Skip to main content

Table 3 Model-fit statistics comparisons

From: Measurement feedback system implementation in public youth mental health treatment services: a mixed methods analysis

Classes

AIC

BIC

aBIC

Vuong-Lo-Mendell LR test p-value

Lo-Mendell adjusted LR test p-value

Entropy

N per class

2

2346.34

2572.63

2273.06

0.26

0.26

0.98

I = 40

II = 40

3

2182.97

2523.60

2072.67

0.58

0.59

0.99

I = 32

II = 19

III = 29

4

2055.08

2510.04

1907.75

0.39

0.39

0.99

I = 6

II = 30

III = 25

IV = 19

  1. AIC, BIC, and aBIC are relative fit statistics, with smaller values indicating better comparative fit. Entropy measures class separation with entropy ≥ 0.80 indicating acceptable class distinction. The Vuong-Lo-Mendell LR test and the Lo-Mendell adjusted LR test provide p values for use in assessing if adding additional classes will result in a statistically significant improvement in fit [75]
Back to article page

Implementation Science Communications

ISSN: 2662-2211