Skip to main content

Table 1 Summary of statistical methods used for the psychometric evaluation of the QoR-15NL scale

From: Validation of the Dutch translation of the quality of recovery-15 scale

Concept Parameter Abbreviation Interpretation  
Construct validity
  Correlation Spearman's rho rs moderate 0.40—0.69
    strong 0.70—0.89
    very strong 0.90—1.00
Reliability
  Internal concistensy Crohnbach's alpha   good 0.70—0.90
  Split-half reliability   unaccaptable  < 0.70
    fair 0.70—0.79
    good 0.80—0.89
    excellent  ≥ 0.90
Responsiveness
  Cohen's effect size   large  ≥ 0.8
  Standardised response mean SRM large  ≥ 0.8
Reproducibility
  Agreement Individual smallest detectible change SDC good SDC < minimal clinically
     important difference
  Test–retest reliability intraclass correlation coefficient ICC good  > 0.7
  (two-way random effect model)    
Clinical feasibility
  Floor effect Percentage of patients with   present if > 15%
  lowest possible score    
  Ceiling effect Percentage of patients with   present if > 15%
  highest posible score