TYPE Original Research PUBLISHED 20 January 2023 DOI 10.3389/fpsyg.2023.1066494 Measuring optimal psychological OPEN ACCESS states: Proposal of two brief EDITED BY Duarte Araújo, University of Lisbon, Portugal versions to measure flow and REVIEWED BY Nathalie André, clutch in athletes University of Poitiers, France Henrique Brito, Faculdade de Motricidade Humana, Portugal Alejandro Sánchez Vara1,2*, José L. Chamorro2,3 , *CORRESPONDENCE Ruben Moreno Castellanos2 and Saul Alcaraz2,4 Alejandro Sánchez Vara 1 alejandro.sanchez@universidadeuropea.es School for Doctoral Studies and Research, Universidad Europea de Madrid, Madrid, Spain, 2 Faculty of Sports Sciences, Universidad Europea de Madrid, Madrid, Spain, 3 Hum-878 Research Team, Department of SPECIALTY SECTION Psychology, Health Research Centre, University of Almería, Almería, Spain, 4 Sport Research Institute UAB, This article was submitted to Autonomous University of Barcelona, Barcelona, Spain Movement Science and Sport Psychology, a section of the journal Frontiers in Psychology RECEIVED 10October 2022 ACCEPTED 04 January 2023 Background: The study of sports experiences with high levels of performance has led PUBLISHED 20 January 2023 to the observation of two psychological states categorized as optimal, denominated CITATION flow and clutch. The objective of this study is to design and validate two brief scales Sánchez Vara A, Chamorro JL, Moreno Castellanos R and Alcaraz S (2023) Measuring version to measure flow and clutch. optimal psychological states: Proposal of two brief versions to measure flow and clutch Methods: Following the MIMO (Maximum Information, Minimum Discomfort) in athletes. protocol, three studies are carried out: In Study 1, we developed the short versions Front. Psychol. 14:1066494. doi: 10.3389/fpsyg.2023.1066494 of the questionnaires based on theory-driven and data-driven criteria; in Study 2, we COPYRIGHT used quantitative criteria to validate the short versions; and in Study 3, we checked © 2023 Sánchez Vara, Chamorro, Moreno the final version of each scale to observe their statistical validity and their relation to Castellanos and Alcaraz. This is an open-access article distributed under the terms of the other variables. Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other Results: The new short flow scale is made up of seven items, while the clutch scale forums is permitted, provided the original is made up of five items. Throughout the three studies, evidence is provided of the author(s) and the copyright owner(s) are internal consistency, invariance of the measurement models and content validity, credited and that the original publication in this journal is cited, in accordance with accepted validity related to the responding process and validity in relation to other variables. academic practice. No use, distribution or reproduction is permitted which does not Conclusion: This study provides two short versions to evaluate flow and clutch, comply with these terms. which will help to continue advancing in the study of optimal psychological states in sport. KEYWORDS flow, clutch, estados óptimos psicológicos, versión corta, validación de cuestionario flow, optimal psychological states, short versions, scale validation Introduction Flow is considered a complex psychological state crucial for the life development of human beings (Csikszentmihalyi, 1992). Although the early research about this phenomenon was conducted in people’s everyday life (Csikszentmihalyi, 1975a,b, 1981), it soon started to be investigated in the sport context and was defined as the optimal state for the sports practice (Chalip et al., 1984; Jackson, 1992; Jackson and Roberts, 1992; Kimiecik and Stein, 1992). Because of the relevance of this state and looking for a deeper understanding of it, Swann et al. (2016a) ran an observation on professional golfers during tournaments. In this research, they found in athletes’ explanations two different psychological states during their best performances. The authors described these two states with the expressions “letting it happen” and “making it happen” (Swann et al., 2016b). They described “letting it happen” as an experience characterized Frontiers in Psychology 01 frontiersin.org Sánchez Vara et al. 10.3389/fpsyg.2023.1066494 by open goals, less contextual structure and related to exploratory and Vlachopoulos et al., 2000). Swann et al. (2012) showed it in their self-referenced behaviors. They considered this state the equivalent systematic review, where only 28.95% of the total athletes in of flow. The other state, “making it happen”, introduces new flow research speak about that specific dimension. These authors characteristics from the nature of the objectives and the context, with explained that variations in the context can modify the subjective fixed goals, a more defined structure and the recognition of non- perception of time (for example, the type of sport). For all that, as we subjective goals related to the context. The researchers considered this said before, the number of variations in the intensity of the subjective second state to be something similar to those performances called experience of the athletes, and the issues on the causality explanation “clutch performance”, described by Hibbs (2010), where the athlete of the state, there is a need for a better explanation of this new state can achieve positive results using their own abilities in contexts with denominated clutch. high pressure. For flow research, different tools were developed during past Thus, currently, we have two optimal psychological states whose years. But in the validation process, some instabilities in the definition contains particular characteristics. On one hand, Flow is conceptualization of flow were observed (see Table 1). After the first the experience of a subjective state