Skip to main content
. 2022 Aug 5;13:901249. doi: 10.3389/fpsyg.2022.901249

Table 3.

Model comparisons of linear-mixed effect models predicting ISR scores.

Outcome variable Model AIC BIC logLik R 2 vs. null model vs. max model
χ2 p -value χ2 p -value
ISR sum score Null Model 1886.70 1901.60 −940.40 0.00
Max Model 1552.90 1622.50 −762.44 0.28 355.84 <0.001
Reduced Model 1548.00 1602.7 −763.01 0.28 354.71 <0.001 1.13 0.768
ISR depression Null Model 2803.10 2818.00 −1398.50 0.00
Max Model 2391.8 2456.50 −1182.90 0.34 431.24 <0.001
Reduced Model 2387.30 2447.0 −1181.70 0.34 428.23 <0.001 3.00 0.391
ISR anxiety Null Model 2679.92 2694.84 −1337.00 0.00
Max Model 2449.30 2504.00 −1213.60 0.21 246.65 <0.001
Reduced Model 2452.90 2487.80 −1219.50 0.20 234.99 <0.001 11.65 0.020

ISR = ICD-10-Symptom-Rating. The null model refers to a random-slope-only model. The max model refers to a model including all predictors and the reduced model only includes significant predictors from the max model.