INTRODUCTION

Psychological well-being is a critical element in students’ performance, particularly in higher education. In the context of a medical degree, which is academically demanding, it can be a determining factor. The Ryff Psychological Well-Being Scale is the most widely used instrument for assessing the eudemonic perspective of well-being. Although it has been adapted for the Spanish population (^{Díaz et al, 2006}), it has not been modified for health science students from the Mexican population.

In this regard, ^{Freire, MarFerradás, Núñez, and Valle (2018)} define psychological well-being as the acquisition of values that enable our self-fulfillment. It is a construct based on the Aristotelian conception of happiness, understood as the achievement of excellence or perfection in oneself, according to individual capabilities and potential.

^{Ryan and Deci (2001)} identify two main strands in the literature: the hedonic tradition, associated with happiness, and the eudemonic tradition, linked to the development and expression of human potential, characterized by feeling positive and thinking constructively about oneself. The latter is characterized by its subjective, experiential, and relational nature, in which specific aspects of physical, mental, and social functioning are interrelated.

Psychological well-being (^{Cuadros, 2019}) makes a person aware of their potential and enables them to cope with the concerns of life and undertake everyday tasks successfully, contributing to their community with a sense of reciprocity and self-fulfillment and acceptance.

^{Medina (2015)} defines it as “functioning or doing well” as opposed to just “feeling well”. In 1989, Ryff proposed the Psychological Well-Being (PWB) scale in a study conducted with 321 subjects, in which she proposed a final structure, in a Likert question format evaluating six dimensions; 1. Self-acceptance, 2. Environmental mastery, 3. Positive relations with others, 4. Personal growth, 5. Autonomy, and 6. Purpose in life. These dimensions are characterized as follows (^{Ryff & Keyes, 1995}):

Growing interest in measuring psychological well-being has led to parts of the world adapting this scale with various samples and countries. Particularly with students, this has required both translating the scale into Spanish and adapting it, with an emphasis on specific dimensions depending on the social context of the subjects (^{Bahamón, Alarcón-Vásquez, Cudris Torres, Trejos-Herrera, & Campo Aráuz, 2019}; ^{Loera-Malvaez, Balcázar-Nava, Trejo-González, Gurrola-Peña, & Bonilla-Muñoz, 2008}; ^{Millán de Lange, García-Álvarez, & D’Aubeterre López, 2014}; ^{Véliz-Burgos, 2012}; ^{Medina-Calvillo, Gutiérrez-Hernández, & Padrós-Blázquez, 2013}; ^{Díaz et al., 2006}; ^{Lindfors, Berntson, & Lundberg, 2006}; ^{Kishida et al., 2004}; ^{van Dierendonck, 2004}).

Given the goal of adapting the scale to Mexican medical students, the study conducted by ^{Díaz et al. (2006)} and undertaken by the universities of Madrid and Rotterdam with 467 students is of interest. The two institutions adopted the van Dierendonck version of the psychological well-being scales proposed by Carol Ryff. Since a satisfactory fit was not obtained with the six original dimensions, it was reduced to twenty-nine items that allowed a better fit of the model (^{van Dierendonck, 2004}).

^{Aranguren and Irrazabal (2015)} subsequently conducted a psychometric analysis of the scale adapted by van Dierendonck, and the scale adapted by Diaz and collaborators, in 396 Argentine students, since the original version proved unsuitable for either population. A twenty-nine-item proposal was designed, comprising three dimensions: autonomy, positive relationships with others and competence, with an adequate fit and reliability values.

Pineda and collaborators evaluated the psychometric quality of a thirty-nine-item version used with a sample of 727 Colombian young adults. They obtained adequate psychometric values and found that the instrument was sensitive enough to discriminate between the results obtained based on educational attainment, gender, and socioeconomic status in several of the dimensions of the scale (^{Pineda Roa, Castro Muñoz, & Chaparro Clavijo, 2018}).

More recently, Meier and Oros adapted the version modified by Diez and collaborators in two stages: first with eighty adolescents to determine the linguistic coherence between Spain and Argentina and second with 825 adolescents to determine its psychometric quality. They obtained a twenty-item structure with four factors: a) Self-acceptance, b) Personal Growth and Purpose in Life, c) Autonomy and d) Positive Relationships with Other People (^{Meier & Oros, 2019}).

In Mexico, Dominguez and collaborators used the original version of the Ryff Psychological Well-Being Scale with 1,700 students from Veracruz, analyzed the different structures proposed in the literature, and only reported a better fit in a one-dimensional model (^{Dominguez-Lara et al., 2019}).

Psychological well-being is one of the main elements linked to academic performance as reported in various studies, in that it is a multidimensional phenomenon with extra-academic aspects that impact students (^{Cuadros, 2019}; ^{Dominguez-Lara & Navarro-Loli, 2018}).

It is also a determining factor in the achievement of the personal and professional goals of university students and their best academic performance (^{Cuadros, 2019}; ^{Correa Reyes, Cuevas Martínez, & Villaseñor Ponce, 2017}).

