Skip to main content

Table 4 Bayesian Model Fit Statistics for Invariance Analyses

From: Measuring health literacy in community agencies: a Bayesian study of the factor structure and measurement invariance of the health literacy questionnaire (HLQ)

Scale Grouping Variable PPP Value 95 % CIs for the Difference between Observed and Replicated Chi-square values Range of PPP Values Across Individual Groups
1. Feeling understood and supported by healthcare providers Gender 0.54 −20.37–19.54 0.51–0.52
Age 0.29 −23.83–43.37 0.31–0.49
Education 0.26 −21.77–46.97 0.19–0.49
Home Language 0.26 −16.47–28.74 0.02–0.49
Organisation 0.14 −14.24─74.27 0.20─0.43
Comment: Strict invariance models were a satisfactory fit across all classifications and across all specific groups within these classifications.
2. Having sufficient information to manage my health Gender 0.37 −18.41–22.12 0.40–0.41
Age 0.38 −28.33–38.36 0.35–0.50
Education 0.30 −24.53─43.38 0.17–0.57
Home Language 0.53 −22.80–18.07 0.49–0.56
Organisation 0.12 −16.82–68.33 0.07–0.60
Comment: Strict invariance models were a satisfactory fit across all classifications and across all specific groups within these classifications
3. Actively managing my health Gender 0.23 −15.20–34.12 0.03–0.32
Age 0.09 −7.73–70.08 0.14–0.36
Education 0.22 −26.39–58.54 0.24–0.54
Home Language 0.46 −23.82–26.67 0.45–0.50
Organisation 0.34 −30.59–66.52 0.26–0.55
Comment: Strict invariance models were a satisfactory fit across all classifications and across all specific groups within these classifications.
4. Social support for health Gender 0.28 −17.35–32.76 0.39–0.35
Age 0.14 −18.41–60.05 0.22–0.45
Education 0.42 −33.56–44.89 0.34–0.65
Home Language 0.36 −21.46–25.62 0.47–0.35
Organisation 0.12 −20.47–79.84 0.10–0.58
Comment: Strict invariance models were a satisfactory fit across all classifications and across all specific groups within these classifications.
5. Appraisal of health information Gender 0.50 −24.74–23.86 0.47–0.54
Age 0.61 −41.51–29.24 0.41–0.65
Education 0.48 −36.51–39.77 0.32–0.61
Home Language 0.11 −10.01–43.82 0.08–0.47
Organisation 0.12 −25.04–85.85 0.15–0.42
Comment: Strict invariance models were a satisfactory fit across all classifications and across all specific groups within these classifications.
6. Ability to actively engage with healthcare providers Gender 0.15 −11.89–40.68 0.20–0.31
Age 0.11 −14.81–70.04 0.12–0.48
Education 0.09 −12.96–67.97 0.10–0.41
Home Language 0.18 −13.98–38.54 0.12–0.47
Organisation 0.02 2.52–107.55 0.07–0.52
Comment: Overall test of invariance across organisations marginally significant (indicating unsatisfactory fit) but fit was satisfactory in all groups separately. Satisfactory fit across all other classifications. Follow-up alignment analysis indicated that the loading of one item (‘Ask healthcare providers questions …’) was non-invariant in Organisation 1, being consistently higher in Organisation 1 than in other organisations.
7. Navigating the healthcare system Gender 0.12 −11.90–49.41 0.16–0.28 -
Age 0.01 15.64–111.02 0.02–0.41
Education 0.02 4.90–99.30 0.12–0.26
Home Language 0.13 −12.53–48.144 0.10–0.398
Organisation 0.01 24.17–144.22 0.02–0.33
Comment: Fit not satisfactory across organisation, education and age groups. Follow-up alignment analyses showed: (a) that the loading of one item (‘Work out what the best care is …’) was non-invariant in Organisation 6, being consistently higher in this organisation compared with the other organisations. Alignment analysis suggested full metric and scalar invariance across education and age groups.
8. Ability to find good health information Gender 0.34 −21.52–27.64 0.39–0.45
Age 0.12 −15.77–62.40 0.14–0.52
Education 0.00 21.38 - 104.91 0.00–0.35
Home Language 0.04 −3.82–52.07 0.03–0.30
Organisation 0.00 21.78–125.37 0.01–0.57
Comment: Fit not satisfactory across organisation, education and, marginally, home language groups. Follow-up alignment analyses showed that: (a) the loading of one item (‘Find information about health …’) was non-invariant in Organisation 4, being consistently lower in this organisation compared with all other organisations; and (b) the loading of one item (‘Get health information … you understand’) was non-invariant across the groups classified according to the language spoken at home, being higher in the group who spoke a language other than English at home. Alignment analysis suggested full metric and scalar invariance across the education groups.
9. Understanding health information well enough to know what to do Gender 0.31 −19.56–31.24 0.31–0.44
Age 0.24 −27.09–50.95 0.25–0.64
Education 0.022 2.15 - 79.32 0.10–0.38
Home Language 0.16 −13.47–42.17 0.36–0.16
Organisation 0.00 26.61–132.36 0.01–0.72
Comment: Fit not satisfactory across organisation and education level. Follow-up analysis showed that: (a) the loadings of two items, both of which referred to understanding information from healthcare providers, were non-invariant in Organisation 8 being consistently lower in Organisation 8 that in all other organisations; and (b) the loading of one item (‘Read and understand … medication labels’ was non-invariant in Education Group 2, being higher than in Groups 3, 4 and 5, but lower than in Group 1, suggesting that this item is most salient in the ‘Understanding health information well enough to know what to do’ factor in the groups with less formal education.