Translation and psychometric validation of the Korean version of the academic nurse self-efficacy scale for Korean bachelor-level nursing students
Article Category: Original article
Data publikacji: 26 cze 2024
Zakres stron: 209 - 219
Otrzymano: 04 kwi 2023
Przyjęty: 26 wrz 2023
DOI: https://doi.org/10.2478/fon-2024-0024
Słowa kluczowe
© 2024 Eunkyung Lee et al., published by Sciendo
This work is licensed under the Creative Commons Attribution 4.0 International License.
Academic self-efficacy has been known to be an important predictor of academic performance.1 In addition to allowing them to choose challenging tasks and put additional effort into performing a given task successfully,2 it plays a prominent role in mediating learners’ judgment of their abilities to organize and implement the actions necessary to perform tasks in an academic situation.3 The concept of academic self-efficacy will be of interest to bachelor-level nursing college students’ capacity to build skills because it can also play a role in formulating professional identity as a nurse among nursing students besides directly affecting their learning performance.4
Studies have shown that college students with high levels of academic self-efficacy are more likely to adapt to university life and have excellent self-directed learning skills.5,6 A study of nursing college students also found that academic self-efficacy was associated with a high level of learning motivation, critical thinking, and problem-solving skills,7 as well as with introspection and insight.8 In particular, as clinical reasoning competencies constitute a complex cognitive process that collects and analyzes health questions, prioritizes them through logical reasoning, and leads to clinical judgment,9 it is classified as an essential nursing competency that professional nurses must possess.10 Therefore, imparting clinical reasoning skills is important in nurses’ training, and academic self-efficacy has been reported as a factor that increases the clinical reasoning capacity of nursing college students.8
In Korea, the nursing profession is presently recognized as one with a high employment rate, and this employment-oriented mindset has led to high competition in recent college entrance examinations. With this, involuntary choices based on the opinions of people around students are also increasing, rather than understanding academics or careers.11 After admission, many nursing students are struggling with the process of studying and clinical practice12 and show a tendency to drop out owing to the burden of large amounts of study and the fear of clinical practicum.11 Academic self-efficacy is known to have a positive effect on academic sustainability and achievement.13 Therefore, it is necessary to identify the degree of academic self-efficacy among nursing college students in Korea and to explore ways to improve and do this, it is imperative to have a measurement tool that has proven its reliability and validity in evaluating academic self-efficacy.
The tool developed by Kim and Park14 has been used to determine the academic self-efficacy of nursing college students in Korea, but it was originally developed to measure the academic self-efficacy in first grade of high school students. Hence, it is imperative to measure accurately the academic self-efficacy of college students who have already chosen their major.
The Academic Nurse Self-Efficacy scale (ANSEs) was developed by Bulfone et al.15 to assess the academic self-efficacy of nursing students. It was developed over 3 years by collecting data from 1129 Italian nursing students and verifying its reliability and validity. When using tools developed in foreign countries having a different culture from one’s own, in addition to sociocultural differences and the translation, the validity of the translation must also be checked carefully.16 Therefore, to use the ANSEs for bachelor-level nursing students in Korea, it is necessary to verify the validity and reliability of the Korean Version of the ANSEs (K-ANSEs). So, the purpose of this study is to translate the original ANSEs developed by Bulfone et al.15 into Korean and test the validity and reliability of the K-ANSEs for Korean bachelor-level nursing students.
This is a methodological study of nursing students in Korea that verifies the validity and reliability of the K-ANSEs.
This study’s participants were conveniently extracted from nursing college students studying in Korea. With respect to the number of participants required to perform a factor analysis for constituent validity verification, Nunnally and Bernstein17 suggest a ratio of 10:1 (the number of cases to measurement variables), but recently, it has been suggested that >200 people are sufficient for structural model equations or factor analysis.18 Additionally, Meyers et al.19 determined that the adequacy of sample sizes for factor analysis is good if the sample size is >300. In this study, 465 nursing students participated and, after excluding insufficient data, finally 444 participants’ data were analyzed.
The general characteristics of the nursing college students included age, year of study, sex, type of residence, stress level, exercise status, motivation for choosing the nursing department, and average grades.
