The term psychological resilience refers to an individual’s capacity to adjust and recover from adversity, trauma, or extreme stress (Fletcher & Sarkar, 2013). Individuals with robust psychological resilience can sustain a sense of well-being and perform their daily tasks efficiently despite facing difficulties. They are adept at problem solving and can devise practical and feasible solutions to overcome obstacles and achieve their goals (Tugade & Fredrickson, 2004).

 

Psychological resilience is particularly crucial for health practitioners who are regularly exposed to high-stress situations and challenges, such as demanding workloads, extended working hours, and moral dilemmas, along with facing constant exposure to human suffering, high-pressure decision making, and emotional demands from patient care. These challenges and stressors can adversely affect their well-being and mental health (McCann et al., 2013), and can lead to burnout, compassion fatigue, and a decline in the quality of patient care (Arimura et al., 2010; Epstein & Hamric, 2009; Shanafelt & Noseworthy, 2017). However, psychological resilience can help healthcare practitioners manage job-related stress, preventing burnout, preserving their physical and mental health, and maintaining their capacity to deliver high-quality patient care (Maunder et al., 2021).

 

Because of the stressors and challenges health practitioners face in the workplace, it is crucial to develop a psychological resilience scale designed specifically for them. Such a scale can identify the precise factors that contribute to psychological resilience among health practitioners (McKinley et al., 2019). Wagnild and Young (1993) developed a 25-item Resilience Scale for a randomly selected sample of older community-dwelling individuals aged between 53 and 95 years. The scale consists of two factors: personal competence and acceptance of self and life. Friborg et al. (2003) developed the Resilience Scale for Adults to assess resilience of adults at an outpatient clinic in Tromsø, northern Norway. They found that the scale was a valid tool for evaluating five factors: personal competence, social competence, family coherence, social support, and the individual’s inclination to uphold regular routines, plan ahead, manage time efficiently, and pursue set goals (personal structure). Connor and Davidson (2003) developed the Connor–Davidson Resilience Scale, which they administered to samples of general psychiatric outpatients, primary care outpatients, and participants in clinical trials. This scale includes five factors: personal competence, high standards, and tenacity; trust in one’s instincts, tolerance of negative affect, and strengthening effects of stress; positive acceptance of change and secure relationships; control; and spiritual influences. With proven good validity and reliability, the Connor–Davidson Resilience Scale has been adopted in healthcare studies to measure resilience among healthcare practitioners. Smith et al. (2008) developed the Brief Resilience Scale to evaluate an individual’s capacity to recover from stress. The psychometric properties of the scale were evaluated across four different groups: patients with a heart condition, patients with chronic pain, and two groups of students. In all groups, the Brief Resilience Scale consistently showed a predictable correlation with personal traits, social relations, coping mechanisms, and health, proving it to be a reliable measure of a unitary construct.

 

Although each of these psychological resilience scales has demonstrated validity for measuring psychological resilience, none of them were designed specifically for health practitioners and students, and they were also developed and validated in Western cultural contexts. Therefore, these scales may not fully capture the resilience factors relevant to Taiwanese health practitioners and students. Thus, developing a psychological resilience measure specifically designed for Taiwanese health practitioners and students becomes imperative. To meet this need, in the current study we proposed the following hypotheses:

Hypothesis 1: The Psychological Resilience Scale for Health Practitioners and Students will be a reliable instrument for assessing the psychological resilience of health practitioners and students in Taiwan.

Hypothesis 2: Analysis of the Psychological Resilience Scale for Health Practitioners and Students will confirm the four-factor model identified in an exploratory factor analysis has satisfactory model fit indices.

Chung Shan Medical University Hospital granted Institutional Review Board approval for the study (CS2-20198). To develop the Psychological Resilience Scale for Health Practitioners and Students (PRS-HPS), we first conducted an extensive literature review using databases such as EBSCO, PubMed, ProQuest, and ScienceDirect. We extracted 87 items related to psychological resilience. Following this, we held panel discussions with three specialists in psychometrics, health education, social science, and medical humanities to ensure that these items adequately covered the theoretical concept. These specialists reviewed and rated the 87 items for relevance to the constructs on a 6-point scale (0–5; 0 = not relevant, 1 = slightly relevant, 2 = somewhat relevant, 3 = moderately relevant, 4 = very relevant, 5 = extremely relevant). Items rated below 4 were removed, and items for which there was not a consensus were discarded. Through three rounds of panel discussions, these 87 items were reduced to 59, using a 9-point Likert scale (9 = always, 8 = almost always, 7 = very often, 6 = often, 5 = neutral, 4 = sometimes, 3 = occasionally, 2 = rarely, 1 = never). Higher scale scores indicated greater psychological resilience. The 59 items were then translated into Mandarin Chinese and back-translated into English to ensure semantic equivalence, with two bilingual English teachers verifying the content and face validity (Boateng et al., 2018).

