Adolescents’ Psychological Well-Being: A Multidimensional Measure

Over the last few years, different theoretical approaches have emerged advocating for a positive understanding of adolescence, recognizing it as a stage characterized by plasticity, the acquisition of competences and the achievement of satisfactory levels of well-being and positive adjustment. Based on Ryff’s multidimensional model of psychological well-being, this study aims: (1) to develop an adjusted measuring instrument for adolescents (Brief Scale of Psychological Well-Being for Adolescents), as well as analysing its psychometric properties; and (2) to conduct a descriptive analysis of the levels of psychological and subjective well-being among adolescent boys and girls. A sample of 1590 Andalusian adolescents (51% girls), aged between 13 and 19 years old participated in this study. The Confirmatory Factor Analysis (CFA) showed the validity of the instrument, with a multidimensional factorial solution of four factors (self-acceptance, positive interpersonal relationships, autonomy and life development) with good levels of internal consistency. Descriptive analyses showed good scores of psychological and subjective well-being among the adolescents, with a significant impact of sex and age in both measures of well-being. The results are discussed in terms of the importance of considering adolescent well-being from a multidimensional view and the need to promote positive development from a multifactorial perspective which takes into account the diversity of the variables involved.


Introduction
Psychological research and practice have traditionally focused on studying and treating pathologies following a deficit approach, which has as its main priority to lessen the unpleasant effects of different individual and social problems [1]. This approach has been particularly evident in studies of adolescents, as this period of life involves many changes and has long been portrayed as a time of stress and hardship. This dramatic view of adolescence has been present since Hall put forward in his seminal work the well-known notion of storm and stress [2]. The idea of adolescence as a period in which young people go through emotional and behavioural upheaval [3] was present in subsequent scientific formulations. Anna Freud, for example, pointed out the inevitable problems arising in this life stage and viewed it as a universal period of developmental disturbance [4]. Similarly, Erikson established the identity crisis as a characteristic feature of adolescence [5,6]. In this way, during the first half of the 20th century, most writing and research about adolescence was based on this deficit conception of young people [7], which still lingers on in current social representations [8]. In the same vein, a number of studies have alluded to the increased conflict with parents [9,10], emotional instability [11,12] or risk behaviour [13], which occur during these years and which have led to our image of adolescence as a period of generalized problems. This approach, however, may have had negative consequences for the boys and girls themselves [8]: the lack of empirical and social attention given to the opportunities that this stage in life offers, as well as the contributions that adolescents can make both to the environment and to themselves, may have built up a deep-set negative bias in our understanding of this stage of development.

The Well-Being of Adolescents
To counter the traumatic, conflictive image traditionally associated with adolescence, a stream of studies has emerged in recent years which portray the adolescent not as a source of problems but rather as a valuable asset in a development process [14][15][16]. One example of these alternative approaches is the positive youth development (PYD) perspective. This strength-based view of adolescents [17], points out that the multiple changes that occur during adolescence represent the plasticity inherent in the system of development [18], which represents a fundamental strength of the adolescent period [19], in that it provides the potential for positive functioning [17]. One essential part of this is promoting beneficial relationships between adolescents and their environment, since such adaptive regulations increase the likelihood that healthy and positive changes will take place [17]. Therefore, this perspective encourages us to study adolescents from a broader perspective, such as the development of human potentialities in different areas-physical, psychic and social-which led to a positive view of their general health, rather than focusing on the presence or lack of diseases or disorders [20]. A close empirical perspective to PYD promotion is the one provided by Positive Psychology in relation to the well-being construct. During the last decade, this field has stimulated considerable research aimed at redressing the scientific and practical imbalance in psychopathology between strengths and well-being [21], while paying greater attention to the complementary study of development and the enhancement of human strengths and environmental resources [22]. Nonetheless, most research into human well-being has been carried out on the adult population. During adolescence, well-being seems to have its own developmental idiosyncrasy and differs significantly from later ages [23]. The manifold changes that occur at these ages (physical, physiological, cognitive, emotional, behavioural, social, relational and institutional), make adolescence an exceptionally plastic period of life [24], in which the interaction with family, peers and the community should be considered as essential aspects of positive development, capable of generating well-being [25] and as good indicators of positive adjustment at these ages. A society which has healthy, dynamic and cooperative young people is essential for our future economic and social well-being [26] and so identifying and investing in the factors which contribute to the promotion of well-being at pre-adult ages continues to be an important area of research [27].

