Abstract
This study aimed to adapt and validate a Korean version of the Moral Identity Scale developed by Aquino and Reed. Intensive care unit (ICU) nurses face several moral conflict or challenging situations where they make difficult decisions related to the rationing of care or prioritization such as a shortage of medical supplies and staffing. To understand the reason behind choosing moral values over other values such as economic and personal ones by nurses during such times, conceptual research on moral identity is required. However, this has not been studied using empirical data analysis despite the increased significance of nurses’ moral identity. A secondary data analysis and a methodological design were employed. Data were collected from 207 ICU nurses. Exploratory and confirmatory factor analyses were used to test the construct validity. The results of exploratory factor analysis showed that the Eigenvalues ranged from 1.63 to 4.47 and comprised 52.17% of the total explained variance. Confirmatory factor analysis showed acceptable model fit indices (χ2 [p] = 28.822 [.051], df = 18, root mean square error of approximation = 0.076, goodness-of-fit index = 0.937, Tucker–Lewis index = 0.93, and comparative fit index = 0.955) and standardized factor loadings (0.45–0.82). As professionals, ICU nurses must protect and advocate for the patient according to their own moral identity. Therefore, the Korean version of the Moral Identity Scale is a valid and reliable instrument that can be used in nursing education programs for improving the moral identity of ICU nurses.
Introduction
Nursing is a profession that provides services centered on human beings, and all situations related to individuals can involve ethical and moral problems (Kim, 2015). Therefore, nursing practice is inherently ethical (Deschenes & Kunyk, 2020). Nurses frequently encounter challenging situations that require difficult decision-making, particularly in scenarios involving care rationing or prioritization. These situations often arise due to factors like shortages in medical supplies and staffing. When caring for patients, nurses may encounter complex scenarios that entail legal and ethical responsibilities (Kleemola et al., 2020). For example, as medical supplies became scarce owing to the coronavirus disease 2019 (COVID-19) pandemic, medical personnel experienced a moral dilemma regarding whom to treat first (Roger et al., 2020). During the early stages of the COVID-19 pandemic, some nurses went on strike, advocating for border closures to prevent the spread of COVID-19 from other countries (Chan, 2020). This event is relevant to nursing moral identity as it highlights the ethical dilemmas nurses face when balancing patient care with broader public health concerns. Conversely, in February 2020, when the medical system collapsed due to the exceeding capacity of patients with COVID-19, volunteer nurses from all over the country went to Daegu City in South Korea to treat patients (Choi, 2020). When nurses had to select between their safety and patients’ health, they seem to select morally valuable behavior. According to Aquino and Reed (2002), the higher a person’s “moral identity,” the higher the likelihood of moral behavior (Aquino & Reed, 2002).
Moral identity connotes a commitment consistent with one’s sense of self to lines of action that support or protect the welfare of other individuals and defined as a self-concept organized around a set of moral traits (Aquino & Reed, 2002; Lapsley & Narváez, 2004). In other words, the greater the strength of an individual’s moral traits that define his/her moral identity, the more probable it is for that identity to be applied in a broad spectrum of situations, thereby increasing the association between moral identity and moral behavior (Aquino & Reed, 2002). Therefore, “moral identity” is a significant concept for nurses who have a professional responsibility to act morally. Particularly, “moral identity” is a significant concept for intensive care unit (ICU) nurses who have conflicting positions among various stakeholders, including medical staff, caregivers, and family members, and may have work difficulties arising owing to various moral pains and frustrations of terminally ill patients (Henrich et al., 2017; Lluch-Canut et al., 2020; Mills & Cortezzo, 2020).
Several empirical studies have examined moral identity in the general population, including the relationship between adolescents’ cyberbullying behavior and their moral identity (Wang, 2019), between incentive manipulation and moral identity in the business context (Aquino et al., 2009), and between stigma and moral identity in the advice of debtors (Andelic et al., 2019). However, despite the increased significance of nurses’ moral identity, it has not been studied using empirical data analysis. Additionally, finding an appropriate instrument for measuring moral identity in Korean nurses is challenging. Generally, moral identity is measured using Aquino and Reed’s (2002) Moral Identity Scale (MIS). Since the scale was developed within a Western cultural context, adapting and validating a new scale for measuring “moral identity” in Korean nurses is necessary.