that merges within the sport qualitative studies to identify flow in sport, (Jackson, 1992, 1995; context that challenge the abilities of an individual. The achievement Jackson and Roberts, 1992), Jackson and Marsh (1996) developed of goals in this state is produced from an exploratory attitude of the Flow State Scale (FSS) with 36 items (four items/dimension). the individual’s possibilities of action, connecting with their context That first scale was focused on specific events, such as training or and flowing with the consecutive application of skills. The original specific competition games. Some variations were introduced later definition of flow describes it as a multidimensional state with nine with the objective of studying the dispositional flow, which means dimensions: (a) perceived challenges that stretch existing skills, (b) the frequency that an athlete can experience flow (Dispositional action-awareness union, (c) clear goals, (d) positive feedback, (e) Flow Scale, DFS; Jackson et al., 1998). After some problems in data concentration in the task at hand, (f) sense of control in what one is reliability and the discovery of some conceptual areas that could be doing, (g) loss of self-consciousness, (h) distortion of time perception, improved (Vlachopoulos et al., 2000), Jackson and Eklund (2002) and (i) autotelic experience (Csikszentmihalyi, 1975a, 1990; Jackson reviewed the FSS and introduced changes in some items, calling the and Marsh, 1996). On the other hand, Clutch is the experience of new tool FSS-2. With the objective of being used and modified in a subjective state derived from the perception of an increase in sporting contexts and other areas of psychology in different cultures, contextual pressure on the individual due to time or results demands the FSS was translated into Spanish for García Calvo et al. (2008). derived from sports practice. The individual will be able to recognize Some researchers considered it interesting to work in short forms of and manage the stress originated, so that he will raise his physical and the FSS, too. So, Rheinberg et al. (2003), proposed a 10-item short- mental energy levels to achieve the proposed objectives. Clutch has form (Flow-Kurzskala – FKS). Despite some reliability problems in been defined as a psychological state characterized by (a) complete their statistical analysis the authors asserted that all components and deliberate attention, (b) intense physical effort, (c) high ego of the flow experience can be measured with 10 items. In their attention, (d) strong physiological activation, (e) absence of negative study, they found two factors (FI & FII) within the scale. They thoughts, and (f) automaticity in the use of one’s own abilities (Swann showed that high values in their scale of total flow indicate that et al., 2016a, 2017a,b, 2019). both factors (FI & FII) were presented, while, with medium values, The study of the optimal psychological states is relevant for accentuations of one factor occurred to the detriment of the other. the relation of their characteristics and the best experiences in This means that the perception of one factor did not permit the human beings. While clutch is still under study, flow state has been experience of the other factor at the same time. Jackson et al. (2008) understood as the intersection between the greatest experiences and developed a short version with one item for dimension, respecting peak performance (Jackson and Roberts, 1992; Kimiecik and Stein, the original structure definition of the concept and promoting the 1992; Jackson and Marsh, 1996; Engeser and Rheinberg, 2008; García use of short forms. They defended that the short versions proved to Calvo et al., 2008; Bakker et al., 2011; Swann et al., 2012; Verner-Filion be just as reliable as the long versions, maintaining the validity of and Vallerand, 2018; Tse et al., 2019). In any case, flow experience the construct and defending its multidimensional explanation with has been considered rare and elusive (Csikszentmihalyi, 1975b, 1986; 1 item for each dimension. Despite some issues in their results, Kimiecik and Stein, 1992; Swann et al., 2017a). The researchers they conclude that “psychometric support” is achieved together with have not been able to explain, for example, the different variations “internal consistency and external validity” (p. 29) in relation to the in the intensity perception during the experience of this state key elements of the construct (Jackson et al., 2008). The authors (Csikszentmihalyi, 1990, 1999; de Nakamura and Csikszentmihalyi, recommended continuing investigating new short forms. 2014), justifying it based on different factors, depending on the In recent years, the methods used to study clutch have been authors or the methodologies (Rheinberg et al., 2003; Swann et al., qualitative methodologies, experimental approaches measuring 2012; Fernández et al., 2015; Stander et al., 2015; Chang, 2016; subjective experiences, introduction of modifications in athletes’ Sinnett et al., 2020). Although the first conceptualization of flow skills or analysis of the competitive results during clutch has been generally accepted and the use of the scales respecting performances. All these methodologies brought us to the situation the original nine dimensions generalized, some theoretical and of defining clutch as a result of a performance, as an ability, as methodological problems have been observed, and some authors a specific situation or as a psychological state (Schweickle et al., have found anomalies in the flow research (Rheinberg et al., 2003; 2020). In that variety of definitions, clutch has been measured Swann et al., 2018). Some examples of these anomalies are the as a skill and/or a performance episode, looking to prove its objections to the dimensions of loss of self-consciousness and