The Need for Validated Research Tools
In operative terms, the conceptual delimitation of this construct is complex, which is why most studies argue in favour of recognizing two perspectives: the hedonic and the eudaimonic [28]. In the first case, the predominant concept is subjective well-being [29], understood in terms of the balance between positive and negative affect and satisfaction with life. Subjective well-being is therefore made up of both an affective and a cognitive component [30]. The eudaimonic angle, on the other hand, puts the focus on psychological well-being and considers the deployment of skills and personal growth as the main indicators of positive functioning [31]. Although, originally, these two perspectives were conceived as independent [32], the most recent works tend to consider that the measurements carried out under the hedonic and eudaimonic concepts result in different but interrelated constructs [30]. As a result, a blend of both perspectives seems to be the most suitable way to evaluate how people function best, whatever the age but in particular during adolescence [32].
Currently, the topic of subjective well-being has received the most empirical support and has been analysed in different stages of life with research tools validated in different languages [33][34][35][36][37][38]. Self-report questionnaires are among the most commonly used measurements to evaluate positive and negative emotions [39], such as the Positive and Negative Affect Schedule (PANAS) [40], or the Multiple Affect Adjective Checklist-R (MAAC-R) [41]. Subjective well-being has also been measured by evaluating satisfaction with life, with instruments such as the Satisfaction With Life Scale (SWLS) [42] and the Personal Wellbeing Index (PWI) [43]. All these instruments have been widely used in the study of adolescent well-being [44][45][46][47][48][49][50]; however, very few studies address the measurement of psychological well-being in adolescence. Consequently, few research tools are backed up by solid empirical support to measure this construct at this stage of development. The increasing interest in this area of research therefore calls for valid and reliable instruments that measure the essential constructs of positive functioning [51].

The BSPWB-A Questionnaire Development: Theoretical Principles
Ryff's [52,53] theoretical contributions regarding psychological well-being have laid the foundations for developing one of the most commonly used instruments for measuring it [54,55]. On the basis of perspectives such as Maslow's conception of self-actualization [56], Rogers's view of the fully functioning person [57], Allport's conception of maturity [58] or Erikson's psychosocial stage model [5] (for review, see [52,59]), Ryff proposed six dimensions to put into practice and measure the eudaimonic approach to the concept of well-being [60]: (1) autonomy, understood by the capacity to have the strength to follow personal convictions, even if they go against conventional wisdom; (2) environmental mastery, which includes being able to manage the demands of everyday life; (3) personal growth, understood by feeling that personal talents and potential are being fulfilled over time; (4) positive relationships with others, in terms of close, valued significant connections with others; (5) purpose in life, in other words, to have goals and objectives that give life meaning and direction; and (6) self-acceptance, or the capacity to see and accept one's strengths and weaknesses (see [52,53] for complete descriptions). Based on these categories, Ryff [52] developed the Scales of Psychological Well-Being (SPWB), a 120-item research tool targeted at the adult population. Given the difficulties involved in using such an extensive questionnaire, later versions with 84, 54 and 18 items were developed [61,62].
Despite the fact that this research tool has been widely used [63][64][65][66][67], van Dierendonck [51] concluded that to achieve good internal consistency, the scales had to be longer than in the 18-item version but factorial validity required that they be shorter than in the 84-and 54-item versions. Therefore, van Dierendonck presented a new 39-item version [51], whose psychometric properties were improved with respect to previous versions, while keeping the theoretical model of six dimensions proposed by Ryff [52]. Following this version, Díaz et al. [31] developed a Spanish adaptation of this scale for an adult population, producing a shorter, 29-item version validated in Spanish. Despite having fewer items, the scales showed an internal consistency similar to that seen in the scales proposed by van Dierendonck [51] and kept the 6 original dimensions proposed by Ryff [52]. These scales, either in the 39-item or 29-item version, have been used with adolescent samples [68][69][70] but there are still very few studies which analyse the psychological well-being in this specific stage of life. On the one hand, there is a single, short 13-item scale (the Psychological Well-Being Scale, known as 'BIEPS' in Spanish) [71,72] to evaluate psychological well-being during adolescence, although there is limited empirical evidence for its reliability, validity and adjustment to the original theoretical model proposed by Ryff [52]. On the other, another adaptation for adolescents of Ryff's scales was proposed in the study carried out by Loera-Malvaez et al. [70]. These authors produced a 4-factor solution using the 39-item version which coincides partly with the categories designed by Ryff [52,53,60]: self-acceptance, interpersonal relationships, autonomy and satisfaction with life. These authors point out that, at these ages, the differentiation between aspects such as control over one's environment or having a purpose in life cannot be measured properly, as boys and girls of this age group find it hard to differentiate between these categories. Other authors point also to the high correlations between these scales and those of self-acceptance and personal growth, which suggests the existence of a degree of overlap and lack of empirical differentiation between them [51,52,61,63,73].

