypothesis: A prediction or educated guess about what will happen.

  • Focuses on the difference or relationship between two or more variables.

  • Comes from theories, past research, or personal observation.

  • Must be testable using data and statistical analysis.

  • Helps researchers stay focused on what they are trying to find out.

Types of hypothesis:

  • Null hypothesis (H0):

    • States there is no effect or no difference.

    • It’s the default or “no change” assumption.

    • Ex: There is no difference in concentration levels between students who listen to music and those who don’t.

  • Alternative hypothesis (H1/Ha):

    • States there is an effect or a difference.

    • What the researcher expects or hopes to find.

    • Ex: Students who listen to music have different concentration levels than those who don’t

Hypothesis testing:

  • Goal: Gather enough evidence from data to reject the null hypothesis (H₀) and support the alternative hypothesis (H₁).

  • A p-value tells us if the result is statistically significant.

  • If p< 0.05, the result is significant.

    • This means the result is unlikely due to chance.

    • We reject H0 and support H1.

Why Not Say “Accept” the Hypothesis?

  • We never truly prove a hypothesis. We only gather evidence.

  • Saying “accept” sounds like we are 100% sure it’s true.

  • Reject H₀: We found enough evidence.

  • Fail to reject H₀: We don’t have enough evidence, but that doesn’t mean H₀ is true.

3 types of errors:

  1. Type 1 error: (Fale positive)

    • Incorrectly rejecting H₀ when it is actually true.

  2. Type 2 error: (False negative)

    • Incorrectly failing to reject H₀ when it is false.

  3. Family-wise error:

    • Happens when you do many tests at once.

    • Increases the chance of false positives (Type 1 errors).

    • Solution: Bonferroni Correction.

      • Adjusts the significance level to reduce errors.

      • Divide 0.05 by the number of tests to get the new p-value threshold.

      • Ex: If you do 5 tests, 0.05 ÷ 5 = 0.01 for each test.