Theoretical Background
In this study, to enumerate moral characteristics and estimate the degree of moral identity of ICU nurses, the trait approach was used. This approach is a way of examining character (morality), emphasizing the individual as a marker of personality (Deary, 2009). Several researchers have discussed the reasons why nurses select moral values as professionals and what characteristics they have (Aquino & Reed, 2002; Killen, 2002; Krishnasamy, 1999). Researchers have proposed the concept of moral identity, which is typically based on the appraisal that an action has moral worth (Lapsley & Narváez, 2004). Furthermore, the American Association of Critical Care Nurses (AACN, 2016) has outlined six essential standards for creating a healthy ICU work environment. These standards include skilled communication, true collaboration, effective decision-making, appropriate staffing, meaningful recognition, and authentic leadership (AACN, 2016). Authentic leaders must possess qualities like confidence, hope, and moral values (Luthans & Avolio, 2003). Therefore, there is a correlation between moral identity and authentic leadership.
The MIS indicates how much a person values the overall image of a moral person. According to Aquino and Reed (2002), the concept of moral identity is divided into the following two components: internalization and symbolization. Internalization refers to the integration of moral values into an individual’s self-concept or identity, whereas symbolization refers to the use of moral identity as a tool for attempting moral action and the degree to the ego based on the moral character expressed externally. According to their theory, two components interact with each other to form a moral identity. The MIS has the advantage of being able to simultaneously measure both “What value I prefer and pursue” and “How do I want to be reflected to others?”
Therefore, when the MIS is used for Korean ICU nurses, it needs to validate the version of moral identity for ICU nurses.
Aim of the Study
This study aimed to adapt and validate a Korean version of the MIS (K-MIS) for ICU nurses in the Korean cultural context.
Methods
Study Design
A secondary analysis of data from the original study on the predictive factors of turnover intention among ICU nurses was done (Lee & Song, 2018). In this study, to test the validity and reliability of the K-MIS, a methodological design was employed. Specifically, to assess the construct validity of the K-MIS, exploratory factor analysis (EFA) and confirmatory factor analysis (CFA) were performed. Convergent validity was assessed through the correlation between the K-MIS and the Authentic Leadership Inventory. Additionally, to evaluate the internal consistency and stability of the K-MIS, a reliability test was performed.
Study Setting
In this study, the participants were a non-probability convenience sample (n = 207) of ICU nurses recruited from four university hospitals with more than 500 beds located in the southeastern region of Korea. Each hospital has two or more ICUs, including internal medicine and surgery ICUs, and approximately 20 to 40 nurses work in each ICU. A total 120 ICU nurses were recruited from hospital A, 38 from hospital B, 31 from hospital C, and 18 from hospital D. Of the 230 questionnaires distributed, 209 were returned; however, two questionnaires lacked responses. Therefore, 207 questionnaires were finally analyzed.
Instruments
MIS
In order to examine the relationship between moral identity, moral cognitions, and the behavior of individuals across different age groups, we employed the MIS developed and validated by Aquino and Reed (2002) in the United States. The MIS examines participants to consider a person with characteristics often associated with moral probity (e.g., fair, kind, caring, and honest). Subsequently, it examines if they “try to answer questions by imagining how people with these characteristics will think, feel, and act.” The MIS consists of 10 items with the following two factors: internalization (five items) and symbolization (five items). The seven-point Likert-type scale ranges from strongly agree to strongly disagree, with higher scores representing higher moral identity. Some examples of items are “Being someone who has these characteristics is an important part of who I am (internalization)” and “The kinds of books and magazines that I read identify me as having these characteristics (symbolization).” The scores of two items were reversed as follows: items numbered 4 and 7 for internalization. Cronbach’s α for the internalization and symbolization in this study were .83 and .82, respectively (rα = .86).