Due to the above and in the context of medical degrees, it is important to have an instrument specifically adapted for its assessment. From the moment they begin their degree course, high school students face major challenges and are forced to adapt. Although they are usually high-performing students, many of them drop out or fall behind in the first two years of their degree because of the demands and academic workload involved. Unlike other professions, the atypical conditions of training in clinical settings have a major impact on the psychological well-being of students during the clinical cycle, undergraduate internship and graduation, and even during their social service.

The objective of this study is therefore to adapt the Psychological Well-Being Scale in the version modified by Díaz and collaborators for students pursuing a bachelor’s degree in medicine at the medical faculty of the National Autonomous University of Mexico (Universidad Nacional Autónoma de México, UNAM) and the Mexican School of Medicine at La Salle University.

METHOD

RESULTS

The instrument was answered by 1,974 students, 63.7% of whom were women and 36.3% men, with a mean age of 23.6 ± 1.75 years, within an age range of seventeen to twenty-seven. Cronbach’s alpha coefficient was calculated to determine the reliability of the instrument, with a standardized value of α equal to .89 being obtained for the instrument as a whole.

An exploratory factor analysis of principal components with Varimax rotation was subsequently performed to determine its structure. The use of orthogonal rather than oblique rotation is due to the fact that as a construct, psychological well-being has various dimensions, characterized by a set of indicators that would theoretically be assumed to be independent of each other.

An analysis of the correlation matrix yielded low to moderate correlations, showing that there was no collinearity between the items. The values of the fit to the structure obtained were adequate, with a KMO of .902, which was significant. Regarding communalities, it was observed that the proportion of explained variance for each item was greater than .40, except for item 23: “If I strive to achieve my goals, I will achieve them,” which only achieved .34, which is why it was removed from the proposed version.

The values of the reproduced correlation matrix, determined by the factor solution, are a reliable indicator that the resulting model was good, and that there were an appropriate number of factors when 100% of the results of the extraction of each statement were reproduced, yielding residuals equal to zero.

In regard to total explained variance, the first four factors were taken into account with a percentage of 54.89%, with the values for each dimension being 30.52%, 11.42%, 7.57% and 5.37% respectively. A four-dimensional structure was obtained, as can be seen in Table 1, showing the dimensions, component items, and factor loads of reliability and discrimination by factor. Items 7, 8, 9, 10, 11, 20, and 21 are not integrated into any factor.    

Discrimination was determined by establishing a cut-off point at the .33 and .66 percentiles. A lower group was obtained in which 651 records were located, with an average of 52.81 ± 5.03 points over the raw score, and a higher group, comprising 742 records, with a mean of 69.27 ± 3.22. Both groups were contrasted using a Student’s t-test, with a significant discrimination of the instrument of [t = -67.18, r < .01] being found. When discrimination was analyzed for each dimension and each of the twenty-one statements comprising the adapted version of the instrument, significant discrimination was found with r < .01 for each of them.

A confirmatory factor analysis was undertaken using a structural equations model developed with AMOS, obtaining a Chi² value of 4.12 (r < .01). The items comprising each dimension achieved significant standardized regression weights, ranging from .571 to .978. The value of the root mean square error of approximation, RMSEA, was considered since, according to ^{Hair, Anderson, Tatham, and Black (2007)}, it corrects the tendency of the Chi² value to reject models with a large n. It has an expected goodness of fit value with the population, obtaining a value of .033 (which is acceptable, since it is less than .08) while the normal fit index (NFI) obtained was .912 (which is adequate, since it is close to 1).

When the version obtained was contrasted with the original structure proposed by ^{Díaz et al. (2006)} in the items comprising each factor, the following changes were observed for the adapted version, as shown in table 2.    

DISCUSSION AND CONCLUSION

It is important to note that any assessment instrument must be suitable for the target population. Hence the importance of adapting, adjusting or validating the instruments according to the needs of the different protocols, and to ensure that they comply with certain psychometric standards that guarantee their impact on the target population, in this case medical students.

This is borne out by the various versions derived from the construction of the Ryff Psychological Well-Being Scale, foremost among which are the following:

Having a scale with these characteristics is important, since psychological well-being affects the variables directly related to student performance, such as stress, coping, motivation, life satisfaction, social well-being, and depression (^{Moreta, Gabior, & Barrera, 2017}; ^{Correa Reyes et al., 2017}; ^{Matalinares et al., 2016}).

Although there are multiple proposals, the fact that every instrument must be appropriate for the characteristics of the target population must be considered, to prevent biases and unsupported interpretations in the interpretation of results, particularly in the area of health, given the profile of the future professionals evaluated and the impact of their practice on the population.

The data obtained in this research allow us to present the adapted version of the Psychological Well-being Scale as a standardized instrument for the population of Mexican medical students in public and private settings for which there is solid evidence of reliability and validity. A four-dimensional structure was obtained, which can be defined as follows:

These dimensions are pertinent and adapted to the environment of undergraduate medicine students, enabling one to have a standardized instrument that contributes to a better assessment of Mexican medical students in regard to critical variables from their admission to their graduation. They can contribute to the development of lines of research and the psychoeducational and training context in training scenarios in both an academic and/or clinical context that will enhance the development of various lines of research.