“Coping self-efficacy scale” evaluation of the Turkish validity and reliability of the scale

ABSTRACT

Background:

The purpose of the study is to evaluate the validity and reliability of the Coping Self-Efficacy Scale (CSES) for family physicians.

Method:

The study is a correlational cross-sectional type of methodological study conducted with 255 family physicians. In the research, data were collected using descriptive data record form and the CSES. The collected data were analyzed using the following statistical methods: “Explanatory Factor Analysis (EFA),” “Confirmatory Factor Analysis (CFA),” “Cronbach’s Alpha.”

Results:

This study involved family physicians (mean age: 38.5 ± 10.1, 59.2% women), with 72.2% married and 59.6% having children. Work settings varied, with 39.2% in family health centers, 38.0% in university hospitals, and 16.5% in state hospitals. Exploratory and confirmatory factor analyses were conducted on the CSES, revealing a three-factor model with strong validity (Kaiser-Meyer-Olkin (KMO) =0.956). The model exhibited good fit indices (χ² =61.432, df = 132, P = 1.000; root mean square error of approximation (RMSEA) =0.000; goodness of fit (GFI) =0.998; comparative fit index (CFI) =1.00; incremental fit index (IFI) =1.003; relative fit index (RFI) =0.997; normed fit index (NFI) =0.997). Reliability was high (Cronbach’s α =0.978), indicating robust internal consistency across factors. This instrument appears reliable and valid for assessing family physicians’ self-efficacy.

Conclusions:

Findings affirm the reliability and validity of the CSES instrument in assessing self-efficacy among family physicians, providing a valuable tool for understanding and enhancing their professional capabilities in diverse healthcare settings.

Introduction

The coronavirus pandemic is rapidly changing the world we live in. The recent coronavirus disease (COVID-19) outbreak has caused significant stress and psychological distress among healthcare professionals. This pandemic is causing fundamental changes, not only in primary healthcare but throughout society.[1] This may pose a risk for burnout syndrome, particularly in crisis situations, due to increased working hours, excessive workload, hazardous working conditions, work pace, and other similar factors.[2,3]

The health profession is characterized by working with people with problems and who are under intense stress and professionals also feel the stress of working alone with individual patients. Even without the disruptions in the healthcare system or patients’ excessive demands, stress can increase physicians’ anxiety and their tendency to burnout. As stress levels rise, job satisfaction, as well as belief in and commitment to the unit’s effectiveness, can decrease.[4]

According to the literature, the high work stress of family physicians causes problems such as burnout and decreased job satisfaction. Among the factors influencing their ability to cope with stress are physicians’ working conditions, their stress levels, and burnout.[5,6,7]

The ability of primary care physicians (PCPs) to cope effectively with stress is crucial for their professional performance and the quality of care they provide to their patients. As the first point of contact in the healthcare system, PCPs often face the challenges of long hours, high patient volumes, and managing a wide range of health concerns, all while fostering continuity of care and building strong patient relationships. Effective coping mechanisms enable them to navigate these demands while maintaining focus, making sound clinical decisions under pressure, and demonstrating empathy in their interactions. This directly enhances their capacity to deliver compassionate, patient-centered care, which is essential for building trust and improving patient satisfaction in the primary care setting. PCPs who manage their stress effectively are more likely to experience increased job satisfaction and professional fulfillment, which not only benefits their own well-being but also contributes to the overall efficiency and sustainability of the primary care system.

When dealing with stress, it is crucial to understand the physicians’ responses. Stress causes emotional, cognitive, behavioral, and physical responses. Emotional reactions include restlessness, anger, sadness, anxiety, hopelessness, and crying; cognitive reactions include difficulties in concentration, memory problems, instability, obsessions, and phobias. Examples of behavioral responses are avoidance, aggression, alcohol consumption, binge eating, and preoccupation with problems.[3]

The study aims to adapt the “Coping Self-Efficacy Scale” (CSES) into N language and to evaluate its validity and reliability among physicians working as family physicians in primary care in Turkey during the COVID-19 pandemic.

Methods

Participants

This is a correlational cross-sectional methodological study. The universe of research consisted of 2200 family physicians who are members of the N Association.