At the time of scale development, an EFA of 10 items for 363 business school students in three universities in the United States showed that the cumulative explanatory power was 71.5% (Aquino & Reed, 2002). CFA was performed on 347 adults graduating from Delaware’s college of business and economics. Therefore, cross-validation was secured. At the time of scale development, the results of CFA showed χ2 = 205.96 (p < .001), root mean square error of approximation (RMSR) = 0.04, comparative fit index (CFI) = 0.87, goodness-of-fit index (GFI) = 0.93, and normed fit index = 0.84 (Aquino & Reed, 2002).
Authentic Leadership Inventory (ALI)
To confirm the convergence validity of the MIS, the concept of authentic leadership, which consistently showed a positive correlation with moral identity (r = .51, p < .001, r = .55, p < .001), was selected (Olsen & Espevik, 2017; Zhu et al., 2011). Authentic leadership has recently attracted attention as a concept that values the morality of a leader; it is a “root concept,” forming the basis of the positive aspects of charismatic, transformational, spiritual, and ethical leadership theories (Ilies et al., 2005). Luthans and Avolio described an authentic leader as someone who is “confident, hopeful, optimistic, resilient, moral/ethical, future-oriented, and provides priority for developing associates to be leaders. An authentic leader is someone who is true to him/herself” (Cameron et al., 2003). This study employed the ALI developed by Neider and Schriesheim (2011), which consists of 14 items. The five-point Likert-type scale ranges from strongly agree to strongly disagree, with higher scores representing higher authentic leadership.
Procedure
Translation and Back Translation Process
Approval for use and adaptation was received by email from the scale developer, Aquino and Reed (2002), and the scale was subsequently translated. The scales were originally written in English; therefore, they were first translated by a Korean expert who is fluent in English at Seoul National University’s Language Education Institute. According to the process of translation and adaptation of instruments guidelines of the World Health Organization (WHO, n.d.), the primary translator should know the English-speaking culture; however, their native language should be the primary language of the target culture. Thus, the primary translator was a Korean who graduated from an American university. The translator was satisfied with the forward translator criteria of the WHO (WHO, n.d.).
The primary translation was revised and supplemented by finding inappropriate expressions or concepts with the advice of a bilingual nursing professor who was fluent in English. The created Korean version was sent to Seoul National University’s Language Education Institute and was subsequently translated back into English by a bilingual translator, whose first and second languages were English and Korean, respectively. This secondary translator was an independent translator with no prior knowledge of the questionnaire. The focus of this translation was on conceptual and cultural identity rather than linguistic identity. The final translation was completed following examining the sentence structure and similarity while comparing the reverse translation result with the original scale.
Feasibility Test
To assess the comprehension and suitability of the questionnaire and review the fluency, readability, and comprehensibility of its items, a feasibility study of the final draft was performed with four Korean-speaking nurses in ICU who were not included in the validation sample. ICU nurses reported no major difficulties. The respondents completed the questionnaires in approximately 3 to 5 min. Minor revisions were made to the K-MIS without any change in the conceptual meanings, and this resulted in the final version of the K-MIS questionnaire.
Data Analysis
Analyses were performed using IBM SPSS Statistics version 26.0 and AMOS version 23.0. Descriptive statistics (i.e., mean, SD, frequency, and percentage) were used to summarize the data. Internal consistency was estimated using Cronbach’s α. EFA and CFA were used to examine the construct validity. To ensure data adequacy for the factor analyses, the Kaiser–Meyer–Olkin (KMO) test and Bartlett’s test of sphericity were used. EFA was extracted through the common factor analysis with direct oblimin. To evaluate the factor structure for adequate model fit with multiple fit indices, CFA was performed. A range of goodness-of-fit criteria for assessing the overall model fit was used, including GFI > 0.95, RMSR < 0.06, Tucker–Lewis index (TLI) > 0.95, and CFI > 0.95 (Hu & Bentler, 1999). Moreover, the convergent validity was estimated by examining Pearson correlations between the K-MIS and authentic leadership. All tests were performed at a p = .05 statistical significance level.
