Skip to main content
Fig. 5 | Health and Quality of Life Outcomes

Fig. 5

From: Comparison of statistical methods for the analysis of patient-reported outcomes (PROs), particularly the Short-Form 36 (SF-36), in randomised controlled trials (RCTs) using standardised effect size (SES): an empirical analysis

Fig. 5

Scatter plot of AIC for different statistical methods against the number of possible observable values of SF-36 domains in three RCT datasets. Note that median and CLAD regression do not have AIC scores, and thus are not compared in this figure. These methods are classified according to their distributional assumptions on SF-36 domain scores. AIC, Akaike information criterion; MLR, multiple linear regression; Tobit, Tobit regression; OL, ordered logit model, OP, ordered probit model, BB, beta-binomial regression; BLN, binomial-logit-normal regression; frac, fractional logistic regression; BR, beta regression

Back to article page

Health and Quality of Life Outcomes

ISSN: 1477-7525

Contact us