TY - JOUR
T1 - Making robust decisions about the impact of health education programs
T2 - Psychometric evaluation of the Health Education Impact Questionnaire (heiQ) in diverse patient groups in Norway
AU - Wahl, Astrid K.
AU - Osborne, Richard H.
AU - Langeland, Eva
AU - Wentzel-Larsen, Tore
AU - Mengshoel, Anne Marit
AU - Ribu, Lis
AU - Peersen, Kari
AU - Elsworth, Gerald R.
AU - Nolte, Sandra
PY - 2016/10
Y1 - 2016/10
N2 - Objective To undertake a rigorous psychometric evaluation of the widely used eight-scale heiQ version 2.0 (evaluating immediate effects of self-management interventions) in diverse patient groups in Norway. Methods Cross-sectional survey data were collected from 1019 Norwegians. Data were extracted from studies among people with musculoskeletal disorders (n = 516), psoriasis (n = 254), heart disease (n = 97), and Type 2 diabetes (n = 152). To investigate the factorial validity of the Norwegian heiQ, confirmatory factor analyses (CFA) were carried out using Mplus. Results One-factor model fit, without modifications, was acceptable for the Emotional distress scale. Only one correlated residual was required to be fitted in each of the other scales to achieve satisfactory model fit. The postulated highly restricted full eight-factor model (no cross-loadings, no correlated residuals) showed good fit to the data. Internal consistency was acceptable for most scales (0.72–0.90) but low for Self-monitoring and insight. Conclusion This study of the Norwegian heiQ replicates the factor structure of the original Australian heiQ, using robust and highly restricted CFA procedures, demonstrating a clean independent clusters model structure. Practice implications: Researchers, program implementers and policymakers could use the Norwegian heiQ with confidence to generate reliable information on program outcomes and support quality improvement activities.
AB - Objective To undertake a rigorous psychometric evaluation of the widely used eight-scale heiQ version 2.0 (evaluating immediate effects of self-management interventions) in diverse patient groups in Norway. Methods Cross-sectional survey data were collected from 1019 Norwegians. Data were extracted from studies among people with musculoskeletal disorders (n = 516), psoriasis (n = 254), heart disease (n = 97), and Type 2 diabetes (n = 152). To investigate the factorial validity of the Norwegian heiQ, confirmatory factor analyses (CFA) were carried out using Mplus. Results One-factor model fit, without modifications, was acceptable for the Emotional distress scale. Only one correlated residual was required to be fitted in each of the other scales to achieve satisfactory model fit. The postulated highly restricted full eight-factor model (no cross-loadings, no correlated residuals) showed good fit to the data. Internal consistency was acceptable for most scales (0.72–0.90) but low for Self-monitoring and insight. Conclusion This study of the Norwegian heiQ replicates the factor structure of the original Australian heiQ, using robust and highly restricted CFA procedures, demonstrating a clean independent clusters model structure. Practice implications: Researchers, program implementers and policymakers could use the Norwegian heiQ with confidence to generate reliable information on program outcomes and support quality improvement activities.
KW - Chronic disease
KW - Confirmatory factor analysis (CFA)
KW - Health Education Impact Questionnaire (heiQ)
KW - Norway
KW - Self-management
UR - http://www.scopus.com/inward/record.url?scp=84969506821&partnerID=8YFLogxK
U2 - 10.1016/j.pec.2016.05.001
DO - 10.1016/j.pec.2016.05.001
M3 - Journal article
C2 - 27211224
AN - SCOPUS:84969506821
SN - 0738-3991
VL - 99
SP - 1733
EP - 1738
JO - Patient Education and Counseling
JF - Patient Education and Counseling
IS - 10
ER -
