Skip to main content

Table 3 Phase II (Steps 4 to 8) Establish the dimension structure

From: A scoping review to create a framework for the steps in developing condition-specific preference-based instruments de novo or from an existing non-preference-based instrument: use of item response theory or Rasch analysis

 

ABC-UI [31, 32]

AQL-5D [33,34,35,36,37,38,39]

IUI [40, 41]

OAB-5D [9, 36, 42,43,44]

EORTC-8D [7, 45,46,47,48]

QLU-C10D [49,50,51,52,53,54,55,56,57,58]

HAQ-PBM, QLQ-PBM, MSIS-PBM [61]

FACT-8D [62, 63]

CP-6D [66]

DEMQOL-U & -Proxy-U [8, 71,72,73,74]

AD-5D [75,76,77]

CARIES-QC-U [112]

DUI [78, 79]

DHP-3D & 5D [80]

HASMID-8 & 10 [81, 82]

DMD-QoL-8D [134, 138, 139]

Step 4. Assess factorability of items

 Item-total correlationsE, C, Pa

X

     

X

    

X

    

 Bartlett test of sphericityE, Pb

     

X

  

X

       

 Kaiser-Meyer Olkin measure proportion of varianceE, Pc

     

X

  

X

       

 Cronbach’s alphad

  

X

           

X

 

Step 5. Select dimension extraction method

 Principal components analysis (PCA)

X

X

X

X

            

 Exploratory factor analysis (EFA)

        

X

X

   

X

  

 Confirmatory factor analysis (CFA)

       

X

        

 EFA and CFA

     

X

    

X

 

X

  

X

Step 6. If there is no hypothesized dimensional structure, select the factors to retain

 Cattell scree testE, P

   

X

 

X

          

 Amount of variance explainedE, P

– eigen values

 

X

X

X

X

   

X

    

X

  

 Parallel analysisE, P

     

X

          

Step 7. Assess interpretability of dimension structure

 Oblimin rotation

     

X

          

 Promax rotation

        

X

       

 Varimax rotation

    

X

    

X

      

 Varimax and promax rotation

X

               

Step 8. Evaluate model fit

 RMSEA+

     

X

    

X

    

X

 SRMR+

                

 CFI++

     

X

    

X

    

X

 TLI++

     

X

    

X

     

 Factor loadings

   

X

X

X

X

   

X

 

X

X

  

 Residual correlations

     

X

          

 Cross loading

X

X

  

X

X

      

X

   
 

NEWQoL-6D [84,85,86]

CORE-6D [88,89,90]

ReQoL-UI [91,92,93,94]

MobQoL-7D [95, 140]

MSIS-8D & -P [27, 97,98,99,100]

Neuro-QoL derived NQU [101, 102]

P-PBMSI [103,104,105,106]

MF-8D [107]

WAITe [108, 109]

POS-E [116]

MHOM RA [117]

Vis-QoL [118,119,120,121,122,123,124]

VFQ-UI [125,126,127,128,129,130]

MacNew-7D [87]

ECOHIS-4D [114]

BUI [64]

CFQ-8D [69]

Step 4. Assess factorability of items

 Item-total correlationsE, C, Pa

        

X

      

X

X

 Bartlett test of sphericityE, Pb

   

X

     

X

   

X

X

  

 Kaiser-Meyer Olkin measure proportion of varianceE, Pc

   

X

     

X

   

X

X

  

 Cronbach’s alphad

   

X

      

X

X

  

X

X

 

Step 5. Select dimension reduction method

 Principal components analysis (PCA)

 

X

       

X

X

  

X

   

 Exploratory factor analysis (EFA)

X

X

 

X

X

  

X

   

X

  

X

 

X

 Confirmatory factor analysis (CFA)

  

X

            

X

 

 EFA and CFA

        

X

        

 Cattell scree testE, P

 

X

 

X

X

  

X

     

X

X

 

X

Step 6. If there is no hypothesized dimensional structure, select the factors to retain

 Amount of variance explainedE, P

– eigen values

  

X

X

X

        

X

X

 

X

 Parallel analysisE, P

        

X

X

   

X

X

  

Step 7. Assess interpretability of dimension structure

 Oblimin rotation

                 

 Promax rotation

 

X

 

X

      

X

      

 Varimax rotation

    

X

    

X

    

X

  

 Varimax and promax rotation

             

X

   

8. Evaluate model fit

 RMSEA+

  

X

        

X

   

X

 

 SRMR+

               

X

 

 CFI++

  

X

        

X

   

X

 

 TLI++

               

X

 

 Factor loadings

   

X

   

X

X

X

   

X

X

X

X

 Cross loading

                 

 Residual correlations

               

X

 
  1. E: exploratory factor analysis, P: principal components analysis, C: confirmatory factor analysis
  2. Factor loading: > 0.3 or > 0.4. + root mean square error of approximation (RMSEA) and standardized root mean squared residual (SRMR) < 0.08 acceptable, < 0.05 good, + + comparative fit index (CFI) and Tucker-Lewis index (TLI) > 0.9 acceptable, > 0.95 good
  3. aInter-item correlations range: 0.3–0.8
  4. bBartlett’s test of sphericity: Chi-sq, p-value < 0.01, < 0.05, confirm presence of correlations among items, to indicate whether EFA would be plausible
  5. cKMO: > 0.5, or > 0.8, close to 1.0
  6. dCronbach’s alpha: > 0.7 for group use, > 0.8 for individual use