This study aimed to translate and evaluate the psychometric properties of the Nurse Competence Scale in the Thai language and context.
Methods:
A cross-sectional descriptive design was used in this study. A convenience sample of 600 registered nurses (RNs) working in different areas of practice from one private and one public hospital in Chiang Mai province, Thailand were recruited. Demographic data were collected. The NCS was used to measure the nurse competence. The 73-item NCS consists of 7 components: helping role (7 items); teaching-coaching (16 items); diagnostic functions (7 items); managing situations (8 items); therapeutic interventions (10 items); ensuring quality (6 items); and work role (19 items).
The NCS was originally developed in Finish and translated to English by the developers using the Brislin’s back-translation approach. Data were analysed using Mplus 7 and SPSS version 20.0 (Chicago, IL). Using the Mplus 7 software, a confirmatory factor analysis (CFA) was performed to evaluate the factor model of the NCS. As the CFA did not confirm the original model structure of the NCS, then EFA with varimax rotation was performed. Parallel analysis was performed to determine the number of factors to retain. Further construct validity was established by the known-groups method. The Mann-Whitney U test was tested to identify difference in factor sores between groups classified according to work experiences. Internal consistency of the NCS was evaluated by determining the Cronbach’s alphas for the overall scale and subscales.
Results:
A total of 571 RNs completed the questionnaire. They comprised 6% males (n=34) and 94% (n=532) females, whose ages ranged from 24-59 years (M=38.52, SD=9.56. Regarding education 84% (n=476) had a 4-year nursing degree from a university or college, while 16% (n=91) were master-prepared nurses. Their work experience ranged from 3-37 years (M=14.69, SD=9.32).
Results from the confirmatory factor analysis revealed that the original structure of the NCS was not confirmed. Exploratory factor analysis was therefore considered appropriate for further analysis. After the initial EFA for the 73 items, 15 factors with eigenvalues >1.0 were retrieved. An inspection of the scree plot revealed that six or seven factors could be interpreted as just above the elbow of the curve. Parallel analysis was performed and the results revealed that the eigenvalues of six factors were larger than the 95th percentile in the distribution of eigenvalues derived from the random data. Applying these multiple factor extraction criteria, the final model of the NCS Thai-version consisted of 36 items that met the criterion of a factor loading of ≥0.40.
The first factor consisted of seven items. Five items were from the ensuring quality component in the original NCS and two items were from the therapeutic interventions component. Considering the items in this category, we labelled it as ‘research-oriented’. The second factor of the EFA contained 8 items of which seven were identical to the original items in the NCS work role component. Based on items included, this factor was similarly labelled the ‘work role’ component as on the original NCS. The third factor, labelled ‘diagnostic functions’, comprised six items from two of the original NCS components. In the fourth factor, there were seven items, four were from the original managing situations category in the original NCS whereas the other three items came from the therapeutic interventions. Based on the shared characteristics of each item, they were called ‘managing situations’. The fifth factor, ‘patient education’, contained 5 items. The original NCS teaching-coaching component consisted of items measuring competency in patient, family, and student teaching. However, in the Thai version, five items loaded in this factor contained only a patient education element. The last factor, ‘mentoring functions’, consisted of five items, all dealing with mentoring nursing students.
These six factors explained 58.45% of the variance. The factors of research-oriented, work role, diagnostic functions, managing situations, patient education, and mentoring functions accounted for 33.63%, 6.69%, 6.23%, 4.61%, 3.87%, and 3.42% of the variable variance, respectively.
The Mann-Whitney U test revealed significant differences between low and high-work experience groups for all 6 factors; nurses with low work experience showed lower scores for all 6 factors than those with high work experience (U = 22806.50, 20973.50, 25726.50, 27341.50, 24325.00, and 25844.00; P<0.05 for all associations).
The Cronbach’s alpha overall was 0.93 and for each competence component: 7-items research-oriented (0.88), 8-items work role (0.87), 6-items diagnostic functions (0.83), 7-items managing situations (0.82), 5-items patient education (0.82), and 3-items mentoring functions (0.86).
Conclusion:
This study was the first attempt to translate and validate the NCS into Thai. Results revealed that psychometric evaluation is crucial after the translation process, and that the 7-factor structure of the original NCS could not be confirmed. The EFA of the Thai-version of NCS led to 6 factors incorporating 36 items: research-oriented, work role, diagnostic functions, managing situations, patient education, and mentoring functions. This version also demonstrated promising psychometric properties, indicating that it might be useful for the assessment of nurse competence in Thailand. Cross-cultural adaptation of an instrument requires high methodological rigor and attempts should be made to further establish the construct validity of the NCS into the new language.