Latent class analysis of ‘good citizenship’ forms among youth in Bosnia and Herzegovina

Figures & data

Table 1. Information criteria and statistical indices for models.

	M1 1 Class	M2 2 Class	M3 3 Class
Number of binary variables	7	7	7
LL	−3221.344	−3069.410	−3063.452
Npar	7	15	23
AIC	6456.687	6168.820	6172.905
BIC	6489.746	6239.659	6281.525
SSABIC	6467.516	6192.024	6208.485
Entropy	n/a	0.790	0.676
aLMRT	n/a	298.321	11.697
p-value	n/a	0.0000	0.0625
BLRT	n/a	303.868	11.915
p-value	n/a	0.0000	1.0000

Note: Sample size: n = 831; LL - Log-Likelihood values; Npar – Number of parameters; AIC - Akaike’s Information Criterion; BIC - Bayesian Information Criterion; SSABIC - Sample size adjusted Bayesian Information Criterion; aLMRT - Lo-Mendell-Rubin adjusted likelihood ratio test; BLRT - Bootstrap Likelihood Ratio Test.

Source: Author’s own calculation.

Table 2. Two classes of young citizens derived from the LCA and respective.

Indicator	Class 1 Enthusiastic citizens	Class 2 Outsiders
Voting (1 = regularly; 0 = otherwise)	0.66	0.40
News exposure (1= high degree; 0= low degree)	0.53	0.32
Contacting activities (1= Yes; 0= No)	0.47	0.00
Conventional participation (1= Yes; 0= No	0.64	0.03
Unconventional participation (1= Yes; 0= No)	0.65	0.07
Institutional trust (1 = high level; 0 = low level)	0.54	0.39
Political gender stereotype (1 = high level; 0 = low level)	0.61	0.52
Proportion of the sample (in %)	16.0%	84.0%

Source: Authors’ analysis.

Figure 1. Conditional probabilities of indicators of 'good citizenship': two-class model.

Source: Authors’ analysis.