time occurrence and evidence, and the obtained results have been distortion dimensions (Jackson and Marsh, 1996; García Calvo et al., very unsatisfactory. Until today, definitions of clutch still lack 2008), or specific problems in the time distortion dimension in specificity and clarity, creating misunderstandings about what other authors (Jackson and Marsh, 1996; Marsh and Jackson, 1999; this concept is (Schweickle et al., 2020). For all this, the authors Frontiers in Psychology 02 frontiersin.org Sánchez Vara et al. 10.3389/fpsyg.2023.1066494 TABLE 1 Flow questionnaires, characteristics, and limitations, over the years. Name Reference Items Language Characteristics Limitations Flow State Scale – Jackson and Marsh, 1996 36 English 9 factor model, with 4 items Lower factor loadings in Transformation of FSS for each dimension. Likert Time and Loss of Self-Consciousness. No scale of 5 points, from 1, clarification if one global factor is valid or strongly disagree to 5, must be used always using 9 factors strongly agree. individually. Retrospective approach to data collection. Heterogeneous sample may not generalize to specific subgroups. Dispositional Flow Jackson et al., 1998 36 English Based on FSS, respect 9 Sample was limited to older athletes, not Scale - DSS factors and 36 items (4 each having control over when participants dimension). Changes in the completed the state assessment. No answer to wording and the tense of the global factor to measure flow. Lack of phrases designed to assess relationship with the motivation scales and frequency. variables. Lack of relationship between challenges ratings and flow subscales. Flow State Scale-2 – Jackson and Eklund, 2002 36 English Original 36-item FFS along The higher-order factor loadings of time FSS-2 with 13 additional items transformation remain relatively weak, being devised as potential unclear whether the item is problematic. replacements. 5-point Likert Differences in factor loadings expresses scale ranging from 1, never to differences in flow experiences, but there is no 5, always. specific answer on why this can be supported (maybe differences in sports or physical activity). Special problems in the perception of the self. Limited control over when the questionnaires are completed. Flow State Scale García Calvo et al., 2008 36 Spanish Translation of the original Lowest value in sense of time and loss of (Spanish Version) FSS to Spanish. 36 items and self-consciousness, where the authors 9 factors. 10-point propose to revise them in future research. Likert-type scale. Different correlations to motivational variables compare to DFS. Heterogeneous sample produce unexpected relations, with heterogeneous values in its results. Flow-Kurzskala – Rheinberg et al., 2003 10 German 10 items adapted from the Include new items in the scale that were not FKS original FSS where selected. related to flow dimensions or construct, that 7-point Likert-type scale. have weak psychometric stability. Not achieve desired fit limits and had cross-loadings and unsatisfactory factors loadings. Short Flow Scale - Jackson et al., 2008 9 English 9 items, with one item from Did not reach acceptable criterion levels of fit SFS the four-item measures of across all indices. The situation-specific state each on the nine flow short measure requires more discrimination dimensions. between factors (fit is reduced). Limitations of the self-report and retrospective recall. Problems in validity and reliability and additional research is needed to examine its psychometric properties. of the clutch definition that we follow invite the researchers to In addition, short questionnaires improve data quality, as long develop new questionnaires that can distinguish flow and clutch tools have been found to lead to boredom and apathy (Coste et al., states, since the methodologies and tools used until today may not 1997; ; Jackson et al., 2008; Alcaraz et al., 2013, 2020; García-Alba be differentiating characteristics from both experiences, being all et al., 2021). In addition, it will be following the recommendations aspects assigned to flow. This is happening due to the little validity of authors who defend that shortening existing tools is better than for the discrimination of states that have the current tools (Swann creating new ones since it saves phases in their creation and the et al., 2017b, 2019). instrument will be familiar for the participants (Coste et al., 1997; The objective of this study is to design and validate two Alcaraz et al., 2013, 2020). For that, the Spanish translation of the FSS independent short instruments to measure the psychological states by García Calvo et al. (2008) will be used to develop a short form for of flow and clutch in Spanish. One short form for flow and a flow. And for clutch, a new scale is going to be developed since there is new scale, based on the Integrated Model of Flow and Clutch still no questionnaire to measure this phenomenon as a psychological States, for clutch. Our main focus during the development of these state in Spanish. scales was to develop instruments that assessed the core aspects of Our process of development and validation of the short forms each psychological state. In this line, we wanted to create efficient relies on the Maximum Information, Minimum Discomfort method single-factor measures that were able to capture the most important (MIMO, Alcaraz et al., 2020). This protocol is initially based on characteristics of each state. Boateng et al., 2018 recommendation to develop new questionnaires Frontiers in Psychology 03 frontiersin.org Sánchez Vara et al. 10.3389/fpsyg.2023.1066494 and is divided into four phases. In this study, we present three on a 10-point Likert-type scale, where 1 means totally disagree different studies that cover all the stages of MIMO development. and 10 totally agree. This translation of the Spanish FSS obtained Specifically, in Study 1 we selected the initial pool of items optimal reliability values, where Cronbach’s alpha reached values based on theory-driven, data-driven and comprehension-related above 0.70, and in several cases, above 0.80. In addition, for its criteria (Phase I), refined the short-form with the evaluation of a translation, a sample had been used in which there were many panel of experts (Phase II), and assessed the target population’s participants dedicated to the practice of team sports, especially understanding of the items (Phase III). Subsequently, in Studies football. 