The Present Study
Although the study of eudaimonic well-being has been given a strong empirical impulse in recent years [31,52,53,61,[74][75][76][77] there are still gaps in our knowledge which need to be filled. Currently, as far as we know, there is no short research tool, translated into Spanish and validated with adolescents, which enables us to carry out studies linking Psychological Well-being with other personal and contextual variables in boys and girls of these ages. For this reason, the purpose of this work was to develop an instrument for measuring psychological well-being which would meet three basic requirements: (1) it should be in Spanish; (2) it should be adapted to the adolescent population; (3) it should be a short validated version which can be used together with other measuring instruments.
Therefore, the objective of this work is twofold: (1) to design a short research tool for measuring well-being which is in Spanish and is adjusted for adolescents (Brief Psychological Well-Being Scale for Adolescents, BSPWB-A) and analyse its psychometric properties; and (2) to carry out a descriptive analysis of both psychological and subjective well-being in a sample of Spanish adolescents.

Participants
This study was carried out in two stages (pilot scheme and final implementation). A total of 1590 adolescents (49% boys, 51% girls) took part in the final implementation stage, all of whom attended school in state and private schools in the Autonomous Community of Andalusia (Spain) and were aged between 13 and 19 years old (mean age = 15.34; SD = 1.24).

Measures
The research tool we designed was made up of the following scales: Socio-metric data: a series of open questions which collected information on the personal variables of the participants: sex, age, type of school, school ownership, course and location.
Scale of subjective well-being: the Personal Wellbeing Index (PWI-7) [78] was used, in its adapted version in Spanish for adolescent populations [34]. The scale consists of seven items assessed on a 10-point Likert scale ranging from no satisfaction at all to completely satisfied, which measure adolescents' satisfaction with different areas of their lives: health, standard of living, achievements in life, personal safety, groups of belonging, security for the future and interpersonal relationships. The average scores derived from the PWI-7 constitute a measure of Subjective Well-Being (α = 0.88). In order to calculate and interpret the global index of subjective well-being, the scores are standardized; the normal range for Western populations is between 70 and 80 points [43].
Scale of psychological well-being: BSPWB-A. The adaptation of this instrument was the first stage of this study. Using the version by van Dierendonck [51] as a starting point, the following two adaptations in Spanish have been used: (1) the short 29-item version aimed at the adult population developed by Díaz et al. [31] and (2) the longer, 34-item version validated with an adolescent population developed by Loera-Malvaez et al. [70]. We worked on both questionnaires and all the necessary adaptations and combinations were carried out before ending up with a final version which met all the established requirements (short version, in Spanish, for adolescents). A preliminary 21-item version of the BSPWB-A was obtained and subjected to a piloting process to assess its suitability in terms of linguistic, semantic and functional validity [79]. In this pilot study, 197 adolescents (51% boys, 49% girls) participated, all of whom were enrolled in state schools in the province of Córdoba (Andalusia, Spain) and aged between 12 and 18 years old (mean age = 15.14; SD = 2.48). This version kept the 4 scales which have been identified as necessary for measuring psychological well-being by Loera-Malvaez et al. [70], although by making a few adjustments to the theory and the content, some of the categories were renamed. The final version of the instrument was taken from the pilot study and included a total of 20 items, after one item, which negatively affected the questionnaire's internal consistency, was deleted. Six more items were reformulated after finding that the participants did not understand what they meant. Finally, taking the model established by Loera-Malvaez et al. [70] as a reference, we decided to validate a 4-factor model which allowed us to identify levels of well-being. The final scale was entitled the Brief Scale of Psychological Well-Being for Adolescents (BSPWB-A), which is an adaptation of the Psychological Well-Being Scales developed by Ryff [52]. The final questionnaire and its relationship with the scales which were used as reference values can be found in Appendix A. The scale is made up of 20 items assessed on a 6-point Likert scale ranging from completely disagree to completely agree, which measure the degree of agreement or disagreement with different statements about 4 areas: self-acceptance, positive interpersonal relationships, autonomy and life development.

