Skip to main content

Table 2 Summary of latent class model fit for with and without sampling weights (indices suggesting best model fit are marked bold) of a population-based cross-sectional survey in Italy, August to September 2014

From: Health information-seeking behavior associated with linguistic group membership: latent class analysis of a population-based cross-sectional survey in Italy, August to September 2014

     

Latent class sizes based on modal assignment

No. of classes

AIC

BIC

LMR

adj. LMR

LC1

LC2

LC3

LC4

LC5

Without sampling weights

 1

5918.6

5960.8

-

-

504

-

-

-

-

 2

5656.4

5745.1

 < .0001

 < .0001

182

322

-

-

-

 3

5587.0

5722.1

0.010

0.011

156

107

241

-

-

 4

5554.8

5736.4

0.213

0.219

136

111

206

51

-

 5

5539.8

5767.9

0.603

0.607

85

144

62

64

149

With sampling weights

 1

6041.1

6083.3

-

-

504

-

-

-

-

 2

5836.2

5924.9

 < .001

 < .001

222

282

-

-

-

 3

5771.9

5907.0

0.203

0.209

197

103

204

-

-

 4

5736.3

5917.9

0.749

0.749

127

125

142

110

-

 5

5714.7

5942.7

0.742

0.743

86

122

58

106

131

  1. AIC Akaike information criterion, BIC Bayes information criterion, LMR P-value of the Lo-Mendel-Rubin test, adj. LMR p-value of the adjusted LMR