In methodological studies, it is recommended that the sample size is at least five to ten times greater than the number of variables, that is, the number of items, in testing the validity and reliability of measurement tools.[8,9,10] Generally, sample sizes of 300 or more are preferred.[11]

In this study, the sample size is 255, which seems to be suitable for total item score correlation. Those who voluntarily agreed to participate in this study were included; those who declined or did not answer all questions were excluded. Considering the conditions of the COVID-19 outbreak, the data were collected by an online survey via a Google Forms (Google LLC, Mountain View, CA, USA) between March and June of 2021. The study was conducted in accordance with the Declaration of Helsinki.

Measurements and definitions

In the study, data collection tools were the descriptive data record form and the CSES.

The descriptive data record form

It consists of 26 questions about the sociodemographic and individual characteristics of family physicians, as well as their professional experiences. The form was created by the researchers themselves, following the literature.

CSES

The scale was developed by Chesney et al.[12] in 2006.

Coping self-efficacy is a 26-item scale developed to evaluate participants’ perceived self-efficacy in coping with difficulties and threats.

This scale is a self-report tool, and each item represents a level of coping self-efficacy. Participants assign a score to each item ranging from 0 (“I cannot at all”) to 10 (“I absolutely can”). The item scores are added together to determine the respondent’s perceived coping self-efficacy level. There is no cutoff point on the scale. A higher level of total score indicates a high level of self-efficacy in coping. The adaptation process of the 26-item scale was followed in this study, in accordance with the recommendations of the authors who developed the CSES.

Data analysis

Statistical Package for the Social Sciences 28 (IBM Corp., Armonk, NY, USA) and JASP (JASP Team, University of Amsterdam, Amsterdam, The Netherlands; Version 0.18.1) were used in the statistical analyses. The description of participant information included frequency (%) for categorical values and mean and standard deviation (SD) for scale values. Kaiser-Meyer-Olkin (KMO) and Bartlett’s test were used to assess the appropriateness of data for factor analysis. Factor structure was investigated using exploratory factor analysis (EFA) with an ordinary least square factoring method with promax rotation based on a correlation matrix. A scree plot was generated, and parallel analysis based on factors was used to decide the number of factors. Factor loadings (>0.4) were listed. We then performed confirmatory factor analysis (CFA) to test the EFA-detected structure. Model indices were compared to find the best fit. The CFA used a diagonally weighted least squares (DWLS) estimator and standardized estimates were reported. Fit indices were model Chi-square (X2; P > 0.05), goodness of fit (GFI > 0.95), Tucker Lewis index (TLI > 0.95), comparative fit index (CFI > 0.90), root mean square error of approximation (RMSEA < 0.08), standardized root mean square residual (SRMR < 0.08). The reliability of the scale was assessed using Cronbach’s alpha, both for the total scale and for each individual factor.

Ethical aspect of research

Chesney was informed about the scale’s usage, and the necessary permission was obtained via e-mail. The Ethics Committee granted written permission for the study. The consent form on the first page of the questionnaire contained information about the research, including assurances of the right to refuse to participate in the research, and a statement that all participants’ personal information would be confidential. By checking “I agree” item, the participants confirmed reading and understanding the consent information and agreed to participate in this study on their own accord. After checking the “I agree” item, participants gained access to the rest of the survey.

Results

Sample characteristic

The mean age of the family physicians participating was 38.5 ± 10.1, and 59.2% were women. About 72.2% of those who took part reported being married, and 59.6%, having children. 39.2% of family physicians worked in a family health center, 38.0% in a university hospital, and 16.5% in a state hospital, with an average working experience of 13.7 ± 10.2. About 30.6 percent of doctors reported a chronic illness, 82.7 percent reported “poor-very bad” overall health, and 25.5 percent reported contracting COVID-19.

Validity analyses of CSES

Construct validity of CSES

The data were subjected to EFA and CFA in order to determine the construct validity of the scale. Before the EFA, the KMO value was 0.956 in terms of sample adequacy and the Barlett test result was X² =5829.573, P = 0.001. The KMO of all items was found to be greater than 0.90. As a result of factor analysis, three factors were identified. I1, I2, I6, I9, I13, I14, I16, I18, I25 were excluded due to cross-loading in multiple factors. Table 1 has summarized the measure of sampling adequacy (MSA) according to Bartlett’s analyses, and Table 2 and Figure 1 have summarized the three-factor models [Tables 1 and 2].

Table 1.