2 and 3 we targeted Phase IV. On the one hand, in Study 2 we assessed the dimensionality and reliability of the questionnaires. Clutch State Scale (CSS) On the other hand, in Study 3 we confirmed the factor structure A scale with six items was developed, one for each characteristic of both questionnaires and provided validity evidence based on that distinguishes the state of clutch with that of flow, based on the relation to other variables. Ethical approval for this investigation proposal in the work of Swann et al. (2016a). These six items created was obtained from the University Ethics Committee (reference: based on the theoretical development of the clutch characteristics CIPI/19/011). were presented to the focus group for discussion and verification of their understanding. At the same time, the 10-point likert-type scale is maintained, where 1 is totally disagree and 10 totally agree. Study 1: Pilot study. Short-form Procedure development The development of short forms following MIMO criteria explains that in Phase I, Initial Item Selection, the process For scale validation, the MIMO strategy (Maximum Information, must respect three conditions for the item selection: theoretical, Minimum Discomfort, Alcaraz et al., 2020) has been followed and psychometrics, and comprehension. Study 1 will complete Phase 1: Initial Item Selection, Phase 2: Expert Our theory-driven criteria came from a literature review and Group, and Phase 3: Comprehension of Focus Group. the advice of an expert committee. After analyzing the original version of the FSS, they were entrusted with a double task: order numerically, under their criteria, which items best represented the Method definition of each dimension for flow, and, at the same time, provide qualitative justification of the selected order. At this stage, three Participants new item proposals were included for those dimensions in which We obtained qualitative and quantitative data from participants it was considered that the existing items did not make “reference in Study 1 to develop the short forms. Qualitative data came to the core of the construct” (p. 55, Alcaraz et al., 2020). From from an expert committee with extensive experience in applied these three items, after semantic analysis and discussion, finally were research and sports psychology counseling, all of whom usually included two related to the dimensions of clear goals (item 3: It was work in different Universities in Spain and are part of different very clear what I needed to do, and I wanted to achieve) and sense research groups of psychology. The other source was a focus group of control (item 6: I was feeling a sense of control over my body formed by four athletes, three men and one woman, with ages and mind). The shortening process continued with the focus group, between 18 and 21 years old, all of them practitioners of team which was conducted to explore the opinion of our target population. sports (football, basketball, volleyball, and handball), with more than During these interviews, three items selected of the FSS were specially five years of experience in regular training and competition, and discussed: those related to the dimensions of clear goals, loss of self- University students. consciousness, and distortion of time. In the dimension of clear goals, Qualitative data came from a sample of 128 young football players the first proposal did not distinguish between self-perception and the belonging to a club in the Community of Madrid (Sample 1). The perception of the coach to define the quality of the performance. In sample was totally male and ranged in age from 14 to 23 years the loss of self-consciousness, there was a misunderstanding between (M = 16.28, SD = 2.60). All the players trained regularly (3 weekly the worried thoughts related to the self and coming from the peers. training sessions) for 1 hour and 30 minutes and competed in their And in the time distortion, it seems that athletes have two kind of respective categories (represented by a total of 7) within the Madrid time perceptions that represents different levels. One level, that we Royal Football Federation (RFFM) championships on weekends. The can call it micro, refers to the temporal perception related to the skills accumulated experience of the players varied from 2 to 18 (M = 8.48, execution. And the second level, more general, that can be called SD = 3.82) training seasons. macro, refers to the general perception of the time needed to realize the complete the task. These conceptual differences were discussed Instruments with the expert committee, and the changes considered relevant and For the flow study, we administered the translation of the original notorious were introduced. The answers were given on a 10-point version of the FSS (Jackson and Marsh, 1996). For the clutch study, we Likert-type scale. developed a new scale. To measure the clutch experience, following the theoretical principles exposed in Swann et al. (2016a), a new scale was developed. Flow State Scale (FSS) Being a first version of the scale in Spanish for the study of the To ensure that the sample responded to the concept of new theoretical construct developed in English, the characteristics