Procedure
The research design was cross-sectional, ex-post facto and retrospective, with a single group and multiple measurements [80]. Before the data was collected, the participants' families gave their informed consent and permission was given by the school. The students answered the questions during school hours in groups: they had previously been informed that their participation would be anonymous, confidential and voluntary. The average time taken to complete the questionnaire was 20 min. The data collection process was carried out by research assistants, who did not belong to the schools. The study was carried out in accordance with the Declaration of Helsinki. Ethics approval was obtained from the Research Ethics Committee of the University of Cordoba, Spain.

Data Analysis
The descriptive and correlational analyses were carried out using the SPSS 20.0 statistical software program (IBM, Armonk, NY, USA), while EQS 6.2 was used for the factorial validity analyses (Multivariate Software, Inc., Encino, CA, USA).
For the Confirmatory Factorial Analysis (CFA), the works of Diaz et al. [31] and Loera-Malvaez et al. [70] were used as a theoretical framework and the suitability of their models was tested with the BSPWB-A. As the instrument was measured on a 6-point Likert scale, we decided to apply the corresponding polychoric correlations [81]. The Maximum Likelihood (ML) estimation method was used, considering that, when used with structural equation models, it was shown to be fairly robust in situations where basic assumptions such as multivariate normality are violated [82]. However, with a Mardia coefficient which indicated an abnormal multivariate distribution, the data analysis was performed with robust coefficients. To evaluate the model, the Satorra-Bentler parameter values were used, corrected according to a robust covariance matrix, as these were seen to provide the most reliable adjustment statistics [83]. In addition to the Satorra-Bentler Chi-Square value, the adjustment values of these models suit the values of the comparative fit index (CFI), the Bentler-Bonett non-normality fit index (NNFI) (≥0.90 suitable, ≥0.95 optimum) and the root mean square error of approximation (RMSEA) (≤0.08 adequate, ≤0.05 optimal) [84]. Finally, we estimated the internal validity of each factor, in addition to the total scale with the ordinal Alpha [85], using a minimum cut-off point of 0.70 [86]. To evaluate convergent validity, the standardized factorial weights were examined, following the criterion of Worthington & Whittaker [87] (values > 0.40 indicate a good convergent validity).
Descriptive analyses on well-being were performed. Comparison analyses with age and gender were also carried out using ANOVA and Student's t-test, since these two tests are both robust in cases of non-compliance with normality [88]. The effect size was controlled using the Cohen's d and Eta-Squared statistics.

Validation of the Instrument: Brief Scale of Psychological Well-Being for Adolescents (BSPWB-A).
The final 20-item version of the BSPWB-A was validated, after carrying out a CFA. A Mardia coefficient score of 115.6968 showed that the data did not meet the assumption of multivariate normality. Inter-item polychoric correlation analyses showed low-mid values, which was indicative of low collinearity (see Table 1).  αt otal well-being = 0.95).
All the items showed high values in terms of standardized factorial weight, which was over 0.49 and significant in all cases, which indicated a good convergent validity (see Table 2). The low but significant covariances between the factors indicated related but not overlapping dimensions.

Psychological and Subjective Well-being in Adolescence by Age and Gender
The second main objective was to make progress in the analysis of the levels of psychological and subjective well-being found in adolescent boys and girls, taking into account any possible differences depending on sex or age. The data produced medium-high scores on all the psychological or subjective welfare scales considered.
The analysis of the influence of gender showed significant differences in the scales of subjective well-being (see Table 3  Age also proved to have a significant impact on adolescent well-being, with differences in the subjective well-being scales of satisfaction with one's standard of living (F (1559,1564) = 2.35; p = 0.039,  Table 4), except for the life development scale, in which older adolescents obtained the highest scores.