For cross-validation, the participants in the EFA and CFA were different in this study. Therefore, 207 participants were randomly selected using IBM SPSS Version 26 and AMOS Version 23 programs to construct different participants for EFA and CFA. Approximately 100 participants were considered appropriate for EFA and CFA (MacCallum et al., 1999; Mundfrom et al., 2005). Therefore, 103 and 104 participants were used for EFA and CFA, respectively.
Ethical Considerations
This study was approved by the Ethics Committee following the ethical guidelines established by the Helsinki Declaration. The nursing department granted study approval following explanation of the research purpose and questionnaire contents. The researcher visited the head nurses and ICU nurses of each participating ICU, explained the purpose of the study, and subsequently obtained informed consent. Participants were informed regarding the purpose of the study, personal information protection, and voluntary withdrawal from the study. To protect their privacy, they were assigned a unique identification number.
Results
Demographic Data
A total of 207 ICU nurses were included in this study. Demographic data are shown in Table 1. The participants’ average age was 29.9 ± 6.25 years, of whom 201 (97.1%) and 6 (2.9%) were female and male participants, respectively. The following were the most prevalent characteristics in each category: Bachelor’s degree, 154 (74.4%); unmarried, 143 (69.1%); no religion, 140 (67.6%); and clinical experience of <3 years, 70 (33.8%).
Characteristics of Participants.
Note. *TICU = trauma intensive care unit; NICU = neonatal intensive care unit; EICU = emergency intensive care unit.
Construct Validity
In this study, factor analysis was performed following the classical theory test to verify the construct validity, which means whether or not the scale measures abstracted constructive concepts well. Unlike the original scale, this study was conducted on ICU nurses in Korea. As participants’ responses to measuring scales may vary depending on the characteristics, circumstances, and cultural differences of the survey target, EFA was first performed to construct an appropriate model or structure by searching for the characteristics inherent in the data without any special assumptions about the number or structure of the factors of the scale (Pett et al., 2003). To secure cross-validation, the participants undergoing EFA and CFA were differently distributed. EFA included 103 participants who were randomly extracted by generating random numbers using SPSS programs among 207 participants. The KMO and Bartlett’s test of sphericity were performed to evaluate the suitability of EFA before analysis. The KMO fit index of 0.81 and Bartlett’s test of sphericity confirmed that the factor structure was a good fit for the data (approx. = 472.822, df = 45, p < .001) (Andy, 2017). Two factors were extracted, with an Eigenvalue of >1.00 and a factor loading value of >0.4 after a common factor analysis with direct oblimin (Andy, 2017). The factor consisted of six items (1, 2, 3, 7, 8, 9, and 10), tapping the degree to which respondents prefer the characteristics of moral behavior, which is called “preference.” The second factor consisted of two items (4, 5, and 6), tapping the degree to which respondents indirectly express their moral characteristics, and this is called “preference.” The two extracted factors explained 52.17% of the total variance in the MIS. In the case of item 6, a reliability analysis was performed owing to cross-factor loading. When item 6 was put in factor 2, the Cronbach’s α was higher at .645; therefore, it was finally incorporated into factor 2 (indirectness) (Table 2).
Result of Exploratory Factor Analysis of K-MIS (n = 103).
Note. Factor 1 = preference; Factor 2 = indirectness; df = degree of freedom; Factor loading >0.4. Items in bold represent those included in each factor.
CFA was performed to determine the construct validity using a structural equation model following performing EFA using a different sample (Table 3). Item 8 of “preference” and item 4 of “indirectness” were deleted from model 1 as their standardized estimates (β) were low at 0.17 and 0.01, respectively. The standardized estimates (β) for the remaining eight items, excluding items 4 and 8, were 0.46 to 0.82, which met the minimum criteria (Kline, 2015). To verify the reliability and validity of the K-MIS, the critical ratio (CR) of the non-standardized lambda (λ) value was evaluated and observed to be in the range of 3.921 to 7.977 at the significance level of α = 0.05, thereby satisfying the analysis conditions of 1.96 or higher in model 2. Therefore, goodness-of-fit was examined for model 2, and the model fit indices were RMSEA = 0.113, GFI = 0.902, TLI = 0.847, and CFI = 0.896.
Goodness-of-Fit for Comparative of the Korean Version-Moral Identity Scale.
