Table 4 Measurement invariance (configural, metric, scalar, and residual) comparison for all study models over three distinct time intervals (P1-P2).

From: The mediating role of cognitive arousal in the relationship between impulsivity and sleep quality among college students: a random intercept cross-lagged panel analysis

 

Fit Indices

Model Comparison

 

CMIN/df

RMSEA

SRMR

CFI

TLI

\(\:\triangle\)CFI

\(\:\triangle\)RMSEA

Impulsivity

Configural

1.31***

0.000

0.015

1.000

0.965

  

Metric

1.21***

0.000

0.016

1.000

0.958

0.000 a

0.000 a

Scalar

1.28***

0.000

0.022

1.000

0.960

0.000 b

0.000 b

Residual

1.35**

0.000

0.023

1.000

0.966

0.000 c

0.000 c

Cognitive arousal

      

Configural

1.42**

0.026

0.014

0.991

0.972

  

Metric

1.31**

0.032

0.024

0.992

0.967

0.001 a

0.006 a

Scalar

1.25***

0.022

0.027

0.989

0.970

−0.003b

−0.001 b

Residual

1.20***

0.025

0.019

0.990

0.969

0.001 c

0.003 c

Sleep Quality

      

Configural

1.46**

0.048

0.030

0.989

0.970

  

Metric

1.39**

0.041

0.033

0.990

0.968

0.001 a

−0.007 a

Scalar

1.29***

0.036

0.035

0.989

0.966

−0.001 b

−0.005 b

Residual

1.18***

0.032

0.033

0.987

0.963

−0.002 c

−0.004 c

  1. Note: As the chi-square statistic for assessing model fit is sensitive to sample size, we compared the CFIs (< 0.01) and RMSEAs (< 0.015) of the models82. a Comparison of metric invariance and configural invariance models. b Comparison of scalar invariance and metric invariance models. c Comparison of residual invariance and scalar invariance models. *p < 0.05. **p < 0.01. ***p < 0.001.
Back to article page