Skip to main content

Table 4 Loadings and thresholds for measurement invariant model, 2018

From: Measuring public knowledge on nuclear weapons in the post-Cold War: dimensionality and measurement invariance across eight European countries

  Unidimensional Bifactor
  General Threshold General Specific 1 Specific 2 Specific 3 Threshold
United States 0.90 − 0.97 0.94     − 0.93
Russia 0.88 − 0.93 0.92     − 0.89
China 0.79 − 0.41 0.82     − 0.37
Israel 0.74 0.11 0.67 0.51    0.07
India 0.84 0.47 0.78 0.51    0.43
Pakistan 0.72 0.38 0.65 0.51    0.34
Fire 0.60 − 0.34 0.56   0.62   − 0.58
Blast 0.65 − 0.58 0.62   0.62   − 0.82
Hiroshima 0.78 − 1.03 0.75    0.57 − 0.85
Nagasaki 0.76 − 0.47 0.73    0.57 − 0.28
  1. Note: Parameters for models with invariant thresholds, loadings, and intercepts, with the scale of latent variates y* fixed to unit. Right-side panel reports parameters for the bifactor model in Table 3, model B2. Unidimensional model with invariant parameters included in the left-side. All parameters are statistically significant at p > 0.01