Skip to main content

Table 2 ICC and 95% CI for types of simulation datasets

From: Intraclass reliability for assessing how well Taiwan constrained hospital-provided medical services using statistical process control chart techniques

Types of datasets ICC    Data generation
    95% CI    
      Hospital Item  
    Lower Upper Measures Difficulty Approach
Raw data (13 continuous variables)
  Real data 0.999 0.999 0.999    
  Random data 0.040 −0.101 0.170 ~N(0,1) Randomized Sampling
XmR data (13 discrete variables with 5 categories)  
  1. Increasing 0.787 0.756 0.816 ~N(0,1) 2.0 to −1.0 Rasch model
  2. Decreasing 0.792 0.761 0.82 ~N(0,1) −2.0 to 1.0  
  3. Out-of control 0.791 0.76 0.819 ~N(0,1) = 0  
4. In-control 0.533 0.464 0.594 = 0 = 0  
  1 × 2 0.763 0.728 0.795 Combined in number of cases
  1 × 3 0.785 0.753 0.814    
  1 × 4 0.726 0.686 0.763    
  2 × 3 0.783 0.751 0.812 (50% vs. 50%)
  2 × 4 0.714 0.672 0.753    
  3 × 4 0.725 0.684 0.762    
  1. Note. √ = the supply of medical services is well-controlled and has a lower ICC.