The academic nurse self-efficacy tool for nursing students in this study was developed for Italian nursing students,15 and consisted of 14 questions and 4 subscales. The subfactors comprised: “auto-regulatory behavior” (4 questions), “collegiality” (3 questions), “internal emotion management” (3 questions), and “external emotion management” (4 questions). Each question was scored as follows: 1 point for “completely unconfident,” 2 for “very little confident,” 3 for “moderately confident,” 4 for “confident,” and 5 for “completely confident,” with scores ranging from 14 to 70 points, with a higher score denoting greater academic self-efficacy.15 The Cronbach’s alpha of the ANSEs in Italy was 0.84 for total and 0.72, 0.73, 0.80, and 0.83 for each of the subfactors.15
The academic self-efficacy tool is used as the gold standard in Korea for testing the validity of the ANSEs, whose validity and reliability have been verified. The tool, which was developed by Kim and Park14 for high school first-graders consists of 3 subfactors with a total of 28 questions, comprising 8 questions on confidence, 10 on self-regulating efficacy, and 10 on preference for task difficulty, each of which were scored on a 6-point Likert scale. The scale’s scores ranged from 28 to 168, implying that the higher the score, the greater the self-efficacy. At the time of development, Cronbach’s alpha for confidence was 0.74, self-regulating efficacy was 0.76, and task difficulty was 0.84.
To collect data, the researchers asked for the cooperation of professors from nursing departments at 10 universities across the country and conducted the data collection from November 1, 2020, to June 8, 2021. Initially, data collection was carried out in 1 university in Daegu City through questionnaires, but owing to the spread of COVID-19 and the winter vacation, subsequent data collection in the rest of the region was conducted online. Researchers in Daegu verbally explained information about the research process, such as the purpose and content of the study, research ethics, and data collection methods, and informed them that it was irrelevant to the evaluation of the subject. In the other 9 universities nationwide, the researcher asked the professors of the nursing department of the school for cooperation in collecting data and posted the recruitment document on the network to the students through the representative so that students could access the site and check the contents of the research manual and consent form. If the subjects agreed to participate in the study, they were asked to participate in the online survey. When collecting data, all items of the questionnaire must be answered to move to the next page, and only one response per question was allowed. A predetermined electronic coupon was issued to the mobile phone number provided after all responses were completed. Four hundred sixty-five respondents completed the questionnaire online and offline, and 444 copies were used for data analysis.
The data were analyzed using the IBM SPSS 25.0 (IBM Corporation, Armonk, NY, USA) and AMOS 25.0 (IBM Corporation, Armonk, NY, USA) programs. For the content validity analysis of K-ANSEs, a question item content validity index (I-CVI) was evaluated by a group of experts, and items with a CVI coefficient of 0.80 or higher were selected. The general characteristics of the participants were expressed using frequencies, percentages, mean and standard deviation; independent
For the second evaluation based on the data of subsample 2, CFA, convergent and discriminant validity tests were performed, and the goodness of fit of the research model was evaluated. The CFA was performed to examine the structural relationship between the subfactors derived through EFA. Since the values of skewness and kurtosis of individual items are assumed to be normal, the maximum likelihood estimation method was used. The goodness of fit of the research model was confirmed by χ2 statistics, chi-square minimization/degrees of freedom (CMIN/DF), comparative fit index (CFI), Tucker-Lewis index (TLI), the goodness of fit index (GFI), adjusted goodness-of-fit index (AGFI), root mean square error of approximation (RMSEA), and standardized root mean square residual (SRMR) based on previous research.20 For criterion-related validity tests, Pearson’s correlation analysis was used, and internal consistency of the K-ANSEs was tested using Cronbach’s α for reliability test.
Permission to conduct the study was approved by the Institutional Review Board of Daegu Catholic University (CUIRB-2020-0028). For the ethical protection of the participants, information about the research process, such as the purpose and content of the research, research ethics, and data collection methods, was written using terms and expressions that were easy to understand. In addition, the researchers’ contact information was provided in the explanation so that the research participants could contact them at any time if they had any questions about the study. Subjects who voluntarily agreed to participate in the study were informed that they could refuse to participate in the study at any time and there would be no disadvantage to it. A small gift in return was provided to all who participated in the offline and online surveys. The same compensation was also provided to the participants who withdrew from the survey.
