Skip to main content

Table 2 Latent Class Growth Analysis (LCGA) based on the total sample

From: Leisure-time physical activity trajectories from adolescence to adulthood in relation to several activity domains: a 27-year longitudinal study

No. of classes

AIC

BIC

BLRT

VLMR

Entropy

Average Latent Class Probabilities for Most Likely Latent Class Membership (%)

Sample Size Per Class Based on Most Likely Class Membership

The Number of Random Start Values and Final Iterations

1

29,901.46

29,971.54

––

––

––

100

1103

200, 20

2

28,520.17

28,615.28

p < .05

p < .05

0.773

92/94

324/779

200, 20

3

28,248.55

28,368.69

p < .05

p = .07

0.690

89/77/85

612/342/149

200, 20

4

28,090.97

28,236.14

p < .05

p < .05

0.703

84/78/87/75

102/276/597/128

200, 20

5

27,997.88

28,168.08

p < .05

p = .20

0.664

77/72/71/82/87

122/292/169/471/49

200, 20

6

27,928.51

28,123.74

p < .05

p = .61

0.664

81/85/70/71/72/86

483/46/268/121/141/44

200, 20

7

27,871.38

28,091.64

p < .05

p = .11

0.653

87/84/78/74/72/71/62

41/34/432/166/62/190/178

200, 20

  1. AIC Akaike information criterion, BIC Bayesian information criterion, BLRT Bootstrap likelihood ratio test, VLMR Vuong-Lo-Mendell-Rubin test
  2. The class solution considered optimal is presented in bold