Skip to main content

Table 3 Summary of model performance for best OLS Tobit and TPM Item and domain models

From: Mapping the EQ-5D index from the cystic fibrosis questionnaire-revised using multiple modelling approaches

 

Observed EQ-5D

OLS models

Tobit models

TPM models

Domain (3)

Item (5)

Domain (3)

Item (5)

Domain (3)

Item (8)

Mean

0.671

0.671

0.671

0.672

0.672

0.691

0.679

SD

0.282

0.223

0.245

0.225

0.250

0.236

0.254

Range of values

- 0.349 - 1

- 0.099 - 1.04

- 0.165 - 1.183

- 0.053 - 0.985

- 0.167 - 1

- 0.221 - 0.985

- 0.200 - 1

ICC

 

0.715

0.801

0.716

0.811

0.717

0.820

RMSE

-

0.127

0.111

0.173

0.136

0.127

0.096

MSE

-

0.030

0.029

0.030

0.018

0.029

0.017

BIC

-

−214

383

45

626

256, − 201

574, 379

AIC

-

−254

−173

5

71

240, − 247

270, − 150

  1. OLS, ordinary least squares; TPM, two-part model; SD, standard deviation; Adj. R2, adjusted R2; RMSE, root of the mean square error; MSE, mean square error; AIC, Akaike information criterion; BIC, Bayesian information criterion.
  2. Model 3 = CFQ-R domains that are statistically significant at the 10% level + statistically significant squared terms; Model 5 = All CFQ-R items excluding the health domain items; TPM Model 8 = All CFQ-R items excluding the health domain items + age and gender.