Transcultural adaptation and psychometric properties of the Spanish version of the Life Skills Scale For Sport: LSSS-S

Objective: The aim of this study was to translate and validate into Spanish the Life Skills Scale for Sport (LSSS) designed and validated by Cronin and Allen. The scale is divided into eight sub-dimensions: goal-seeking, teamwork, time management, emotional skills, interpersonal communication, social skills, leadership, and problem solving. Method: For the translation process, a validation between languages was performed. A double procedure was carried out with bilingual researchers related to the subject, translating from English to Spanish and, conversely. Subsequently for validation, 1251 students (59.8% women) between 15 and 29 years old from Spain and Chile answered the questionnaire. Results: Comparing initial results, the translated terms matched, and no relevant mismatch was found. The confirmatory factor analysis supported the factorial structure proposed by the original authors, showing acceptable fit indicators (RMSEA: 0.048; CFI: 0.93; TLI: 0.92), satisfactory internal consistency coefficient (0.980) and adequate item-total correlation (0.630). Conclusion: The instrument proves to be consistent in the evaluation of life skills in the Spanish-speaking population, presents adequate psychometric properties and is a reliable and valid instrument for the evaluation of life skills in contexts of physical activity and sport with a young population aged 15 to 29 years.


Introduction
Currently, we live in a context of constant change and uncertainty that involves facing new challenges in education 1 . These new demands have caused numerous authors to focus their efforts on determining what competencies or skills must be learned to face the challenges of a fast-paced society, and that they also prepare individuals to be citizens and function properly in their adult lives.
Focusing in Life Skills (LS), the definition varies across disciplines 2 . For instance, to de World Health Organization 3 , LS are defined as the capacity to adopt a positive behavior that allows to address the challenges of daily life and maintain a state of mental well-being through a positive and adaptable behavior in the interaction with other people and with the social and cultural environment. The development of these competences would favor in individuals a positive management and control of their own health and resistance to social-group pressure 3 . In the context of Physical Education and Sport, definitions close to the above can be found. For Danish et al. 4 LS are those that allow us to deal with the demands and challenges of everyday life. Such skills would allow young people to succeed in different contexts that are part of their daily lives, such as school, their own home and neighborhood. Other authors 5 also point out its importance to function in the work environment. These skills would be behavioral (e.g., effective communication), cognitive (e.g., effective decision-making), interpersonal (e.g., assertiveness), and intrapersonal (e.g., goal setting) Likewise, terms have been used to describe LS such as socialemotional learning, emotional intelligence, positive psychology, resilience and character education 6 .
Childhood and adolescence are critical periods for their development. LS acquisition will influence decisions and trajectories about your health and well-being 7,8 . Regarding the young adult population, although the acquisition and improvement of these skills can be done throughout life, there are few programs and specific curricular configurations that include them, and these are mainly aimed at reducing risky behaviors such as substance abuse or the prevention of sexual diseases 9 , prevention of mental illness 10 or the prevention of violent behavior 11 .
Positive youth development (PYD) is a perspective born from positive psychology 12 and whose fundamental premise is that all children and young people are potential positive contributors to society and its immediate contexts. Sport is one of the ideal contexts for positive development and LS in young people (e.g., leadership, self-control) due to the high participation and nature of its context, which favors the involvement of subjects at an interactive, emotional and social level 13 and when an adequate structure is available 14,15 . Young people feel safe physically and psychologically in sports contexts, which favors the acquisition of LS 16 . For this reason, in the last 15 years, the number of studies investigating the potential of sport in this area has significantly increased [17][18][19] and the PYD has become one of the most studied and developed topics and fields in the area 20 .
As a result, PYD has also become a key element in the Physical Education (PE) curriculum in many countries 21 , emerging a multitude of different programs and models that focus on LS and have proven to be effective 22 , both specific to sports contexts (e.g., Girls on the Run program 23 ) as non-specific but transferred and applied to it (e.g., The Five Cs Model 24 ). The development of LS through sport has been the main topic of a wide variety of studies, which, in turn, involve a great diversity of variables. In some of them, the importance of the climate or context for their learning has been addressed, to try to understand what mechanisms and keys facilitate the development of LS in young people.
Self Determination Theory 25 is one of the promising theories for understanding such mechanisms 26 . The development of LS has been related, among others, to the satisfaction of basic psychological needs 17 , intrinsic motivation and friendship goals 27 , Personal and Social Responsibility 28 , Well-being 29,30 or Prosocial behaviour 31 . Recently, some authors have equally focused on how those skills can be transferred from PE to other everyday contexts, because otherwise they could not be considered LS 32,33 . Learning these skills and transferring them to everyday reality is critical for all young people 29 but especially for those at risk or vulnerable 34 . For Pierce et al. 32 the transfer of LS in sport is the process by which the learning and improvement of a LS is related to this subsequent application in, at least, one or more context of their individual's daily life beyond sports or physical activity.
However, despite the fact that, in recent decades, the development of LS has acquired great prominence in research and a multitude of programs and models have been created and applied, only one instrument has been found for assessing LS in physical activity and sport context. Cronin and Allen develop this scale for youth participants in PYD and LS through sports programs, emphasizing ages between 11 to 21 26 . This scale was developed to measure how the youth perceive their own LS development in 8 dimensions (teamwork, goal setting, time management, emotional skills, interpersonal communication, social skills, leadership, management, and problem solving) through sport in different practice contexts. Regardless of this, there is still a lack of measurement and, above all, a lack of instruments to assess LS for non-English speaking researchers and practitioners 35 . This scale has been only translated and validated for Portuguese population by Nascimiento-Junior et al 18 . The authors made a transcultural validation of the instrument in 413 youth athletes from Brazil with ages between 10 and 21. That said, there is still an important gap assessing LS in Spanish speakers. Responding to this need, the objective of this research is to translate and validate into Spanish the Life Skills Scale for Sport designed and validated by Cronin and Allen 26 .