 

We then conducted a pilot study with 738 participants who were either health students or health practitioners in central Taiwan with health-related training or competence. The mean age of the sample was 35.91 years (SD = 3.69, range 19–62). The mean age for men was 37.2 years (SD = 3.82, range 19–60) and that for women was 35.1 years (SD = 3.61, range 19–62).

The consistency of the scale factors was confirmed using exploratory factor analysis (EFA) with SPSS (Version 24.0; International Business Machines Corporation, 2016) and confirmatory factor analysis (CFA) with AMOS (Version 24.0; Arbuckle, 2016). Eigenvalues, promax rotation, and principal component analysis were utilized to identify the factor structure. To ensure the sample size was adequate for factor analysis, Bartlett’s Test of Sphericity (Bartlett, 1950) and the Kaiser–Meyer–Olkin (KMO) Test (Kaiser, 1970) were employed. The model fit was evaluated using the comparative fit index (CFI; Bentler, 1990), Tucker–Lewis index (TLI; Bentler, 1990), root-mean-square error of approximation (RMSEA; Hooper et al., 2008), and χ²/df ratio (Hooper et al., 2008).

To examine the psychometric properties and perform EFA, data from 430 participants were used. The KMO value for the current study was .971, which is higher than the threshold value of .6 (Kaiser & Rice, 1974). The Bartlett’s Test of Sphericity (Bartlett, 1950) result was statistically significant (11913.27; df = 528; p < .001). These results from the KMO and Bartlett’s tests indicated that the sample size was appropriate for EFA. The scree plot for factor analysis of the PRS-HPS also indicated that four factors were optimal for the scale (see Figure 1).

EFA was conducted to test the construct validity and internal consistency of the PRS-HPS, using eigenvalues greater than 1.0, principal component analysis, and promax rotation. After EFA, 33 items and four factors were identified that accounted for 68.310% of the variance. Factor 1, with nine items related to positive expectations, explained 54.161% of the variance. Factor 2, with eight items related to meaningfulness, explained 5.910% of the variance. Factor 3, with nine items on self-reliance, explained 5.151% of the variance. Factor 4, with seven items on composure, explained 3.089% of the variance (see Table 1). The eigenvalues of the four factors in the principal component analysis were all greater than one: 17.873, 1.950, 1.700, and 1.019 (see Table 1). These findings support the multidimensionality of the PRS-HPS.

Table 2 shows the scale items, average item scores, and standard deviations for the four subscales of the PRS-HPS.

CFA was conducted using AMOS (Arbuckle, 2016) with data from 308 participants to further validate the factor structure. The CFA confirmed the same four factors. Item 33 was deleted during the CFA process because its standardized factor loading did not meet the set threshold (i.e., it was below .70), indicating that it did not strongly represent the latent construct of composure in our model. Therefore, we decided to exclude it in order to improve both the overall model fit and the construct validity of the composure factor. This resulted in a 32-item model (see Figure 2) as follows: positive expectations (nine items; factor loadings .795–.887), meaningfulness (eight items; factor loadings .809–.979), self-reliance (nine items; factor loadings .815–.895), and composure (six items; factor loadings .790–.863).

The fit indices showed that in the EFA-derived scale, the χ²/df ratio was 3.586, the TLI was .883, the CFI was .892, and the RMSEA was .092. In the CFA-derived scale, the χ²/df ratio was 1.222, the TLI was .990, the CFI was .993, and the RMSEA was .027. The CFA-derived scale demonstrated outstanding reliability, with Cronbach’s alpha values for the subscales of positive expectations, meaningfulness, self-reliance, composure, and the overall scale, respectively of .960, .949, .964, .933, and .984, and composite reliability values of .959, .959, .962, .926, and .988. Average variance extracted (AVE) values and composite reliability were used to examine the convergent validity of the PRS-HPS, determining whether the items converge and accurately reflect the underlying construct. The AVE values for the four PRS-HPS factors were: positive expectations (.725), meaningfulness (.747), self-reliance (.739), and composure (.676). All AVE values were lower than the corresponding composite reliability values (.959, .959, .962, and .926), each of which exceeded .70 (Hair et al., 2018; Pallant, 2013).

In this study we developed the PRS-HPS, a psychological resilience scale for health practitioners and students in Taiwan, with an emphasis on cultural sensitivity and fit specific to Taiwanese health settings. The highest mean score was for the positive expectations factor, with an average score of 6.364 per item. The self-reliance factor had the second-highest mean score, at 6.152. The lowest mean score was 5.893, that of the composure factor. This suggests that the participants tended to exhibit a strong sense of optimism and hope for the future (positive expectations), a strong belief in their own abilities (self-reliance), and a significant sense of purpose and meaning in their activities (meaningfulness). However, they seemed to struggle with maintaining steadiness and composure in the face of stress or adversity (composure).

