Please enable JavaScript.

Coggle requires JavaScript to display documents.

Calculating Sample Size, Proportion, %, Standardized Gain Score, chi…

- - - - usually take as 80%
    - - sample size
      - effect size
- - - - assume p= 0.5 and q=0.5
        
        P=0.5 is the maximum variability (proportion) for the population
- - - - Pilot Study
      - Previous study (eg: literature review)
      - ES Conventions (as suggsted by Cohen)
    - - Measure the magnitude of effect.
      - a way of quantifying the difference between two groups
      - eg: if one group has an experimental tx, and the other not (control), then ES is a measure of the effectiveness of the treatment
    - - depends on specific objectives
        
        Standardized Mean Difference
        
        comparing group research
        
        either
        
        treatment group
        
        naturally occurring group
        
        Inherently cotinuous construct (continuous outcome)
        
        Odds-Ratio
        
        comparing group research
        
        either
        
        treatment groups
        
        naturally occurring group
        
        inherently dichotomous construct (categorical outcome-rare)
    - - each "family" of inferential statistics has specific ES index
        
        t test : Cohen's d
        
        standardized mean difference
        
        formula
        
        size of effect:
        
        small: 0.2
        
        medium: 0.5
        
        large : 0.8
        
        F test : eta squared, Cohen's f
        
        eta squared size:
        
        small : 0.01
        
        medium: 0.06
        
        large: 0.14
        
        Cohen's f
        
        size:
        
        Small: 0.10
        
        Medium: 0.25
        
        Large: 0.40
        
        Correlation : r
        
        based on correlation coefficient (r)
        
        associatiom between variable research
        
        strength of association
        
        size:
        
        Small: 0.1
        
        Medium: 0.3
        
        Large : 0.5
        
        Regression: Cohen's f²
        
        size
        
        Small: 0.02
        
        Medium: 0.15
        
        Large: 0.35
        
        Chi square: Cohen's w
        
        size
        
        Small: 0.1
        
        Medium: 0.3
        
        Large: 0.5