Common Research Scenarios
Real research questions matched to the appropriate statistical test.
🔬 Research Questions → Tests
Scenario 1: Coffee and Reaction Time
Question: Does drinking coffee affect reaction time compared to no coffee?
Setup: - Two groups: Coffee vs No Coffee - Outcome: Reaction time (continuous) - Different people in each group
Test: Independent t-test
See: T-Tests
Scenario 2: Teaching Methods
Question: Do three teaching methods produce different test scores?
Setup: - Three groups: Lecture, Discussion, Flipped - Outcome: Test scores (continuous) - Different students in each group
Test: One-Way ANOVA
See: ANOVA
Scenario 3: Before and After Training
Question: Did scores improve after the training program?
Setup: - Same people measured twice: Before and After - Outcome: Skill scores (continuous) - Paired measurements
Test: Paired t-test
See: T-Tests
Scenario 4: Caffeine × Time of Day
Question: Do caffeine AND time of day both affect performance?
Setup: - Two factors: Caffeine level (None/Low/High) × Time (Morning/Evening) - Outcome: Performance score (continuous) - Testing main effects and interaction
Test: Two-Way ANOVA
See: ANOVA
Scenario 5: Study Hours → GPA
Question: Can I predict GPA from study hours?
Setup: - Predictor: Study hours per week (continuous) - Outcome: GPA (continuous) - Want to predict one from the other
Test: Simple Linear Regression
See: Regression
Scenario 6: Gender × Major Choice
Question: Is gender related to major choice?
Setup: - Two categorical variables: Gender × Major - Count data (frequencies) - Testing association
Test: Chi-Square Test of Independence
See: Chi-Square
Scenario 7: Drug A vs Drug B vs Placebo
Question: Which treatment is most effective?
Setup: - Three independent groups - Outcome: Symptom severity (continuous) - Want to know if any differ
Test: One-Way ANOVA + post-hoc tests
See: ANOVA
Scenario 8: Stress Over Semester
Question: Does stress change across beginning, middle, and end of semester?
Setup: - Same students measured 3 times - Outcome: Stress score (continuous) - Repeated measurements
Test: Repeated Measures ANOVA
See: ANOVA
Scenario 9: Non-Normal Data, Two Groups
Question: Do meditation and control groups differ on anxiety?
Setup: - Two independent groups - Outcome: Anxiety scores (highly skewed, non-normal) - Assumptions violated
Test: Mann-Whitney U Test
See: Non-Parametric
Scenario 10: Multiple Predictors
Question: Can I predict burnout from workload, social support, and sleep quality?
Setup: - Multiple predictors (all continuous) - One continuous outcome - Want to see unique contribution of each
Test: Multiple Regression
See: Regression
🎯 Quick Pattern Recognition
| Your Situation | Pattern | Test |
|---|---|---|
| 1 group vs known value | Compare to standard | One-Sample t-test |
| 2 independent groups | Group comparison | Independent t-test |
| Same people, 2 times | Before/after | Paired t-test |
| 3+ independent groups | Multiple group comparison | One-Way ANOVA |
| 2 factors | Factorial design | Two-Way ANOVA |
| Same people, 3+ times | Longitudinal | Repeated Measures ANOVA |
| Continuous predictor → continuous outcome | Prediction | Regression |
| 2 categorical variables | Association | Chi-Square |
| Height ↔ Weight | Relationship strength | Correlation |