Life Skills Scale for Sport
We used the English version of the Life Skills Scale for Sports (LSSS) 26 . To translate it to Spanish, we followed a double procedure. First, we asked bilingual and topic-related researchers to translate it from English to Spanish. Second, we asked another group of researchers to make translation from Spanish to English. In total, four bilingual researchers with Spanish as first language (2 Chilean and 2 Spanish) and three bilingual researchers with English as first language worked on the translation. Comparing both results we concluded that the terms translated coincided, and no relevant mismatch was found. The researchers put emphasis on cultural parlance, considering a neutral language for all Spanish speakers. Sequentially, the scale was applied to 124 people to assess the comprehension of the items, without receiving observations about it.
Original LSSS uses 43 items. The instruments items are evaluated through on a 5-point Likert scale, ranging from 1 (not at all) to 5 (very much), over sentences like "sports have taught me that…". The whole scale is divided into eight subdimensions: teamwork (seven items), goal-seeking (seven items), time management (four items), emotional skills (four items), interpersonal communication (four items), social skills (five items), leadership (eight items), and problem-solving and decision-making (four items). The instrument was typed in the "Google Forms" platform. Also, informed consent for ethical compliance and information protection with a presentation letter explaining the objective of the study and emphasizing that the personal information was anonymous and confidential, were uploaded. Ethical approval was obtained from the Ethics Committee of the Community of Aragon (CEICA).

Participants
For this study, a convenience sampling of 1251 students (52% Chilean, 47% Spanish and 1% other nationalities) were surveyed during sports class or training (during school physical education, college training, community practice or sports club) living and studying in Chile and Spain. From the sample, 59.8% were women (n=748) and 40.2% were men (n=503). The average age is 19.7, ranging from 15 to 29. In addition, they were asked to state the sport they usually practice, and more than 20 different ones were answered including individual disciplines (such as weightlifting, climbing, tennis, boxing, etc.) and team sports (futsal, football, handball, basketball, volleyball, among others). The authors have defined the sample in this age range considering definitions of youth of the United Nations (15 to 24) 36 and the National Youth Institute of Chile (15 to 29) 37,38 .