For using the tool, approval was obtained from Dr. Bulfone, the developer of ANSEs. Based on the guidelines for the process of translation and application of the tools set forth by the World Health Organization,21 a draft of the K-ANSEs was prepared through forward and back translations. The original tool was translated by a researcher proficient in Korean and English, and the questions were revised after 2 nursing professors reviewed any sentences that required clarity and/or a cultural reflection of the translation. The translated questionnaire was translated back into English by those fluent in both English and Korean. Finally, the researchers compared it with the original tool to ensure that there was no difference in meaning between the questions.
Each item of K-ANSEs was evaluated by experts consisting of 3 professors and 2 doctoral students with >8 years of clinical experience in the department of nursing. For the translated measurement tool, the I-CVI, which measures the validity of the content on a scale of 4 wherein 1 point represents “not valid at all,” 2 “not valid,” 3 “valid,” and 4 “very valid,” was calculated.22 In addition, experts were asked to present their opinions on the need to revise the items and the composition of the measurement tools. As a result of verifying the content validity of the preliminary tool, one item, “Avoiding committing transgressions even when the risk of sanction is minimal,” had an I-CVI = 0.8, and the remaining questions were all between 0.94 and 1.00. Therefore, none of the 14 questions presented in the preliminary items were removed. However, the K-ANSEs were modified to reflect the opinions that were difficult to understand or awkward in expression among questions.
The translated K-ANSEs retained the same negative and positive question formats, arrangements, and question counts as the original tool. A preliminary survey was conducted on 16 students in juniors and seniors at a nursing school in Daegu. The questionnaire was distributed to the participants who provided their written consent after the purpose and method of the preliminary investigation were explained, and completed questionnaires were then collected. The preliminary survey was conducted to identify problems, such as comprehension of questions in the K-ANSE tool, ease of response, and time required to respond. Based on the results of the preliminary investigation, questions whose content meaning was unclear or difficult to understand were corrected. The final translation was completed and the average time taken to complete the questionnaire was approximately 3 min.
The average age of the 444 participants was 22.94 ± 4.21 years, and sophomores were the most with 234 (52.7%) participants. Three hundred sixty-seven (82.7%) participants were women, 303 participants (68.2%) lived with their families, and 243 (54.7%) reported moderate levels of stress. Overall, 264 participants (59.5%) did not exercise regularly and 165 (37.2%) entered the nursing department because it was easy to get a job. A grade of 3.5–4.0 was the most prevalent with 179 (40.3%) of participants. There was no significant difference between the explorative factor analysis and confirmative factor analysis participants (Table 1).
General characteristics of participants.
| Variables | Total ( |
Participants for EFA ( |
Participants for CFA ( |
||
|---|---|---|---|---|---|
| 22.94 ± 4.21 | 23.07 ± 4.93 | 22.88 ± 3.84 | 0.46 | 0.648 | |
| 7.20 | 0.066 | ||||
| Freshman | 11 (2.5) | 5 (3.5) | 6 (2.0) | ||