Sociodemographic Properties of Participants

Characteristics	n	%
Sex
Woman	151	59.2
Man	104	40.8
Age
Median±SD: 38.5±10.1	Min: 25	Max: 68

The work year has a mean of 17.88 and ranges from a minimum value of 1 to a maximum value of 42 years

Table 2.

Measure of Sampling Adequacy (MSA)

Item	MSA Value
Overall MSA	0.956
I3	0.969
I4	0.945
I5	0.971
I7	0.925
I8	0.939
I10	0.970
I11	0.952
I12	0.946
I15	0.964
I17	0.959
I19	0.961
I20	0.967
I21	0.951
I22	0.976
I23	0.939
I24	0.927
I26	0.973
Bartlett’s χ2 (P)	5829.573 (<0.001)

Figure 1.

Path diagram illustrating the factor structure of the model. Three latent factors (Fc1, Fc2, and Fc3) are shown, each with corresponding observed variables (items I3 to I26). The numbers on the arrows represent factor loadings, indicating the strength of the relationship between each factor and its associated items. The double-headed arrows between Fc1, Fc2, and Fc3 represent the correlations between the latent factors. The numbers inside the circles of observed variables indicate the error variances

The three-factor model’s calculated Chi-square value was 61.432, the degrees of freedom were 132, and P = 1.000. RMSEA 0.000, GFI 0.998, CFI 1.00, incremental fit index (IFI) 1.003, relative fit index (RFI) 0.097, and normed fit index (NFI) 0.997 were the fit indices. As a result of CFA, it was determined that the scale’s factor loads greater than 0.6090 [Table 3].

Table 3.

Model Fit Indices

	χ 2	SD	RMSEA	GFI	CFI	IFI	RFI	NFI
Three-factor model	61.432	132	0.000	0.998	1.000	1.003	0.997	0.997

Reliability of CSES

The Cronbach’s alpha coefficient for the entire scale was determined as 0.978. Factor 1 has a Cronbach’s alpha of 0.957. Factor 2 has a Cronbach’s alpha of 0.975. Factor 3 has a Cronbach’s alpha of 0.913.

Discussion

Validity analysis

Content validity of the scale

In the first stage of the original study of the scale, which established the N version’s validity and reliability, analyses were performed on 26 subitems. During correspondence with the authors, it was recommended that the validity and reliability of the N form be tested on 26 items. This recommendation was accepted, and after the analyses, the comparisons are discussed in light of the most recent findings of the original study. These findings revealed a high level of agreement among experts, that the scale items were appropriate for N culture, that the scale items represented the target to be measured. The findings supported the content validity of the N version of the CSES for a N sample and demonstrated content validity comparable to that of the original scale.[12]

Construct validity of the scale

The KMO and Bartlett sphericity tests were used to assess the data’s suitability and adequacy for factor analysis. For factor analysis, the Bartlett sphericity test value should be statistically significant and the KMO value should be at least 0.60 for both the whole scale and the item basis.[13,14] In this study, the KMO value was > 0.90 for the whole scale and for all items, and the Barlett test result was P < 0.005. These results showed that the data and sample size were suitable for factor analysis.[13,14]

First, I1, I2, I6, I9, I13, I14, I16, I18, and I25, found to have a low load in the exploratory factor analysis (EFA) and multiple correlations in the CFA were removed and analyzed. Parallel analysis based on factors was used to decide the number of factors.[15]

The results of our study showed that the factor loads of the scale were >0.90. Generally, the minimum factor load should be 0.30, and items below this value should be removed from the scale.[13,14] In this study, factor loads are > 0.30. After the validity and reliability studies of CSES, which, in its original form, has 26 items,[16] it was determined that the 1^st, 23^rd, and 26^th items were loaded in neither the (original form) nor in the short form.[12] The factor loads obtained in our study are >0.30 and showed that the scale has a strong factor structure.

The structure obtained by EFA should be analyzed by CFA.[17] CFA is a method based on the evaluation of the fit indices that show the fit between the data and the structure. An acceptable fit is considered to be a CFA fit index of x2/df <5, CFI, GFI, AGFI greater than. 90, and RMSEA less than. 08.[18]

The CFA results in the original scale were calculated on 26 items, then the third month and sixth month results were calculated on different items, and the results of this study were found similar to the CFA results calculated at different times. In addition, the determination with EFA for the N sample was found to be supported by CFA. Our study findings showed values of fit indices were within the desired limits.