Note. df = degree of freedom; RMSEA = root mean square error of approximation; SRMR = standardized root mean square residual; GFI = goodness of fit index; TLI = Tucker-Lewis index; CFI = comparative fit index.
Model 1 = Korean version Moral Identity Scale (10 items-original model). Model 2 = Korean version Moral Identity Scale (8 items-deleted two item). Model 3 = modified model of the model 2.
In this study, the RMSEA and TLI values of model 2 did not meet the criteria; therefore, a modified model 3 was constructed by applying a modification index (MI) to increase the model fit (Figure 1). The resulting fitness indices were RMSEA = 0.076, GFI = 0.937, TLI = 0.93, and CFI = 0.955, which met all criteria (Table 3). To confirm the validity of the composition concept of this scale, the convergent validity and discriminant validity of the items were verified. The results of the convergent validity of the K-MIS under three conditions are shown in Table 4.
The two-factor structure of the Korean version of 8-item Moral Identity Scale.
Confirmatory Factor Analysis of K-MIS.
Note. AVE = average variance extracted; CR = critical ratio; K-MIS = Korean Moral Identity Scale; Factor 1 = preference; Factor 2 = indirectness.
First, the standardized factor load (estimate, β) for all items was approximately 0.45 to 0.82, which satisfies 0.5 or more. Second, the average variance extracted (AVE) value was approximately 0.4 to 0.5, which satisfies the standard of 0.5 or higher. Third, the construct reliability value was approximately 0.63 to 0.65, which satisfies the standard of 0.7 or higher. Therefore, the K-MIS satisfies three conditions, and the intensive validity of the question was confirmed.
To verify the discriminant validity of this scale, two conditions were checked. First, the squared value (p) 2 of the correlation coefficient of all factors was .27, which was less than the AVE value; therefore, the first condition was satisfied. Second, the standard error χ2 + correlation coefficient = .024 and standard error χ2 + correlation coefficient = .78. Therefore, the condition that “standard error χ2 ± correlation coefficient” must not include 1 was satisfied.
Convergent Validity
The correlation between authentic leadership and moral identity was analyzed for 207 individuals to verify the convergent validity of the construct validity. The ALI was correlated with moral identity in a previous study (Olsen & Espevik, 2017). Consequently, a significant positive correlation (r = .203, p = .003) between the K-MIS and authentic leadership was observed.
K-MIS Reliability Test
The reliability Cronbach’s α value for all eight questions in the K-MIS was .705. Furthermore, the Cronbach’s α value for each sub-factor was “preference” at .58 and “indirectness” at .79.
Discussion
This study was performed to adapt and validate the K-MIS for ICU nurses in the Korean cultural context.
In this study, as in the original MIS, it consisted of two factors; however, the items that constituted these factors were different. The reason for this difference is the variance in EFA methods. At the time of scale development, it initially conducted a principal components analysis with Varimax rotation, which is a common approach in natural sciences and does not incorporate an error term. However, given that this study falls within the field of social sciences, we employed common factor analysis with direct oblimin rotation, a method that appropriately accounts for error terms. This methodological difference may have contributed to the variations in item composition within each factor.
Based on these results, CFA was performed with participants who were different from those who participated in EFA. To verify the model fit, models for latent variables and items were constructed; however, the criteria were not met. This difference is because of cultural differences between Korea and the West. Item 4 is “I would be ashamed to be a person who had these characteristics.” Confucianism is one of the most influential social cultures in Korea and has more influence on the lifestyle of individuals than any other religion (Ha, 2018). Confucianism teaches that human destiny can be improved by exercising moral principles (Ha, 2018). In particular, the Confucian culture of Korea regards morality as vital enough to employ moral education as part of educational activities in official subjects or school environments (Han et al., 2018). Therefore, in Korean culture, there is no shame in having moral characteristics. Rather, Korean society expects individuals to act morally. Owing to such cultural differences, item 4 was deleted since it was believed to have a relatively low impact compared with other questions. Item 8, which stated “The fact that I have these characteristics is communicated to others by my membership in certain organizations,” also showed low standardized estimates. The original scale targeted college students who are members of specific organizations that perform various activities. However, in this study, it was believed that it had a relatively low effect as it was targeted at Korean nurses who are already affiliated with an organization called the ICU; therefore, it was deleted. Therefore, the K-MIS was constructed with the final two factors for a total of eight items.
