Statistics: Hypothesis Testing and T-Tests

Hypothesis Testing

• A statistical method that uses sample data to evaluate a hypothesis about a population.
• Consists of four steps: stating hypotheses, knowing the direction, choosing the level of significance, and analyzing the data.

Step 1: State Your Hypotheses

• Research Hypothesis: There is a significant difference/relationship between groups/variables.
• Null Hypothesis: There is no significant difference/relationship between groups/variables.

Step 2: Know the Direction

• No additional information provided.

Step 3: Choose the Level of Significance

• Typically set at α = 0.05, which means a 5% risk of making a type 1 error.
• Alternatively, α = 0.01 may be used, which means a 1% risk of making a type 1 error.

Step 4: Analyze the Data

• Use JAMOVI to find the p-value.
• If the p-value is greater than α, fail to reject the null hypothesis (Ho).
• If the p-value is less than α, reject the null hypothesis (Ho).

Comparing Two Means

One-Sample T Test

• No additional information provided.

Independent Samples T Test

• No additional information provided.

Dependent Samples T Test

• Used when comparing the difference between pre and post test scores or the difference between pairs.
• Assumptions:
  • Independent variable (IV) should be nominal with 2 groups/categories.
  • Dependent variable (DV) should be interval/ ratio (continuous).
  • Random sampling is done.
  • Distribution is normal (Shapiro-Wilk test).

Non-Parametric Test

• Used when data deviates from normal distribution.
• Report the median (Mdn) when using Mann-Whitney U.