 

Regarding model fit indices, for the χ²/df ratio, some experts suggest a range of 5.0 to 2.0 (Tabachnick & Fidell, 2007; Wheaton et al., 1977), whereas others argue that a χ²/df < 2.0 indicates strong model fit (Koufteros, 1999; Schumacker & Lomax, 2010). The EFA-derived scale in our study had a χ²/df ratio of 3.586, which falls within the acceptable range of 5.0 to 2.0 (Tabachnick & Fidell, 2007). The CFA-derived model showed a good fit to the data, with a χ²/df ratio of 1.222, which is much lower than the usual criterion of 2 (Schumacker & Lomax, 2010). The TLI for the EFA-derived scale was .883, slightly below the desirable threshold of .90 (Schermelleh-Engel & Moosbrugger, 2003), indicating that the model could be improved. In contrast, the TLI value for the CFA-derived scale was .990, exceeding the benchmark value of .90, indicating a very good fit. The CFI for the EFA-derived scale was .892, slightly below the recommended level of .90 (Hu & Bentler, 1999), suggesting a marginally acceptable model fit. However, the CFI for the CFA-derived scale was .993, well above .90, indicating an excellent model fit. Fabrigar et al. (1999) stated that an RMSEA index of less than .05 suggests a good fit, an index of .05 to .08 is considered acceptable, an index of .08 to .10 shows mediocre fit, and an index above .10 suggests poor model fit. The RMSEA value for the EFA-derived scale was .092, suggesting a marginal model fit. Conversely, the RMSEA value for the CFA-derived scale was .027, well below .05, indicating an excellent fit. The comparison of the fit indices reveals that the CFA-derived scale had a much better model fit than the EFA-derived scale in regard to χ²/df ratio, CFI, TLI, and RMSEA values, demonstrating the validity and reliability of the scale as a robust and well-structured model with low error, making it a more accurate and reliable assessment instrument for measuring psychological resilience among Taiwanese health practitioners and students. Regarding reliability, both the EFA-derived and CFA-derived PRS-HPS showed strong internal consistency, with Cronbach’s alpha and composite reliability values ranging from .903 to .973, well above the minimum acceptable reliability value of .70 (Hair et al., 2018).

 

In conclusion, the study demonstrated that the EFA- and CFA-derived PRS-HPS are both reliable instruments for assessing the psychological resilience of health practitioners and students in the cultural context of Taiwan. Nonetheless, there may be potential limitations in the study. Participants might have provided socially desirable responses when answering the scale items, which could have influenced the accuracy of their self-reported data and potentially skewed the results. In future studies, researchers could examine criterion validity to further validate the developed scale and explore the feasibility of implementing resilience-building interventions among health practitioners and students in practical settings to provide evidence-based insights into the impact of such interventions on participants’ psychological resilience.

Arbuckle, J. L. (2016). IBM SPSS Amos 24 user’s guide. IBM.

 

Arimura, M., Imai, M., Okawa, M., Fujimura, T., & Yamada, N. (2010). Sleep, mental health status, and medical errors among hospital nurses in Japan. Industrial Health, 48(6), 811–817. https://doi.org/10.2486/indhealth.ms1093

 

Bartlett, M. S. (1950). Tests of significance in factor analysis. British Journal of Statistical Psychology, 3(2), 77–85. https://doi.org/10.1111/j.2044-8317.1950.tb00285.x

 

Bentler, P. M. (1990). Comparative fit indexes in structural models. Psychological Bulletin, 107(2), 238–246. https://doi.org/10.1037/0033-2909.107.2.238

 

Boateng, G. O., Neilands, T. B., Frongillo, E. A., Melgar-Quiñonez, H. R., & Young, S. L. (2018). Best practices for developing and validating scales for health, social, and behavioral research: A primer. Frontiers in Public Health, 6, Article 149. https://doi.org/10.3389/fpubh.2018.00149

 

Connor, K. M., & Davidson, J. R. T. (2003). Development of a new resilience scale: The Connor-Davidson Resilience Scale (CD-RISC). Depression and Anxiety, 18(2), 76–82. https://doi.org/10.1002/da.10113

 

Epstein, E. G., & Hamric, A. B. (2009). Moral distress, moral residue, and the crescendo effect. The Journal of Clinical Ethics, 20(4), 330–342. https://doi.org/10.1086/JCE200920406

 

Fabrigar, L. R., Wegener, D. T., MacCallum, R. C., & Strahan, E. J. (1999). Evaluating the use of exploratory factor analysis in psychological research. Psychological Methods, 4(3), 272–299. https://doi.org/10.1037/1082-989X.4.3.272

 

Fletcher, D., & Sarkar, M. (2013). Psychological resilience: A review and critique of definitions, concepts, and theory. European Psychologist, 18(1), 12–23. https://doi.org/10.1027/1016-9040/a000124

 

Friborg, O., Hjemdal, O., Rosenvinge, J. H., & Martinussen, M. (2003). A new rating scale for adult resilience: What are the central protective resources behind healthy adjustment? International Journal of Methods in Psychiatric Research, 12(2), 65–76. https://doi.org/10.1002/mpr.143

 

Hair, J., Black, W., Babin, B., & Anderson, R. (2018). Multivariate data analysis (8th ed.). Cengage.