Discussion
Scientific interest in the positive aspects of human functioning has increased considerably in recent decades, giving rise to new definitions, models and measurements of a concept that is still rather ambiguous but is currently understood by well-being [89][90][91]. There has been a marked increase in studies which call for a new perspective on the psycho-evolutionary and change process of adolescence, which highlights the human potential and the values of growth and a positive outlook on life, even in a period which has nearly always previously been looked on as problematic [9][10][11]. This positive approach to adolescence raises the notion that even difficulties and challenges are factors which could contribute to the pursuit of happiness and lead to positive changes [14][15][16][17]19]. However, the study of personal and social life during adolescence has been traditionally plagued with an imbalance towards identifying pathologies and damage-repair, following a model of pathology and conflicts and neglecting any potential for trying to induce young people to be well integrated in their communities or well oriented towards the pursuit of happiness [92]. Due to this, these studies still have difficulties to overcome: from the difficulty of defining key constructs such as well-being, to the lack of suitable research tools which could enable us to obtain valid and reliable scientific evidence from these constructs.

The BSPWB-A Validation
With the aim of contributing to the progress in this field of study, the main objective of this work was to design and validate a brief research tool in Spanish to measure psychological well-being in the adolescent population, as well as analysing its psychometric properties. To do this, we based our study on the adaptations made by Díaz et al. [31] and Loera-Malvaez et al. [70] of the Psychological Well-being Scales developed by Ryff [52]. The results obtained support the BSPWB-A as a valid and reliable instrument, with good indexes of internal consistency and a factorial adjustment with a 4-factor model based on the scales of self-acceptance, positive relationships, autonomy and personal growth proposed by Ryff [52]. Because of the high correlations shown in previous studies [51,52,61,63,73], the scales of environmental mastery and purpose in life proposed by Ryff [52] were not included. Loera-Malvaez et al. [70] also indicated that these dimensions, due to their content, could be inappropriate in studies with adolescents. The instrument presented in this work showed better internal consistency indices in comparison with both versions used as references, specifically with respect to the autonomy and personal growth (life development in the BSPWB-A) scales [31,70]. In addition, the self-acceptance scale of BSPWB-A had better internal consistency than the version developed by Díaz et al. [31], as did the positive interpersonal relationships scale in comparison with the study carried out by Loera-Malvaez et al. [70]. It is important to highlight that, in these two studies, the personal growth scale showed the lowest alpha coefficients. Our version therefore appears suitable for use with adolescent samples.

Adolescent Well-Being: Associations with Gender and Age
Taking into account the studies which suggest that measuring both psychological and subjective well-being together is the most suitable way of evaluating positive functioning [32] and that both processes of well-being can work together [76], we established as the second main objective the analysis of the rates of psychological and subjective well-being reported by boys and girls of these ages: high scores on both scales would indicate optimal functioning [32] and a better personal adjustment. In this work, the results showed that the average level of the adolescents' well-being met the requirements to be considered satisfactory, with subjective well-being scores within the established normative range for the western population and medium-high scores for psychological well-being. Although no studies were found which use a similar joint approach to well-being in adolescence, including both subjective and psychological well-being measurements, prior studies have produced similar results regarding subjective well-being [93,94]. The gender differences observed suggest that boys express a higher level of general well-being than girls, although the girls score higher in the items referring to their own life development and satisfaction with life. Previous studies, however, showed no significant gender differences regarding subjective well-being [95,96]. With regard to psychological well-being, studies with adult samples showed that women consistently rate themselves higher in positive relations with others than men and they tend to score higher than men on personal growth [52,61,97], although no studies were found using adolescent samples. In our study, girls scored significantly higher that boys in the life development scale (corresponding to personal growth in Ryff's scale), which could point to a possible developmental trend. Further longitudinal studies should be carried out to verify if there are, indeed, gender differences in well-being across age.
As regards age, there were important differences between the younger age group (representing the first stage of adolescence: 13, 14 and 15 years old) and the older age group (representing the phase of late adolescence or the beginning of adulthood: aged 16, 17, 18 or over). The former showed significantly higher levels of well-being than the latter, which could be explained by the different psycho-evolutionary characteristics between the age groups of young adolescents and early adulthood, such as leaving home, looking for work or making future plans as an adult, which can play an important role in the awareness of the contextual variables which facilitate or hinder the achievement of such objectives and have a direct impact on the feeling of well-being shown. Overcoming the psycho-evolutionary tasks presented in this period of the life cycle, some of which are directly linked to the economic, educational, work and social opportunities offered by their social context, leads to improved levels of well-being and personal satisfaction [98]. The decrease in well-being with age found in this work is consistently with other studies of subjective well-being in adolescence [99]. However, no studies have been found analysing age differences in psychological well-being. On this subject, studies with adult samples have suggested that certain aspects of well-being, such as autonomy and environmental mastery, increase with age, while others decrease, such as personal growth and purpose in life [52,61,97,100]. Nonetheless, our results showed an increase in personal growth with age (or life development).
These findings suggest, on the one hand, that adolescence may not be the problematic and turbulent stage of life that still features so prominently in current social representations and research [7,8], lending therefore empirical support to current theories and models of positive development and well-being during adolescence. On the other hand, they promote the idea of adolescence as a period of life in which valuable resources are created [14][15][16] and emphasizing the importance of the existence of beneficial, mutual and healthy links with the environment in which adolescents grow up.

