| Term 
 
        | Parametric I/R Cross Over
 2
 |  | Definition 
 | 
        |  | 
        
        | Term 
 
        | Parametric I/R Cross Over
 3+
 |  | Definition 
 | 
        |  | 
        
        | Term 
 | Definition 
 | 
        |  | 
        
        | Term 
 | Definition 
 | 
        |  | 
        
        | Term 
 
        | Parametric I/R 3+
 Where is the difference?
 |  | Definition 
 
        | MCM (Multiple Comparison Method) |  | 
        |  | 
        
        | Term 
 
        | Parametric I/R Correct for portential confounders.
 |  | Definition 
 | 
        |  | 
        
        | Term 
 | Definition 
 | 
        |  | 
        
        | Term 
 | Definition 
 
        | McNemar's Chi-Square Chi-Square
 |  | 
        |  | 
        
        | Term 
 | Definition 
 
        | YCF (Yates Correction Factor for Chi-square) - Better than X2 when data is arranged in 2x2. FET (Fisher's Exact Test)-Used when total observations <40 or when <5 in a single "cell"
 Logistic Regression-NOT derived from chi-square
 |  | 
        |  | 
        
        | Term 
 | Definition 
 | 
        |  | 
        
        | Term 
 | Definition 
 
        | Chochran Mantel Haenszel Chi-square |  | 
        |  | 
        
        | Term 
 | Definition 
 
        | Makes the curve: Actuarial method (~meh)
 Kaplan-Meier Method (better)
 
 Analyze the curve:
 Cox-proportional Hazards
 Cochran Mantel Haenszel
 Log-rank (when censored observations occur)
 |  | 
        |  | 
        
        | Term 
 | Definition 
 
        | Mantel Haenszel Test Der-Simonian and Laird Method
 |  | 
        |  | 
        
        | Term 
 | Definition 
 
        | Prospective [a/(a+b)] / [c/(c+d)]
 >1 = outcome is increased
 <1 = outcome is decreased
 |  | 
        |  | 
        
        | Term 
 | Definition 
 
        | Retrospective (a*d)/(b*c)
 >1 = outcome is increased
 <1 = outcome is decreased
 |  | 
        |  | 
        
        | Term 
 
        | Correlation is 0 or 1 or -1 better?
 |  | Definition 
 
        | Closer to 1 or -1 means correlation. Cloers to 0 means not correlated
 |  | 
        |  | 
        
        | Term 
 | Definition 
 
        | Pearson's Product Moment 'r' |  | 
        |  | 
        
        | Term 
 | Definition 
 
        | Spearman Rank coefficient |  | 
        |  | 
        
        | Term 
 
        | Regression analysis Linear
 |  | Definition 
 
        | Provides mathematical relationship between variables: (y=mx + b)
 |  | 
        |  | 
        
        | Term 
 
        | Regression analysis Non-Linear
 |  | Definition 
 
        | Provides mathematical relationship between variables: Logistical Regression
 Log-linear analysis
 |  | 
        |  | 
        
        | Term 
 | Definition 
 | 
        |  | 
        
        | Term 
 | Definition 
 
        | (#new cases) / (#patients available) |  | 
        |  | 
        
        | Term 
 
        | ARR (Absolute Risk reduction) |  | Definition 
 
        | AR(control) - AR(exposure) or
 AR(control) - AR(exposure)
 |  | 
        |  | 
        
        | Term 
 
        | Ordinal/Non-normal IR Cross Over
 2
 |  | Definition 
 | 
        |  | 
        
        | Term 
 
        | Ordinal/Non-normal IR Cross Over
 3+
 |  | Definition 
 | 
        |  | 
        
        | Term 
 
        | Ordinal/Non-normal IR Parallel
 2
 |  | Definition 
 
        | Mann-Whitney U test 
 Provides U value that p-value is derived from
 |  | 
        |  | 
        
        | Term 
 
        | Ordinal/Non-normal IR Parallel
 3+
 |  | Definition 
 | 
        |  | 
        
        | Term 
 
        | Independent Cumulative Frequency test |  | Definition 
 | 
        |  | 
        
        | Term 
 
        | Cochran Mantel Haenszel Chi-Square |  | Definition 
 
        | Used in: 2x2 over time, survival analysis,Meta Analysis Can adjust for confounding factors
 |  | 
        |  |