Preliminary Findings from an Instrument Development Study to Measure Perceived Competence and Confidence of Clinical Nurse Educators

Friday, 24 July 2015: 2:10 PM

Van N. B. Nguyen, MN, BN, RN1
Mohamaddreza Mohebbi, PhD2
Thai Thanh Truc, MPH3
Maxine Duke, PhD, MEd, BAppSc (AdvNsg), RN1
Helen Forbes, PhD, RN1
(1)School of Nursing and Midwifery, Deakin University, Melbourne, Australia
(2)Faculty of Health, Deakin University, Melbourne, Australia
(3)Faculty of Health Sciences, University of Sydney, Sydney, Australia

Background: Clinical education is central to nursing education (American Association of Colleges of Nursing 2008). In the dynamic and complex clinical environment, clinical educators need to be prepared and supported in order to teach and facilitate students effectively (Gaberson and Oermann 2010). The transition of nurse clinician to the nurse educator role in Western countries however has been found to lack sufficient orientation, preparation and support (Aston et al. 2000, Cangelosi, Crocker and Sorrell 2009, Williamson, Webb and Abelson-Mitchell 2004). There is however little evidence regarding this transitional journey in developing countries. In addition, in several Asian countries, the educator role is undertaken by those moving from the student role and thus, the transition may present different challenges. In addition, to date, there is currently no valid and reliable scale to measure perceived competence and confidence following transition to the role of clinical nurse eduactor. 

Purpose: To develop a scale to measure clinical nurse educator perceived competence and confidence in clinical teaching. 

Methods: A structured two-phase approach has been employed to develop the Clinical Nurse Educator Skill Acquisition Assesment (CNESAA) instrument based on the platform of another questionnaire developed by Ramsburg and Childress (2012). Phase one (item identification, piloting, reliability and validity establishment and scale modification) will be discussed in this paper. 

Six-stage approach by Hair et al (2010) was chosen to guide the factor analysis using  SPSS version 20 software. “Partial least square structural equation modeling” (PLS-SEM) (Hair et al. 2014) rarely used in health was also used to compliment the understanding of factor analysis using SmartPLS software version 2.0. 

Results:

Demographic:

Five institutions offering undergraduate nursing degrees from both public and private sectors located in three distinct regions (North, Center and South) of Vietnam were included in this study. Of the 104 clinical nurse educators who participated in phase one, 78 were females (75%) and 26 were males (25%). A majority of the participants aged under 30 (n= 71, 68.3%). 90 participants (86.5%) had background in nursing and 14 (13.5%) majored in medicine or other disciplines in health.

Validity & reliability: Content validity was established by a panel of experts in nursing education. Reliability and convergent validity were established by two different statistical techniques:

Factor analysis:

From the pool of 38 items that were piloted, a conceptual and statistical model of an instrument with 24 items and five sub-scales was constructed. The instrument’s internal reliability was achieved with Cronbach’s alpha fluctuating from .828 to .920 for all sub-scales and .952 for the overall scale. Convergent and discriminant validity were established with high-loading items (ranging from .511 to 1.002), no cross-loadings items and low correlation between five subscales (<.7). The stability of the construct via an internal replication technique on two randomly split subsets of the original dataset indeed validated the factor analysis results.

PLS-SEM:

High internal reliability of the 24-item model was established with composite reliability values ranging from .899 to .938 for all five sub-scales. Outer loadings for all 24 items from .756 to .918 demonstrated indicator reliability. Convergent validity of the instrument was confirmed with Average Variance Extracted (AVE) criterion substantially above .5 (ranging from .645 to .748) for five subscales. The structural equational modeling analysis was validated with t-statistics greater than 3.29 (p<0.001, two-tailed test).

Further to statistical merit; expert ideas, content cohesion and participant voices were taken into consideration. Modification was made to simplify CNESAA’s format (labelling) and re-wording where necessary. The modified version of CNESAA with 24 items and five subscales (“Enhancing student learning”, “Relating theory and practice”, “engaging in scholarship”, “functioning as a leader”, and “participating in professional development”) is undergoing the second phase for validation purpose.  

Conclusion:

Through two different but complimentary approaches, the instrument was found to be conceptually and statistically reliable and valid to potentially measure perceived competence and confidence of nurse educators to teach nursing in clinical settings. An additional stage is to be conducted to confirm the purified version of the CNESAA. The structured, rigorous and robust approaches that were conducted to design and validate the CNESAA could be a recommendation for future high quality studies in instrument development in nursing.