Archived Comments for:
Patient safety in surgical environments: Cross-countries comparison of psychometric properties and results of the Norwegian version of the Hospital Survey on Patient Safety
Confirmative factor analysis (CFA), not explorative factor analysis (EFA), would have been more appropriate to examined the applicability of facture structure of the Norwegian translation of the HSOPS
Kim Lyngby Mikkelsen, The National Board of Health, Denmark
13 October 2010
The authors used explorative factor analysis (EFA) to examine the applicability of the HSOPS factor structure in operating theatre settings. EFA is used to explore the possible underlying factor structure of a set of observed variables without imposing a preconceived structure on the outcome. By performing EFA, the underlying factor structure is identified. However, the factor structure of HSOPS is already known.
CFA is used to verify the factor structure of a set of observed variables. CFA allows the researcher to test the hypothesis that a relationship between observed variables and their underlying latent constructs exists. Therefore using CFA, the hypothesised facture structure of HSOPS could have been confirmed.
EFA or CFA, - does it matter? Using the original 12-factor model of the HSOPS the authors found low reliability scores for two factors, and their data seem to suggest that these two factors could be collapsed into one factor. In another sample data would possible suggest other modifications to the original 12-factor model of HSOPS. The point is, - EFA is simply not the most appropriate statistical tool to use here. Using CFA, it would have been possible to test if the data from the Norwegian translation of the HSOPS could confirm the original 12-factor model. If not the model misfit could have been examined, and eventually EFA could be used to hypothesise another facture structure if needed.
Further, two of the 12 factors of HSOPS are considered outcome variables. Therefore not only a CFA model but a structural equation model (SEM) could have been used to test both the 12-factor model and the hypothesised structural relationship between the factors.
The authors rightly conclude: “The psychometric properties of the questionnaire need further investigations to be regarded as reliable in surgical environments.”
Competing interests
No competing interests
Reply to comments: Confirmative factor analysis (CFA), not explorative factor analysis (EFA), would have been more appropriate to examined the applicability of facture structure of the Norwegian translation of the HSOPS
Arvid Steinar Haugen, Haukeland University Hospital
16 November 2010
On the behalf of the authors I would like to thank professor Kim Lyngby Mikkelsen for his thoughtful comments. In our work with exploring the safety climate factors of this surgical environment, we also understand the need for a structural equation model (SEM) test of the 12-factor model and the hypothesised structural relationship between the factors of the HSOPS. This is to be addressed in a follow up study of this surgical environment
Confirmative factor analysis (CFA), not explorative factor analysis (EFA), would have been more appropriate to examined the applicability of facture structure of the Norwegian translation of the HSOPS
13 October 2010
The authors used explorative factor analysis (EFA) to examine the applicability of the HSOPS factor structure in operating theatre settings. EFA is used to explore the possible underlying factor structure of a set of observed variables without imposing a preconceived structure on the outcome. By performing EFA, the underlying factor structure is identified. However, the factor structure of HSOPS is already known.
CFA is used to verify the factor structure of a set of observed variables. CFA allows the researcher to test the hypothesis that a relationship between observed variables and their underlying latent constructs exists. Therefore using CFA, the hypothesised facture structure of HSOPS could have been confirmed.
EFA or CFA, - does it matter?
Using the original 12-factor model of the HSOPS the authors found low reliability scores for two factors, and their data seem to suggest that these two factors could be collapsed into one factor. In another sample data would possible suggest other modifications to the original 12-factor model of HSOPS. The point is, - EFA is simply not the most appropriate statistical tool to use here. Using CFA, it would have been possible to test if the data from the Norwegian translation of the HSOPS could confirm the original 12-factor model. If not the model misfit could have been examined, and eventually EFA could be used to hypothesise another facture structure if needed.
Further, two of the 12 factors of HSOPS are considered outcome variables. Therefore not only a CFA model but a structural equation model (SEM) could have been used to test both the 12-factor model and the hypothesised structural relationship between the factors.
The authors rightly conclude: “The psychometric properties of the questionnaire need further investigations to be regarded as reliable in surgical environments.”
Competing interests
No competing interests
Reply to comments: Confirmative factor analysis (CFA), not explorative factor analysis (EFA), would have been more appropriate to examined the applicability of facture structure of the Norwegian translation of the HSOPS
16 November 2010
On the behalf of the authors I would like to thank professor Kim Lyngby Mikkelsen for his thoughtful comments. In our work with exploring the safety climate factors of this surgical environment, we also understand the need for a structural equation model (SEM) test of the 12-factor model and the hypothesised structural relationship between the factors of the HSOPS. This is to be addressed in a follow up study of this surgical environment
Competing interests
No competing interest