Exam Notes: Hypothesis Testing, Contingency Tables, Descriptive Statistics

Hypothesis Testing: Randomization Test for Completion Rates

• Null Hypothesis (H0): There is no difference in completion rates between groups. Any observed difference is due to chance.

• Observed Difference Calculation:

  • Determined from a contingency table, with groups based on rows (e.g., suburban vs. urban locations).

  • Example: Suburban proportion ($355/963$) minus Urban proportion ($114/450$).

  • Calculated observed difference example: $0.1153$.

Setting up Simulations for Randomization Test

• Card Analogy: Cards or sticks are colored based on the total counts from the rows (e.g., total women in suburban clinics, total in urban clinics).

• Total Cards Needed: Equal to the total sample size ($1420$ in the example) for the entire population.

• **Number of