The findings indicate that the scale demonstrates robust construct validity for use in assessing family physicians. CFA results support the three-factor structure proposed in the model, affirming that the items within the scale effectively capture and define their respective factors. EFA further bolsters the construct validity of the scale in this study. This suggests that the scale is not only applicable but also exhibits good validity when employed in the context of N family physicians.

Reliability analysis of the scale

Cronbach’s alpha coefficient shows whether the items measure the same feature and whether these are related to the subject. This value should be as close to one as possible. Values between 0·60 and 0·80 show the scale is quite reliable, and of between 0·80 and 1·00, highly reliable.[19,20] In our study, Cronbach’s α value of the scale was found to be >0·70 and thus showed a high level of reliability. These results revealed that the items were able to adequately measure the desired subject, that the items were related to the subject, and that the scale had a very good level of reliability. Cronbach’s alpha values of the scale were reduced to 17 items according to the EFA and CFA results of the original scale; these were determined as follows: using problem-focused coping (6 items, α = 0.91), stopping unpleasant emotions and thoughts (4 items, α = 0.91), and getting support from friends and family (3 items, α = 0.80). These were shown to be similar to our study results. In our reliability analysis, Cronbach’s alpha values for both the overall scale and its subgroups exceeded 0.90. This high level of internal consistency indicates strong reliability in our measurements for the general scale as well as its specific subcategories.

Coping self-efficacy, defined as an individual’s belief in their capacity to effectively manage challenging situations, particularly in the face of stress and adversity, emerged as a significant predictor of health-related behaviors and outcomes. This aligns existing literature that highlights the importance of self-efficacy in promoting mental health and overall well-being.

The role of PCPs is pivotal in this context. By employing strategies such as providing education about health conditions, offering resources for stress management, and encouraging personalized coping strategies, physicians can significantly enhance their patients’ coping capabilities. Actively involving patients in their care and decision-making processes fosters resilience, equipping individuals to confront life’s challenges with increased confidence and competence.

By addressing the psychological dimensions of health, PCPs can promote a more holistic approach to care. The interplay between coping self-efficacy and patient engagement highlights the importance of integrating psychological support into primary care practices.

Limitations of the study

The study utilized the same sample for both EFA and CFA. While this approach ensures consistency in the dataset, introducing a different sample for CFA could enhance the robustness of the results, providing a more comprehensive validation of the identified factor structure.

Additionally, it is essential to consider the context in which data collection occurred, particularly during the COVID-19 pandemic. Given the unprecedented challenges and stressors faced by healthcare professionals during this period, it is plausible that the reported CSES scores may be higher than the norm. Recognizing the potential influence of external factors on self-efficacy assessments becomes crucial for a nuanced interpretation of the findings and underscores the need for contextual sensitivity in the evaluation of results.

Future Aspects

This study provides a foundation for further exploration of coping self-efficacy among family physicians. Future research could focus on longitudinal studies to assess changes in coping mechanisms over time and the effectiveness of tailored interventions in reducing burnout. Expanding the use of the scale to other healthcare professionals and adapting it to different cultural contexts would enhance its applicability.

Conclusion

The findings of this study strongly affirm the reliability and validity of the CSES instrument as a robust tool for assessing self-efficacy among N family physicians. This instrument emerges as a valuable resource, offering insights into the professional capabilities of family physicians across diverse healthcare settings.

Ethical permissions from the Izmir University of Economics Medical Faculty Clinical Research Ethics Committee (B.30.2.EÜSB.0.05.05-20-129) were obtained for the study’s conduct. The research was carried out in accordance with the Helsinki Declaration. To conduct the research, permission was obtained from the TR Ministry of Health’s COVID-19 Scientific Research Evaluation Commission.

Authors contributions

Author 1: Conception and design of the study, data collection, data analysis and interpretation, manuscript drafting, and critical revision for important intellectual content.

Author 2: Development of methodology, data collection, statistical analysis, and manuscript drafting.

Author 3: Literature review, data interpretation, manuscript drafting, and approval of the final version.

Author 4: Data collection, technical support, and critical revision of the manuscript.

Conflicts ofinterest

There are no conflicts of interest.

Funding Statement

Nil.