flow, the Spanish version of the FSS translated by García Calvo proposed by the original authors were translated into Spanish. et al. (2008) was presented. The original version of the FSS was The Spanish version of the clutch characteristics was analyzed and developed by Jackson and Marsh (1996) and is composed of 36 compared with the flow dimensions by the expert committee: changes total items, 4 for each of the 9 dimensions, which are answered in some lexical and meaning nuances were introduced with the Frontiers in Psychology 04 frontiersin.org Sánchez Vara et al. 10.3389/fpsyg.2023.1066494 objective to differentiate both states in the same questionnaire. correlation is observed between automaticity of skills and heightened At the same time, several versions of the scale introduction were arousal. developed, introducing the specific characteristics that distinguish flow from clutch and that are originated in variations in the sporting context. Thus, a 6-item scale was generated in which each item Discussion responded to each of the characteristics proposed by Swann et al. (2016a). The answers were given on a 10-point Likert-type scale, In Study 1, we developed the short forms by using theory-driven thereby maintaining the structure of the FSS. To help participants as well as data-driven and comprehension criteria. Initial short forms identify clutch as a state other than flow, the scale was presented for flow and clutch were created (Version 0). A future study was with a different introduction to flow based on its phenomenological needed to validate these new short forms. characteristics. For clutch, the emphasis was placed on the contextual variation that affects the perception of the experience by the individual, as this state was related to a sudden assessment of Study 2: Preliminary psychometric situational demands due to a change in them. For this reason, in the scale presentation, the participant is reminded that “during a validation of the short forms training session or match there are variations in the context due to Following with the MIMO structure (Alcaraz et al., 2020), external causes”, exposing several possibilities below (result, orders in Study 2, we developed the first part of Phase IV, Evaluation from the coach or pressure from the public) and asking if, based on of the Short Version. With the application of the two scales, its these situational changes, “you have changed the way you face your statistical analysis is developed, in which its reliability and validity participation”. are assessed together with the relationship between dimensions and The psychometric criteria were obtained from the data of the characteristics. original questionnaire, together with the data of its application in Sample 1. In this phase, the instruments were administered using paper and a pen at the club’s facilities, where they had a meeting room used for the application of the questionnaire. All minors had Participants previously been given the informed consent document, which was Sample 2 consisted of 286 young soccer players, aged between returned with the signature of their parents or guardians before 14 and 18 years, (M = 16.56, SD = 0.94), and of whom 217 (75.9%) completing the questionnaire. The objectives and characteristics of were male and 69 (24.1%) females. These players were distributed the study were reported, as well as the confidentiality and anonymity in a total of 16 different teams within 12 different clubs located in of the responses, remembering that their participation was voluntary. different cities of Spain and covering up to nine different categories, The questionnaires were always kept and guarded by the principal between men’s and women’s football. The experience of the group of investigator of the study. Once data were collected, data-driven players of both sexes ranged between 1 and 13 seasons (M = 7.68, criteria were obtained from results of internal reliability and internal SD = 3.28) of training. structure. More information about these analyses can be found in In addition, the aim was to introduce gender differences in the results section. The data-driven criteria helped us to determine the sample, since differences between boys and girls have been observed items that conform the short-form questionnaire and observe the in other studies on optimal psychological states (Elbe et al., 2010; validity of the clutch scale. Chang, 2016). In the clutch scale, no conceptual issues were found, and focus group found it clear and comprehensible. Instruments Flow state short-scale (FSSS) The short version obtained from the quantitative and qualitative Results analysis of Study 1 was used (Version 0). This process led us to develop a short flow scale of nine items, one for each As a result of the procedure conducted with the expert dimension that is part of the state. The scale was answered in committee, the focus groups with the target populations, and the same 10 points Likert scale, as in Study 1, respecting the the research meetings, we obtained a great number of criteria original construction. to select the items that would be included in flow and clutch questionnaires. Such criteria were complemented with the data Clutch state scale (CSS) collected with Sample 1. First, we obtained the descriptive statistics After the analysis of the data in Study 1 (Version 0), this scale for each item and dimension and then, we assessed the correlation included six items, each of them related to a dimension of clutch. between dimensions. As can be observed in Table 2 for flow and The stem of the questionnaire was “In some specific moment of Table 3 for clutch, almost all correlations between dimensions and training/match.”