| Sophomore | 234 (52.7) | 76 (54.3) | 158 (52.0) | ||
| Junior | 105 (23.6) | 39 (27.9) | 66 (21.7) | ||
| Senior | 94 (21.2) | 20 (14.3) | 74 (24.3) | ||
| 1.33 | 0.248 | ||||
| Female | 367 (82.7) | 120 (85.7) | 247 (81.3) | ||
| Male | 77 (17.3) | 20 (14.3) | 57 (18.8) | ||
| 0.31 | 0.577 | ||||
| With family or relative | 303 (68.2) | 93 (66.4) | 210 (69.1) | ||
| Independent | 141 (31.8) | 47 (33.6) | 94 (30.9) | ||
| 0.59 | 0.745 | ||||
| Little | 78 (17.6) | 22 (15.7) | 56 (18.4) | ||
| Moderate | 243 (54.7) | 77 (55.0) | 166 (54.6) | ||
| A lot | 123 (27.7) | 41 (29.3) | 82 (27.0) | ||
| 0.003 | 0.96 | ||||
| Yes | 180 (40.5) | 57 (40.7) | 123 (40.5) | ||
| No | 264 (59.5) | 83 (59.3) | 181 (59.5) | ||
| 11.83 | 0.066 | ||||
| Guaranteed employment | 165 (37.2) | 45 (32.1) | 120 (39.5) | ||
| Advice from others | 60 (13.5) | 16 (11.4) | 44 (14.5) | ||
| Aptitude and interest | 66 (14.9) | 29 (20.7) | 37 (12.2) | ||
| Dreams from childhood | 50 (11.3) | 11 (7.9) | 39 (12.8) | ||
| Help others | 56 (12.6) | 22 (15.7) | 34 (11.2) | ||
| Economically guaranteed | 29 (6.5) | 12 (8.6) | 17 (5.6) | ||
| Can keep my major | 18 (4.1) | 5 (3.6) | 13 (4.3) | ||
| 1.55 | 0.817 | ||||
| 2< GPA ≤2.5 | 8 (1.8) | 4 (2.9) | 4 (1.3) | ||
| 2.5< GPA ≤3.0 | 31 (7.0) | 9 (6.4) | 22 (7.2) | ||
| 3.0< GPA ≤3.5 | 84 (18.9) | 28 (20.0) | 56 (18.4) | ||
| 3.5< GPA ≤4.0 | 179 (40.3) | 55 (39.3) | 124 (40.8) | ||
| 4.0< GPA | 142 (32.0) | 44 (31.4) | 98 (32.2) |
For EFA, subsample 1 (
Items’ distribution and EFA of the K-ANSE’s (
| Items | M(SD) | Skewness | Kurtosis | External emotion management | Internal emotion management | Sociality | Auto-regulatory behavior | Communalities |
|---|---|---|---|---|---|---|---|---|
| Item1 | 3.33 (1.01) | −0.40 | −0.23 | 0.352 | 0.145 | −0.014 | 0.742 | |
| Item2 | 3.64 (0.96) | −0.29 | −0.41 | 0.144 | 0.129 | 0.151 | 0.786 | |
| Item3 | 3.41 (1.14) | −0.38 | −0.73 | 0.266 | 0.162 | 0.119 | 0.702 | |
| Item4 | 3.61 (1.08) | −0.54 | −0.44 | 0.308 | 0.025 | 0.262 | 0.618 | |
| Item5 | 4.45 (0.84) | −1.38 | 0.88 | −0.279 | 0.151 | 0.026 | 0.536 | |
| Item6 | 3.72 (1.05) | −0.60 | −0.22 | 0.477 | −0.066 | 0.064 | 0.764 | |
| Item7 | 4.39 (0.82) | −1.30 | 1.13 | −0.138 | 0.118 | 0.102 | 0.597 | |
| Item8 | 2.84 (1.18) | 0.24 | −0.74 | 0.312 | −0.024 | 0.047 | 0.695 | |
| Item9 | 2.70 (1.16) | 0.27 | −0.79 | 0.151 | −0.002 | 0.003 | 0.804 | |
| Item10 | 3.08 (1.09) | −0.12 | −0.61 | 0.160 | 0.213 | −0.016 | 0.718 | |
| Item11 | 2.94 (1.08) | 0.16 | −0.63 | 0.314 | 0.172 | −0.003 | 0.809 | |
| Item12 | 4.00 (0.91) | −0.69 | 0.00 | 0.115 | 0.087 | 0.266 | 0.685 | |
| Item13 | 3.83 (0.94) | −0.64 | 0.07 | 0.073 | 0.260 | 0.001 | 0.769 | |
| Item14 | 3.94 (0.92) | −0.32 | −0.96 | 0.056 | 0.062 | 0.098 | 0.664 | |
| Eigenvalue | 3.36 | 2.28 | 2.16 | 2.09 | ||||
| Variance (%) | 24.00 | 16.29 | 15.43 | 14.93 | ||||
| Cumulated total variance (%) | 24.00 | 40.28 | 55.71 | 70.64 |
After establishing a measurement model based on the results of the EFA, a CFA was conducted on the data of the remaining 304 participants. First, the normality verification for CFA showed that skewness was from −1.92 to 0.32 and kurtosis was −0.98 to 3.37, which is not severely non-normal.18 Thus, the assumption of univariate normality was satisfied (Table 3). CFA was performed on 4 subfactors of the K-ANSEs to evaluate the fit index of the measurement model (Table 4, Figure 1). The fit index was not reached, except SRMR, with χ2 = 250.03 (
Korean version of the ANSE questionnaire scores (N = 304).