Accordingly, CFA was performed with eight items, and as a result of calculating and analyzing the AVE, the convergent and discrimination validity of the items were also secured. Consequently, it was confirmed that six items of factor 1 and two questions of factor 2 were appropriate to measure each corresponding factor and were clearly distinguished from the questions included in other factors. In conclusion, it was judged that the construct validity of the eight items comprising the two factors constituting moral identity was secured.
To verify the convergent validity of the scale, the relationship with the MIS was analyzed by setting authentic leadership, a variable that was correlated with the MIS in previous studies, as the golden standard (Olsen & Espevik, 2017). Authentic leadership is a concept that values morality as a leader (Neider & Schriesheim, 2011). Because of analyzing the correlation between the two variables, the K-MIS showed a significant positive correlation with “authentic leadership,” suggesting that the convergence validity of this scale has been secured.
Finally, because of measuring the internal consistency of the scale, the reliability of the K-MIS was .67, and the reliability of factors was approximately .54 to .73. The overall reliability of the original MIS is unknown since it was not reported; however, the factor was approximately .76 to .77, and the reliability of the K-MIS was slightly lower than that of the original scale. In general, when the number of items is small, the Cronbach’s α value tends to be low. In this study, as two questions were deleted, reliability was measured using only eight questions. Therefore, the reliability may be reduced owing to the small number of items; however, the overall reliability of the scale was high. The K-MIS is a useful scale for measuring the moral identity of ICU nurses in Korea.
Limitations
Since moral identity can be influenced by individual temperament variables, including the personality characteristics of participants, conducting repeated studies with various participants and situation is necessary. Furthermore, this study has a limitation in generalizing the research results because data were collected through convenient sampling of ICU nurses. Therefore, repeating the study with an expanded sample size is necessary. Finally, the instrument was not completed repeatedly by the same sample; therefore, the level of test-retest reliability was not determined.
Implications for Research and Practice
The K-MIS holds significant implications for both research and practical applications in nursing. For research, the validated and reliable K-MIS provides a robust tool for investigating the moral identity of ICU nurses in Korea. This instrument can be used to explore the relationship between moral identity and various aspects of nursing practice, contributing to the academic understanding of nursing ethics.
In practical applications, the K-MIS is valuable for developing and assessing moral improvement training programs for ICU nurses. Nursing educators can use this scale to measure the effectiveness of interventions aimed at enhancing moral identity and ethical decision-making.
For nursing management, the K-MIS is an essential tool for nurse leaders and administrators. It enables them to evaluate and strengthen the moral identity of their nursing staff, leading to improved ethical practices in patient care. By identifying areas where additional support or training is needed, nursing managers can implement targeted strategies to bolster ethical decision-making and moral values among their teams. Ultimately, the K-MIS helps foster a morally driven, patient-centered healthcare environment, enhancing both nurse satisfaction and patient outcomes.
Conclusion
This study adapted the MIS developed by Aquino and Reed (2002) to construct the K-MIS, following WHO’s instrument translation procedure. Although the overall factor structure of all items was slightly different from that of the original MIS, the validity of the scale was secured as it is shown to explain each factor well. Moreover, reliability was secured by examining the internal consistency of the scale. Therefore, it is expected that the completed K-MIS will contribute to revitalizing the study of the moral identity of ICU nurses in Korea.
Footnotes
Acknowledgements
Gratitude is expressed to the nurses who participated in this study.
Ethical Considerations
This study was reviewed and approved by the Ethics Committee of Kyungpook National University (IRB NO. 2016-105).
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
Declaration of Conflicting Interests
The author(s) declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: This paper was previously published as a preprint in Research square and is now published in SAGE Open. The authors declare no potential or perceived conflicts of interest. 
Data Availability Statement
Data sharing not applicable to this article as no datasets were generated or analyzed during the current study.