. The scale was answered in the same 10 points Likert characteristics are statistically significant. Only in the dimension scale, as in Study 1. distortion of time, some relations were smaller and not significant, although they were being positive as expected. Specifically, such smaller relationships are only observed with three dimensions: Procedure positive feedback, concentration in the task at hand, and loss of self-consciousness. Related to clutch items, all correlations are The different sports facilities of the participating clubs were positive, as expected, and significant, being ≥ 0.377. The lowest visited, and the questionnaires were administered using paper and Frontiers in Psychology 05 frontiersin.org Sánchez Vara et al. 10.3389/fpsyg.2023.1066494 TABLE 2 Correlations between items for FSS shortening. Balance Union Clear goals Feedback Concentration Control Loss of self Time Autotelic Balance Union 0.197** Clear goals 0.401** 0.291** Feedback 0.489** 0.298** 0.335** Concentration 0.250** 0.256** 0.493** 0.462** Control 0.520** 0.415** 0.546** 0.460** 0.371** Loss of self 0.298** 0.179** 0.227** 0.361** 0.541** 0.345** Time 0.297** 0.217** 0.282** 0.174 0.123 0.272** 0.158 Autotelic 0.263** 0.314** 0.303** 0.398** 0.361** 0.289** 0.203* 0.263** The nine dimensions of the original FSS scale: (a) Challenge-skill balance, (b) Action-awareness merging, (c) Clear goals, (d) Unambiguous feedback, (e) Concentration on the task at hand, (f) Sense of control, (g) Loss of self-consciousness, (h) Transformation of time, (i) Autotelic experience. *p < 0.05; **p < 0.01. TABLE 3 Item correlations between items for clutch. Complete Intense effort Heightened Heightened No negative Automaticity focus awareness arousal thoughts of skills Complete focus Intense effort 0.643** Heightened awareness 0.595** 0.582** Heightened arousal 0.523** 0.527** 0.534** No negative thoughts 0.676** 0.564** 0.559* 0.491** Automaticity of skills 0.394** 0.431** 0.389** 0.377* 0.417** *p < 0.05; **p < 0.01. a pen at the same venues. The main researcher was always with to refine the Version 1 of both questionnaires. We started from an the different teams solving all doubts that could arise when reading initial 9-item measurement model for the FSSS and from a 6-item and completing the scales. The questionnaires never took more measurement model for the CSS, and then we deleted the items that than 20 minutes to answer, and the technical staff of the teams failed to produce sufficient measurement requirements. As the main helped in the organization. All minors had previously been given the idea was to assess the core aspects of flow and clutch, all measurement informed consent document, which was returned with the signature models were expected to fit a single-factor structure. We considered of their parents or guardians before completing the questionnaire. items for deletion when they included target standardized factor The objectives and characteristics of the study were reported, as well loadings of < 0.40. We also inspected modification indices to identify as the confidentiality and anonymity of the responses, remembering problematic items. Finally, we considered theory-driven criteria that their participation was voluntary. when necessary. All the items shared similar scrutiny and were deleted after we reached consensus. This process resulted in the Version 2 of FSSS and CSS. Data analysis All CFA models were estimated based on the Maximum Likelihood (ML) estimator. Different fit indices were used to test Data were analyzed using SPSS version 17 and Mplus version the fit of the CFA measurement models to the data: χ2 statistic, 7.4 (Muthén and Muthén, 2015). First, preliminary analyses included Comparative Fit Index (CFI; Bentler, 1990), Tucker–Lewis Index the detection of missing values and the analysis of data distribution. (TLI; Tucker and Lewis, 1973), and root mean square error of Subsequently, we tested and refined the measurement models of the approximation (RMSEA; Steiger and Lind, 1980) including its 90% Flow State Short-Scale (FSSS) and Clutch State Scale (CSS). As our confidence intervals (CI). A non-significant χ2 value indicates a focus was to develop efficient questionnaires that assessed the core close fit between the observed and the expected model values. The aspects of each psychological state, both instruments were expected threshold of acceptable fit for the RMSEA is ≤ 0.08 (for an excellent to fit single-factor structures. However, to obtain evidence supporting fit, £0.06; Browne and Cudeck, 1993). Additionally, CFI and TLI such assumption, we first conducted exploratory factor analyses values > 0.95 are considered as indicators of excellent fit (Hu and (EFA) to assess whether a single-factor structure was more suitable Bentler, 1999). than a multi-factor structure. Varimax was selected as rotation To assess the reliability of the questionnaires, we used method. Such analyses were conducted using IBM SPSS statistics 23. McDonald’s (1970) coefficient omega (ω), which is computed Then, the single-factor structure of both instruments was from the standardized parameter estimates of the model. Finally, examined with Mplus using a Confirmatory Factor Analysis (CFA) concurrent validity of both questionnaires was analyzed via a approach. To that end, we tested a series of single-scale CFA models deattenuated correlations between their latent factors. To do so, the Frontiers in Psychology 06 frontiersin.org Sánchez Vara et al. 10.3389/fpsyg.2023.1066494 two latent factors were included in a single measurement model to in Supplementary Table 1 (Supplementary material). Regarding the compute these correlations. reliability of this scale, McDonald’s (1970) coefficient omega was excellent1 :  = 0.789. From this point forward, we refer to the 7-item, single-factor FSSS as FSSS –Version 2. Results Regarding the CSS, the analysis began with a 6-item single- factor model (see Table 5). This measurement model provided an acceptable