| Factors and items | How much are you confident with | M (SD) | Min | Max | Skewness | Kurtosis |
|---|---|---|---|---|---|---|
| Factor 1 | Internal emotion management | 3.40 (0.81) | ||||
| Item1 | Controlling anxiety in front of a problem | 3.25 (0.94) | 1 | 5 | −0.23 | −0.40 |
| Item2 | Keeping calm during an exam | 3.67 (0.94) | 1 | 5 | −0.57 | −0.06 |
| Item3 | Avoiding discouraging myself in adversity | 3.29 (1.09) | 1 | 5 | −0.17 | −0.73 |
| Factor 2 | Auto-regulatory Behavior | 4.08 (0.65) | ||||
| Item4 | Resisting the pressure of friends for doing some things that risk getting you into a trouble | 3.63 (1.12) | 1 | 5 | −0.55 | −0.55 |
| Item5 | Resisting the temptation not to go to the lesson if you feel bored |
4.51 (0.84) | 1 | 5 | −1.92 | 3.37 |
| Item6 | Avoiding the insistence of friends who ask you to do some things that you think would be better to avoid | 3.69 (1.05) | 1 | 5 | −0.52 | −0.54 |
| Item7 | Avoiding committing transgressions even when the risk of sanction is minimal |
4.47 (0.77) | 2 | 5 | −1.44 | 1.50 |
| Factor 3 | External emotion management | 2.92 (0.98) | ||||
| Item8 | Do not spiritless when you are criticized | 2.93 (1.20) | 1 | 5 | 0.14 | −0.98 |
| Item9 | Containing shame after making a bad impression in front of the class | 2.75 (1.12) | 1 | 5 | 0.32 | −0.65 |
| Item10 | Overcoming the embarrassment of having made a “gaffe” with a person to the judgment of which you care a lot | 3.04 (1.13) | 1 | 5 | −0.07 | −0.91 |
| Item11 | Dominating shame when your frailties have highlighted in front of the class | 2.94 (1.08) | 1 | 5 | −0.02 | −0.88 |
| Factor 4 | Sociality | 3.93 (0.76) | ||||
| Item12 | Ensuring me the help of other students when necessary | 3.95 (0.95) | 1 | 5 | −0.80 | 0.14 |
| Item13 | Helping a colleague in difficulty in the study | 3.90 (0.96) | 1 | 5 | −0.82 | 0.56 |
| Item14 | Helping in creating a good atmosphere among students | 3.93 (0.97) | 1 | 5 | −0.77 | 0.38 |
Model fit indices for K-ANSEs subscales from confirmatory factory analysis.
| Items | χ2 (P) | Normed χ2 | CFI | TLI | GFI | AGFI | RMSEA | SRMR |
|---|---|---|---|---|---|---|---|---|
| Original model K-ANSEs | 250.03 (<0.001) | 3.52 | 0.89 | 0.86 | 0.89 | 0.84 | 0.09 | 0.08 |
| Modified model K-ANSE (deleted items 5 and 7) | 141.08 (<0.001) | 2.94 | 0.94 | 0.92 | 0.93 | 0.89 | 0.08 | 0.06 |
| Modified K-ANSE (deleted items 5 and 7) Second-order hierarchical model | 148.43 (<0.001) | 2.97 | 0.94 | 0.92 | 0.93 | 0.89 | 0.08 | 0.07 |
Results of CFA, convergent and discriminant validity, and reliability test of final K-ANEs (N = 304).
| Dimensions and items | β | B | SE | C.R. | P | CR | AVE | Cronbach’s α |
|---|---|---|---|---|---|---|---|---|
| 0.76 | 0.52 | 0.76 | ||||||
| 1 | 0.79 | 0.94 | 0.09 | 10.99 | <0.001 | |||
| 2 | 0.63 | 0.75 | 0.08 | 9.46 | <0.001 | |||
| 3 | 0.72 | 1 | Fix | Fix | Fix | |||
| 0.74 | 0.59 | 0.77 | ||||||
| 4 | 0.81 | 1.12 | 0.13 | 8.88 | <0.001 | |||
| 6 | 0.77 | 1 | Fix | Fix | Fix | |||
| 0.86 | 0.61 | 0.89 | ||||||
| 8 | 0.8 | 1.09 | 0.07 | 14.89 | <0.001 | |||
| 9 | 0.82 | 1.03 | 0.07 | 15.27 | <0.001 | |||
| 10 | 0.79 | 1 | Fix | Fix | Fix | |||
| 11 | 0.84 | 1.01 | 0.07 | 15.64 | <0.001 | |||
| 0.74 | 0.49 | 0.7 | ||||||
| 12 | 0.5 | 0.64 | 0.09 | 7.15 | <0.001 | |||
| 13 | 0.76 | 0.99 | 0.11 | 8.7 | <0.001 | |||
| 14 | 0.77 | 1 | Fix | Fix | Fix | |||
| 0.87 |
Figure 1.
