Skip to main content

Table 2 Latent class growth analysis and growth mixture modelling: comparisons of models

From: Perinatal depressive symptoms among low-income South African women at risk of depression: trajectories and predictors

Classes

BIC

AIC

Entropy

Size (%) of smallest class

LMRT statistic (p-value)

BLRT statistic (p-value)

Quadratic LCGAa

 2

7825.996

7770.687

0.742

26.0

203.838 (< 0.001)

− 3977.544 (< 0.001)

 3

7809.547

7738.435

0.750

4.6

38.630 (0.034)

− 3871.344 (0.030)

 4

7817.215

7730.301

0.718

4.6

15.483 (0.304)

− 3851.218 (0.288)

 5

7828.413

7725.696

0.584

3.9

12.097 (0.657)

− 3843.151 (0.647)

Quadratic GMM b

 2

7797.269

7722.207

0.816

8.6

38.789 (< 0.001)

− 3862.313 (< 0.001)

 3

7799.594

7708.730

0.807

4.2

20.612 (0.503)

− 3842.104 (0.488)

 4

7814.876

7708.209

0.762

1.6

8.177 (0.234)

− 3831.365 (0.229)

 5

7822.705

7700.235

0.771

1.3

15.330 (0.123)

− 3827.105 (0.115)

  1. AIC = Akaike Information Criterion, BIC=Bayesian Information Criterion; BLRT = Bootstrap Likelihood Ratio Test, LMRT = Lo-Mendell-Rubin Test; aLatent curve growth analysis; bgrowth mixture modelling
  2. Note: the final model selected is indicated in bold