fit to the data: χ2 (df ) = 30.649 (9), p < 0.001, RMSEA Percentages of missing data were < 1.4% for all the items. [90% CI] = 0.092 [0.057 –0.129], CFI = 0.954, TLI = 0.924. However, According to Graham’s (2009) criterion, our missing data had no Item 6 (I wasn’t thinking in the execution, I was focus on the general psychometric effects on data analyses. Subsequently, we conducted context) presented a factor loading below 0.40 and were considered an EFA with each questionnaire to test whether a single-factor for deletion (l = 0.290). Considering the theory-driven criteria, there structure was more convenient than a multi-factor structure. In the was a problem with item 2 of the Flow scale. This item in the clutch case of FSSS, the results of the EFA returned two factors with an scale refers to the dimension called Automaticity of skills, during the eigenvalue > 1. However, the rotated solution showed one factor with subjective experience. This sense of automaticity is similar to that 7 items and a second factor with only two items (items 2 and 8; further one that we are looking for in dimension 2 of flow, where the item explanation on these two items is provided in the next paragraph). states, “things were happening automatically”, whereas in the clutch The percentages of explained variance were as follows: Component item, this sense is described by “I wasn’t thinking in the execution”. 1 = 36.8%, Component 2 = 13.2%. Regarding the CSS, the EFA results Previous researchers have related those problems with the different also supported the preference for a single-factor structure. In fact, experiences of the 9 dimensions to different causes across settings and only one factor presented an eigenvalue > 1 (% of variance = 49.7%). depending on the type and kind of sports or individual differences Further description of these results is available from the first author (Swann et al., 2012). We understand that the specific cultural context upon request. related to groups can influence the interpretation of this subjective Then, a series of single-factor CFA models were tested to refine experience. the measurement model of both questionnaires. For the FSSS, the The measurement model without Item 6 showed a good fit to a priori 9-item model showed a fit to the data of χ2 (df ) = 73.247 the data: χ2 (df ) = 9.859 (4), p = 0.043, RMSEA [90% CI] = 0.072 (27), p < 0.001, RMSEA [90% CI] = 0.077 [0.056–0.099], CFI = 0.912, [0.012 –0.130], CFI = 0.987, TLI = 0.967. Such model also benefited TLI = 0.882 (Table 4). Thus, deletion of some of the items appeared from the inclusion of a correlated uniqueness between Items 1 and 2. to be necessary. Specifically, the model significantly improved when Furthermore, all the items included in this model showed satisfactory Items 2 (Things just seemed to be happening automatically) and 8 (At factor loadings > 0.40 (range = 0.575–0.742; see Supplementary times, it almost seemed like things were happening in slow motion) Table 2 in Supplementary material), no modification indices were removed from the analysis. Those items were considered suggested further item deletion, and McDonald’s (1970) coefficient for deletion as they presented insufficient factor loadings in the omega was excellent1 ( = 0.813). Once we finished this process, we initial model (litem 2 = 0.193, litem 8 = 0.143). In addition, deletion labeled this 5-item, single-factor CSS as Version 2. of those items was also supported by the previous exploratory factor analysis. From a theoretical standpoint, problems with the items that refer to the dimension of Transformation of time were observed from the very beginning of flow research. In the same FSS Discussion development (Jackson and Marsh, 1996), the authors considered that In Study 2, we began the process of validating the short forms dimension “less universally important” (p.30), due to less support developed in Study 1. Statistical evidence concerning their reliability, from the athlete’s experiences and factor loadings. Swann et al. concurrent validity and improvement of data quality was obtained. (2012) confirmed that point in their systematic review about flow: Evidence supporting the single-factor structure of FSSS, and CSS Transformation of time was the last dimension in terms of athlete’s was obtained from both exploratory and confirmatory approaches. citation (only 28.95%, when the first, related to concentration, Our results showed that two items presented fit problems, which achieved 80.7%). In the same line, in the Spanish version of the coincides with previous research. The use of the different type of FSS (García Calvo et al., 2008), those two dimensions were the sports and cultural contexts have shown that these variations inside dimensions with the lowest values (Transformation of time and the sample can bring variation to the data results. This reinforces the Balance of action and awareness,0.42,0.59, respectively). In other need to continue studying both subjective states, looking for which research using Spanish scales with swimmers (López-Torres et al., dimensions can be described as universal, and which are more related 2007), the dimension Balance of action and awareness is the 6th , to the individual characteristics inside the samples. A future study was out of the 9 dimensions, in terms of the experience score when then necessary for the last version of the short forms. athletes describe flow. It seems that the significance of the sense of union between action and awareness could be different in the Spanish context. The item-deletion process resulted in a 7-item, single-factor Study 3: Psychometric validation and FSSS, which showed a good fit to the data: χ2 (df ) = 25.409 (14), criterion validity p = 0.031, RMSEA [90% CI] = 0.053 [0.013–0.086], CFI = 0.976, TLI = 0.964. In addition, all the items included in this