Psychometric properties
To validate the subdimensions proposed for the LSSS we used a confirmatory factor analysis over two different fitting models: bifactorial and second-order, following recent and similar research 18,39 . For the estimation procedure we used the R package lavaan (0.6-8), proposed by Roseel 40 which allows for maximum likelihood robust estimations, that follows literature recommendations for when data does not follow a normal distribution 18,41,42 . We preliminary checked for normality, and our data showed that multivariate normal distribution did not fit adequately to our sample (kurtosis test: 2908, p-value < 0.001; skewness test: 252, p-value < 0.001; normality test: 4.44, p-value < 0.001).
Two confirmatory factor analysis models were proposed according to Nascimiento-Junior et al 18 which are similar to those proposed by Wang et al. 39 in educational research. These models are i) an eight factor model which each one represents a sports-related skill by the group of items presented above, and the collection of all these abilities represents the whole LSSS (second order model), and ii) a bifactorial model in which the eight sports-related skills are represented by the sample items presented above and the collection of every item (43) in the scale also represents the whole LSSS (bifactorial).
Other studies also propose the study of a bifactorial and second-order models, such as Cronin & Allen 26 and Myers et al. 38 . Models that contain over one order of effect prevent loadings between factors to be excessively high, because the higher-order dimension allow each loading to be interpreted as a loworder to higher-order loading. Instead, in bifactorial models, loading between factors are forced to zero and the items are allowed to show higher loadings between factors (for a deeper discussion on psychometric differences between these models in sports psychology, see Ntoumanis et al. 39 ).
To evaluate the fit of these two models to the factors behind our data, we used common psychometric indicators, following Cronin & Allen 26 . Table 1 shows the benchmark indicators we use as a reference, where the second column gives a brief description, and the third column presents the literature suggested range for evaluation.

Initial Exploratory Analysis
An exploratory factor analysis was first run to check whether the eight expected factor were seen in the data. Contrasting information between the Cattell scree plot criterion and Horn parallel analysis, they both suggest that eight factors is the best possible fit over randomness in a normal distribution (the difference between factor and parallel analysis is lowered in the 8th factor). However, Kaiser method and Velicer minimum average partial method suggest lower number of factors: four and three respectively. Using eight factors explain over 99% of variance, and the Kaiser-Meyer-Olkin indicator was 0.988. Using regular correlations and polychoric correlations the conclusions are the same (because debate is held about using regular correlations in Likert scales in low-level categories and no normality, we compared both methods, reaching similar conclusions). The discrepancy between methods allows us to expect that using eight factors is adequate, though some of them will present lower loadings in general. Estimating the exploratory factor analysis we conclude that the first five factors show higher loadings than the last three, which is expected when the exploratory factor analysis shows this kind of results [44][45][46] .

Results
Results show that both models proposed fit adequately to the data presented. Nevertheless, the bifactorial model shows a better fit than the second-order one, and it relates better to the data according to the following indicators: root mean square error of approximation (RMSEA), the comparative fit index (CFI), and the Tucker-Lewis index (TLI) (see Table 2). Psychometric properties of the scale aim to an adequate fit to the data. Internal consistency is high (Cronbach Alpha of 0.980) and item-total correlation (0.630). The exploratory analysis (see Table 3) converges to the use of eight factors, showing that some of them would load higher than others, which is also similar to the findings in previous uses of this scale 18,26 .