Confirmatory Factor Analysis original and modified model of K-ANSEs.
Figure 2.
Final model of K-ANSE: Second-order hierarchical model with modified K-ANSEs.
The convergent/discriminant validity of the 12 items was verified using construct reliability (CR) and average variance extracted (AVE). As a result, the CR was 0.74–0.86, and for AVE, the sociality was slightly lower at 0.49, but for the other items, the AVE was concentrated between 0.52 and 0.61, confirming the convergent validity of the K-ANSEs. The correlation coefficient of the subfactors was found to be 0.33–0.56, and by squaring it, the correlation between the latent variables and the AVE value was confirmed. As a result, discriminant validity was confirmed (Table 5).
To verify the criterion-related validity of the K-ANSEs, correlation analysis with Kim and Park’s14 academic self-efficacy scale showed that the correlation coefficient between the 2 measuring tools was 0.63 (
Correlation coefficients between the 12-item K-ANSEs and ASEs (N = 304).
| Items | 12-item K-ANSE total | F1 = Internal emotion management | F2 = Auto-regulatory behavior | F3 = External emotion management | F4 = Sociality |
|---|---|---|---|---|---|
| Academic self-efficacy total | 0.63 (<0.001) | 0.53 (<0.001) | 0.36 (<0.001) | 0.48 (<0.001) | 0.51 (<0.001) |
| F1 = Task difficulty preference | 0.38 (<0.001) | 0.35 (<0.001) | 0.16 (0.005) | 0.34 (<0.001) | 0.24 (<0.001) |
| F2 = Self-regulatory efficacy | 0.44 (<0.001) | 0.31 (<0.001) | 0.29 (<0.001) | 0.27 (<0.001) | 0.49 (<0.001) |
| F3 = Self-confidence | 0.59 (<0.001) | 0.52 (<0.001) | 0.35 (<0.001) | 0.46 (<0.001) | 0.43 (<0.001) |
The overall Cronbach’s α for the K-ANSEs was 0.87, whereas for the sub-factors, it ranged from 0.70 to 0.89 (Table 5).
The purpose of this study was to verify the validity and reliability of the K-ANSEs by translating the ANSEs developed by Bulfone et al.15 to measure the academic self-efficacy of Korean nursing college students. As the validity of tools cannot be sufficiently supported by one type of validity verification, a variety of logical analyses must be presented to provide a rational rationale for the use of the tool.25,26 This study was conducted to verify the validity and reliability of the K-ANSE among 444 Korean nursing college students. We confirmed that the K-ANSEs consisted of 12 items, in which 2 of the 14 items in the original tool were deleted, were the same as the constituent factors of the original tool, and convergent and discriminant validity at the factor and item levels were confirmed.
EFA, CFA, convergent/discriminant validity and criterion-related validity were performed to determine the constituent validity of the K-ANSEs. The EFA results were extracted from 4 factors. The factor load for each item was ≥0.66 and the explained variance of factors 1, 2, 3, and 4 were 24.0%, 16.3%, 15.4%, and 14.9%, respectively. The total explained cumulative variance of the K-ANSE question was 70.6%, which was >65.3% of the original tool15 and was generally judged to be a valid question when the factor load was >0.60 with the relevant factors.27 Therefore, we confirmed that the corresponding questions for the 4 sub-factors of the K-ANSEs measure each factor appropriately.
