The multidimensionality of caring: a confirmatory factor analysis of the Caring Nurse–Patient Interaction Short Scale
Version of Record online: 21 FEB 2008
© 2007 The Authors
Journal of Advanced Nursing
Volume 61, Issue 6, pages 699–710, March 2008
How to Cite
Cossette, S., Pepin, J., Côté, J. K. and De Courval, F. P. (2008), The multidimensionality of caring: a confirmatory factor analysis of the Caring Nurse–Patient Interaction Short Scale. Journal of Advanced Nursing, 61: 699–710. doi: 10.1111/j.1365-2648.2007.04566.x
- Issue online: 21 FEB 2008
- Version of Record online: 21 FEB 2008
- Accepted for publication 23 August 2007
- Caring Nurse–Patient Interaction Short Scale;
- confirmatory factor analysis;
- instrument development;
- nurse–patient interaction;
- nurse–patient relationships;
- nursing theory
Title. The multidimensionality of caring: a confirmatory factor analysis of the Caring Nurse–Patient Interaction Short Scale
Aim. This paper is a report of a study to evaluate the construct validity of the four-dimensional Caring Nurse–Patient Interaction–Short Scale using confirmatory factor analysis.
Background. Validating theoretical structures of caring is an ongoing challenge in the discipline of nursing. Our previous work has contributed to this literature by the exploration of the dimensionality of the Caring Nurse–Patient Interaction Short Scale via an exploratory factor analysis. The Caring Nurse–Patient Interaction Short Scale comprises 23 items reflecting four caring domains: humanistic care, relational care, clinical care and comforting care.
Method. A methodological study was conducted involving a convenience sample of 531 nursing students in a baccalaureate nursing programme (20% were already Registered Nurses). Data were collected in 2002 and 2004. Confirmatory factor analysis of the Caring Nurse–Patient Interaction Short Scale was performed.
Findings. As expected with large samples and models, the chi-squared-associated P-value was statistically significant (χ2 = 811·43, d.f. = 224, P < 0·01). However, the other indices reached acceptable levels with 0·054 for the standardized root mean-squared residuals, 0·070 for the root mean-square error of approximation, 0·88 for the goodness of fit index, 0·98 for the comparative fit index and 0·97 for the normal fit index. The factor loadings for all items with their hypothesized factor were ≥0·48 and statistically significant at the 0·01 level.
Conclusion. The Caring Nurse–Patient Interaction Short Scale model was judged to fit the data adequately. Although further testing of the scale with different samples of patients is warranted, our model emerged as a middle-range theory during the construct validity process and still reflects Watson’s theory while offering a structure that is testable in clinical research.