Discussion
The aim of this study was to translate and validate into Spanish the LSSS designed and validated by Cronin and Allen 26 . Firstly, the translated terms were consistent across evaluators and no relevant discrepancies were found. Secondly, we used a sample of 1251 Spanish-speaking sports practitioners between ages 15 and 29. Results show that both models proposed fit adequately to the data presented.
Nevertheless, the bifactorial model shows a better fit than the second-order one, and it relates better to the data according to RMSEA, CFI and TLI indicators. These results give evidence that LSSS should be considered as a whole dimension, identified by all 43 items, instead of being only a collection of eight subdimensions. In other words, is better to consider LSSS as a general factor rather than it being influenced indirectly by the eight subdimensions. In future research, LSSS are to be considered as a factor itself.
Regarding the psychometric properties of the scale, this has an adequate fit and a high internal consistency, showing similar numbers as English 26 and Portuguese 43 version of the scale. Bifactorial model shows that the fit of the general factor is high and stable among all items. However, factor loading of specific item tends to be lower than the general model. For example, items three and six of teamwork show low loadings (which is also similar to the findings of Nascimiento-Junior et al. 18 . These authors show similar low and non-significant loadings for half of the leadership dimension (items five, six, seven and eight), while our findings shows that all its items' loadings are low and non-significant. Cronin & Allen 26 shows that leadership dimension general loading is the second lowest after emotional skills dimension. The bifactorial model shows higher loadings for the interpersonal communication dimension, whereas using the second-order model shows low and non-significant items and general loading. Nevertheless, the general fit of the bifactorial model suggests that the whole scale should be used and no subdimension should be excluded. Likewise, the use of the bifactorial model is more robust in "communication" subdimension inclusion, when using the bifactorial model, the overall load allows the subdimension to be included in the measurement; on the other hand, in the second-order model, neither the specific loads of the item nor the effect of the subdimension allow a good fit.
According to the evaluation of psychometric properties of the LSSS applied in Spanish speakers' population, the LSSS reproduces results collected in the publication of the original scale and maintains all the dimensions and the different elements-items that compose them. As several authors have said [47][48][49] , it is essential to evaluate interventions and programs that have an impact on young people lives.
Having this instrument would give the opportunity to sports educators, coaches, and teachers in the area of physical education to measure the levels of these skills in their athletes and/or sports students. Also, it can be a complement for other scales that allow us to understand the sport atmosphere in a globality way such as Teaching Personal and Social Responsibility scale 47 , Self Determination Scale 50 , and the transfer of the LS in coaches, such as Coaching Life Skills in Sports Questionnaire for Coaches 49 or Life Skills Self-Assessment Tool for Coaches 51 . In addition, it provides an evaluation tool for researchers in the field, which will allow a) to improve the research and evaluation of the effectiveness of programs or interventions that aim at the development of LS in sport, and b) explore new theoretical frameworks that involve such skills.
Unlike the Portuguese instrument and the original instrument, the scale was applied to a larger sample, including more than 20 sports, such as football, canoeing, chess, climbing, bicycle, futsal, badminton, weightlifting, volleyball, boxing, kickboxing, handball, paddle tennis, among other disciplines. This diversity gives a window to expand the field of application of the instrument, bringing more opportunities for assessing LS.

Limitations and Strengths
As main limitations of this study, we found the type of sampling (non-probabilistic by convenience) and the lack of balance in terms of sex (59% are women). The applicability efficacy of this instrument in other age ranges and various socio-sports contexts should be evaluated in the future. However, once the psychometric properties of this scale translated into Spanish have been analyzed and according to the results, despite the limitations indicated above, we can say that the strength of this study is that the LSSS it is a useful scale to evaluate the levels of LS in contexts of physical activity and sport in the Spanish-speaking population aged between 15 and 29 years of age.

Conclusions
The LSSS, whose translation and validation into Spanish is addressed in this study, has shown to have adequate psychometric properties and is a valid instrument for the evaluation of LS in contexts of physical activity and sport with a young population of 15 to 29 years. This instrument facilitates Spanishspeaking experts and researchers belonging to the field of physical activity and sport to generate knowledge and scientific production.