 

Hooper, D., Coughlan, J., & Mullen, M. (2008). Structural equation modeling: Guidelines for determining model fit. Electronic Journal on Business Research Methods, 6, 53–60. https://academic-publishing.org/index.php/ejbrm/article/view/1224

 

Hu, L., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling: A Multidisciplinary Journal, 6(1), 1–55. https://doi.org/10.1080/10705519909540118

 

International Business Machines Corporation. (2016). IBM SPSS statistics for Windows (Version 24.0). IBM Corporation.

 

Kaiser, H. F. (1970). A second generation Little Jiffy. Psychometrika, 35(4), 401–415. https://doi.org/10.1007/BF02291817

 

Kaiser, H. F., & Rice, J. (1974). Little Jiffy, Mark IV. Educational and Psychological Measurement, 34(1), 111–117. https://doi.org/10.1177/001316447403400115

 

Koufteros, X. A. (1999). Testing a model of pull production: A paradigm for manufacturing research using structural equation modeling. Journal of Operations Management, 17(4), 467–488. https://doi.org/10.1016/S0272-6963(99)00002-9

 

Maunder, R. G., Heeney, N. D., Strudwick, G., Shin, H. D., O’Neill, B., Young, N., Jeffs, L., Barrett, K., Bodmer, N. S., Born, K. B., Hopkins, J. P., Jüni, P., Perkhun, A., Price, D. J., Razak, F., Mushquash, C. J., & Mah, L. (2021). Burnout in hospital-based healthcare workers during COVID-19. Science Briefs of the Ontario COVID-19 Science Advisory Table, 2(46). https://doi.org/10.47326/ocsat.2021.02.46.1.0

 

McCann, C. M., Beddoe, E., McCormick, K., Huggard, P., Kedge, S., Adamson, C., & Huggard, J. (2013). Resilience in the health professions: A review of recent literature. International Journal of Wellbeing, 3(1), 60–81. https://doi.org/10.5502/ijw.v3i1.4

 

McKinley, N., Karayiannis, P. N., Convie, L., Clarke, M., Kirk, S. J., & Campbell, W. J. (2019). Resilience in medical doctors: A systematic review. Postgraduate Medical Journal, 95(1121), 140–147. https://doi.org/10.1136/postgradmedj-2018-136135

 

Pallant, J. (2013). SPSS survival manual: A step-by-step guide to data analysis using SPSS (4th ed.). Allen & Unwin.

 

Schermelleh-Engel, K., & Moosbrugger, H. (2003). Evaluating the fit of structural equation models: Tests of significance and descriptive goodness-of-fit measures. Methods of Psychological Research Online, 8(2), 23–74. https://www.stats.ox.ac.uk/~snijders/mpr_Schermelleh.pdf

 

Schumacker, R. E., & Lomax, R. G. (2010). A beginner’s guide to structural equation modeling (3rd ed.). Routledge.

 

Shanafelt, T. D., & Noseworthy, J. H. (2017). Executive leadership and physician well-being: Nine organizational strategies to promote engagement and reduce burnout. Mayo Clinic Proceedings, 92(1), 129–146. https://doi.org/10.1016/j.mayocp.2016.10.004

 

Smith, B. W., Dalen, J., Wiggins, K., Tooley, E., Christopher, P., & Bernard, J. (2008). The Brief Resilience Scale: Assessing the ability to bounce back. International Journal of Behavioral Medicine, 15(3), 194–200. https://doi.org/10.1080/10705500802222972

 

Tabachnick, B. G., & Fidell, L. S. (2007). Using multivariate statistics (5th ed.). Allyn & Bacon.

 

Tugade, M. M., & Fredrickson, B. L. (2004). Resilient individuals use positive emotions to bounce back from negative emotional experiences. Journal of Personality and Social Psychology, 86(2), 320–333. https://doi.org/10.1037/0022-3514.86.2.320

 

Wagnild, G. M., & Young, H. M. (1993). Development and psychometric evaluation of the Resilience Scale. Journal of Nursing Measurement, 1(2), 165–178. https://pubmed.ncbi.nlm.nih.gov/7850498/

 

Wheaton, B., Muthen, B., Alwin, D. F., & Summers, G. F. (1977). Assessing reliability and stability in panel models. Sociological Methodology, 8(1), 84–136. https://doi.org/10.2307/270754