The CFA was conducted to confirm whether the components of the original ANSE were suitable for Korea. In the initial model, the SRMR value was 0.08, and all fitness indices, except the SRMR, did not meet the standard.18 As a result of checking the standardized regression weights of the question, item 5 (“Resisting the temptation not to go to the lesson if you feel bored”) is 0.23 and item 7 (“Avoiding committing transgressions even when the risk of sanction is minimal”) is 0.24. In the modified model, these items were deleted because they did not meet the criterion of 0.5.28 Only the AGFI was 0.89, which did not meet the AGFI criterion of 0.90 or higher,18 but all other fitness indices met the criterion. The 2 deleted questions were all auto-regulatory behaviors of the subfactors and it seems that the standardized regression weights were low due to the situational difference between Italy and Korea. First, item 7 has a CVI value of 0.8 in the content validity verification, which is lower than the CVI values of 0.94–1.00 for the other questions but is in line with the adoption criteria,22 which is an item included in the CFA. Looking at the domestic situation, the number of nursing departments in Korea has doubled over the past 10 years from about 10,000 in 2008 to over 20,000 in 2019.29 Consequently, owing to the high employment rate compared to other majors that do not include personal values, such as aptitude or interest, the number of nursing students who want to enter the department is also increasing,30 indicating a high level of competition for entrance examinations. As a result, the number of students entering nursing schools with excellent grades has increased and hence the number of new nurses produced each year is also increasing. Hence, the competition for employment is intensifying. It seems that the standardized regression weights of the 2 items were low because the 2 deleted items did not reflect the high sensitivity to attendance and grades owing to the so-called credit war experienced by nursing students in Korea. This reasoning is also supported by the fact that the mean ± SD of the item at the time of tool development was 3.91 ± 1.11 and 4.06 ± 0.97, respectively, lower than the score of 4.51 ± 0.84 and 4.47 ± 0.77 for the subjects in this study. However, this study looked at some nursing students in South Korea. Also, information can be lost or tampered with when moving from one language to another, so it is necessary to reconfirm the standardized regression weights through repeated studies in the future.
In this study, the discriminant validity of the questions was evaluated. Although the AVE of sociality was 0.49, slightly lower than the recommended standard of 0.50,31 the AVE in the remaining 3 factors met the recommended criteria, and CR also met the recommended criterion of 0.70 or higher.31 This means that the contents of each item constituting the K-ANSEs consistently measure the constituent concepts of the subfactors. At the time of the development of the ANSEs,15 discriminant validity was not tested, so the discriminant validity of the K-ANSEs and ANSEs could not be directly compared. The coefficients ranged from 0.33 to 0.56, indicating that the degree of correlation between the sub-factors was similar. Consequently, even though the two studies targeted nursing students from different countries, the influence of the subfactors of the ANSEs and the K-ANSEs was judged to be similar.
To confirm the criterion-related validity of the K-ANSEs, the academic self-efficacy scale of Kim and Park14 which is widely used in domestic academic self-efficacy measures, was used as the gold standard. The correlation between the 2 tools indicated a statistically significant positive correlation, confirming that it is an appropriate tool for measuring the academic self-efficacy of nursing college students in the field of nursing education in Korea. The total explained that the cumulative variance of K-ANSEs was 70.6%, which was lower than the 78% of Kim and Park14. However, Kim and Park’s14 study was designed for high school students, whereas K-ANSE was designed for nursing students and the total cumulative variance explained met the recommended level. Therefore, this tool was used to determine the level of academic self-efficacy of home nursing students.
Based on the result of the internal consistency analysis of the K-ANSEs, the reliability of Cronbach’s α was 0.76 and the subfactors were 0.70–0.76, which was slightly lower than the reliability of the original ANSEs.15 However, as the values are higher than the recommended reliability standard of 0.70,17 overall, the tool can be judged as a reliable tool. However, there are certain limitations of this study. First, since this scale is a self-assessment scale and the study participants often select the correct answer instead of what they really think subconsciously, and so bias is inevitable. Second, the sample was selected for convenience and the findings may reflect local reality. Third, the total number of female nursing students in the sample of this study was much larger than that of male nurses, and the distribution of participants’ academic performance was uneven, with only a small percentage of students performing poorly. It is proposed to conduct research by selecting subjects with similar gender composition and academic achievement distribution.
Academic self-efficacy—the ability to adapt and perform the necessary skills using one’s knowledge and own judgment with confidence, when faced with a difficult problem in an academic situation—is a significant factor in adapting to school life. This study examined the ANSEs developed for Italian nursing students through a tool translation process, constructing the K-ANSEs with 4 factors, 12 questions, and verifying their credibility and validity. To determine the validity of their constituents, EFA and CFA were used to extract subfactors and test their suitability and theoretical validity. However, 2 items with low standardized regression weights were excluded from this study. Future testing of K-ANSE is recommended to reassess the stability of the tool. Based on our findings, this tool is intended to be used to develop and evaluate various programs to improve scientific self-efficacy. Academic self-efficacy monitoring can also help to assess the long-term impact of different teaching methods and curricula in nursing education and to improve the quality of nursing degree programs.