Limitations and Future Directions
Despite all this, it is important to note the limitations of this study. The use of self-report questionnaires on aspects related to happiness or well-being could be affected by what some authors refer to as a positivist bias, or a tendency to overestimate the real values of well-being [101,102]. Although this bias has proved to be cross-cultural, future studies using the BSPWB-A in other populations than the Spanish one could help us, on the one hand, to understand how marked this effect is in the study population and on the other, to prove its psychometric properties and invariance to other samples, thus contributing to the generalization of the results. The cross-sectional nature of this work may represent another limitation. Thus, it would also be interesting to develop longitudinal studies which enable us to assess the role of well-being as a cause or consequence of some of the most widely used contextual variables linked to risk and deficit. Finally, future studies including both subjective and psychological well-being measurements will be needed to advance our scientific knowledge about the positive aspects of adolescent development. According to the literature, both types of well-being can operate in tandem, providing evidence that the best psychological outcomes can be obtained through the synergy of both hedonic and eudaimonic processes [32,76,103]. The use of distinct labels and terms therefore leads to thinking about constructs that are separated from one another, which may represent a major obstacle in the progress towards the understanding of positive aspects of human functioning.

Conclusions
This study supports the idea of considering adolescence as a stage of opportunities and positive development, which in turn allows contemporary systemic models of development to be adopted, focusing on the study of the links between the individual and their context [24]. The focus on the organism-context as a unit of analysis [104] works on the principle that the process of human development involves mutually influential relationships with the environment [105,106] which, when mutually beneficial, become adaptive regulations for development [105,[107][108][109]. Although relationships with our environment occur throughout our lives [110,111], they are of special importance in the study of well-being and positive functioning during adolescence, given the plasticity which characterizes this stage of human development [24]. The promotion of well-being during adolescence can help to obtain positive results, as well as acting as a buffer against negative results, such as psychological disorders [112]. In this way, well-being not only represents a key indicator of positive development, but it can also serve to ensure optimal mental health [113] and to discover the beneficial adjustment pathways between adolescents and their context, leading to greater chances of achieving positive changes in the transition to adulthood [114][115][116]. Acknowledgments: The authors thanks to Eva Romera for the project administration and the students who participated in this study.

Conflicts of Interest:
The authors declare no conflict of interest. 23. Tengo clara la dirección y el objetivo de mi vida (moved to self-acceptance)

Appendix A. Reference Scales and Final Questionnaire: Brief Scale of Psychological Well-Being for Adolescents (BSPWB-A)
Note: The changed items are in italics; the inversed score items are in bold; the item deleted after the pilot study is in between parenthesis. 1 In 34 items version for adolescents [70], this factor is named Interpersonal Relationships; in the validated version of BSPWB-A it is called Positive Interpersonal Relationships. 2 In 34 items version for adolescents [70], this factor is named Satisfaction with life; in the validated version of BSPWB-A it is called Life development. * Labels of Ryff's original scales [53].