3.2.2. Confirmatory Factor Analysis

The CFA confirmed the EFA model with eight variables. Standardised and nonstandardised solutions are presented respectively in Figures 3 and 4. A convergent validity the strength of relationships of the model's factor variables is not supported by the average variance extracted (AVE), which has to be greater than 0.5. The AVE for accessibility equals 0.461 and AVE for effectiveness equals 0,401. This confirmed the low values of the Cronbach's alpha coefficients calculated during the EFA. Nevertheless, the convergent validity might be confirmed using the composite reliability (CR) index, which should be higher than 0.7. CR values for accessibility and effectiveness equal 0.771 and 0.722, respectively, which means that the convergent validity of the model is confirmed.

**Figure 3.** CFA model for the treatment effectiveness and the access to healthcare services during the COVID 19 pandemic (standardised estimates).

**Figure 4.** CFA model for the treatment effectiveness and the healthcare services accessibility during the COVID 19 pandemic (non-standardised estimates).

The model's discriminant validity is also confirmed using the Fornell Larker criterion, since the square root of the AVE for each factor is higher than the correlation between factors (Table 7). Also, the hetero trait − mono trait method points to a correlation between factors of 0.539, which should be smaller than 0.85. Taking into account the above results, it is possible to confirm the model's reliability and validity.


**Table 7.** Correlations between accessibility and effectiveness.

\*\*. Correlation is significant at the 0.01 level (2-tailed).

The model fit measures indicate that the model is correct. CMIN = 20.918; DF = 19.000. CMIN/DF = 1.101 (>1; <3), CFI = 0.989 (>0.95); SRMR = 0.054 (<0.08); RMSEA = 0.032 (<0.06); *p*-value = 0.605 (>0.05).