model In Study 3, the last part of MIMO Phase IV, Evaluation of the showed satisfactory factor loadings > 0.40 (range = 0.487–0.697; Short Version, is applied. Study 3 has two main purposes: test the see Supplementary Table 1 in Supplementary material) and no psychometric properties of FSSS-Version 2 and CSS-Version 2 and modification indices suggested further item deletion. The data observe the relationship of flow and clutch with burnout and the distribution and factor loadings for the retained items are available subjective perception of performance. Frontiers in Psychology 07 frontiersin.org Sánchez Vara et al. 10.3389/fpsyg.2023.1066494 TABLE 4 Comparison among CFA measurement models (Flow state short-scale). Model χ2 (df) p RMSEA (CI 90%) CFI TLI Study 2 Original 9-item Flow State Short-Scale 73.247 (27) <0.001 0.077 (0.056 –0.099) 0.912 0.882 Deletion of item 2 and item 8, item 3 (Version 2) 25.409 (14) 0.031 0.053 (0.016 –0.086) 0.976 0.964 Study 3 Version 2 79.207 (14) <0.001 0.093 (0.074 –0.114) 0.932 0.898 TABLE 5 Comparison among CFA measurement models (Clutch state scale). Model χ2 (df) P RMSEA (CI 90%) CFI TLI Study 2 Original 6-item Clutch State Scale 30.649 (9) <0.001 0.092 (0.057 –0.129) 0.954 0.924 Deletion of item 6 (Version 2)1 9.859 (4) 0.043 0.072 (0.012 –0.030) 0.987 0.967 Study 3 Version 2 14.515 (4) 0.006 0.070 (0.034 –0.111) 0.985 0.961 1 This model includes a correlated uniqueness between Item 1 and Item 2. Participants Procedure Sample 3 comprised 551 athletes, 382 men (72.07%) and 148 In this phase, the response to the questionnaires was done women (27.92%), with an age range of 16 to 35 years (M = 18.61, digitally, through Google Forms. The data was stored in a SD = 3.83) for both males and females (Men: M = 18.53, SD = 3.87; computer file and subsequently analyzed with the SPSS application. Women: M = 18.82, SD = 4.09). In this case, the sample is formed Confidentiality was guaranteed throughout. by different team sports. Football is repeated, and we introduce basketball, handball, volleyball, roller hockey, rugby and indoor football, and paddle and tennis as individual sports. The practice Criterion validity experience in training of the sample comprised from 1 to 24 years (M = 8.37, SD = 4.31; Men: M = 9.09, SD = 4.31; Women: M = 6.58, Based on previous theoretical and empirical results, it is SD = 3.78). In addition, research on gender differences is maintained, hypothesized that higher levels of experience of the optimal since differences between men and women have been observed in psychological states will reduce the possibility of developing burnout other studies on optimal psychological states (Elbe et al., 2010; Chang, process in athletes, whereas greater experiences of the two optimal 2016). psychological states will produce an increase in the subjective perception of performance during sports practice. Finally, subjective Instruments perceptions of high performance will be negatively related to Short version of the Athlete Burnout Questionnaire (ABQ) advanced states of burnout. A short version of the original 15-item version of the ABQ developed by Arce et al. (2010) was used to measure burnout. De Francisco (2015) studied a 9-item version, in comparison with Data analysis versions of 12 and 15 items of the same scale, finding that the model with nine was the one that presented a higher value of Preliminary analyses included an examination of missing values, variance, in addition to a better fit in the confirmatory factor analysis distribution of the data, and scale reliability using McDonald’s (1970) and maintaining the levels of relationship between factors found in coefficient omega. Then, to obtain further evidence regarding the previous research (both in the original and in the Spanish version). internal structure of the questionnaire, we used Mplus software This 9-item version was used as scale. The response format was version 7.4 with ML estimator (Muthén and Muthén, 2015) to Likert type, with four alternatives: “Almost never” (1), “Rarely” (2), confirm the single-factor structures described in Study 2. “Sometimes” (3), “Often” (4). To examine whether the FSSS-Version 2 and CSS-Version 2 displayed invariance across different groups of athletes (i.e., sex and Subjective perception of performance age), we tested a series of models using Mplus 7.4. Specifically, we The performance of the players was measured from two items tested three models for each subdivision: configural (i.e., expected adapted from the study of Bakker et al. (2011), who also wanted to same factorial structure across groups), metric (i.e., expected same observe the individual performance of football players. The response factor loadings across groups), and scalar (i.e., expected same item scale was adapted to a family scale within the culture of the study intercepts across groups) invariance. We compared those models by participants. From 1 to 10, in the first item, the participants had to examining Satorra–Bentler χ2 , CFI, TLI, and RMSEA differences. assess how their performance was being “given their abilities”. And The more parsimonious model was selected only when changes in in the second, using the same 10-point scale, they had to judge their CFI were < 0.01 and increases in RMSEA were < 0.015 (Cheung “performances in competitive matches.” and Rensvold, 2002; Chen, 2007). Changes in the TLI were evaluated Frontiers in Psychology 08 frontiersin.org Sánchez Vara et al. 10.3389/fpsyg.2023.1066494 following similar guidelines associated with changes in CFI (Marsh Results et al., 2009). Finally, to obtain validity evidence based on relation to other Preliminary analysis and measurement variables of the FSSS-Version 2 and CSS-Version 2 scores, we models tested two separate structural models. In those models, Flow and Clutch were expected to mediate the negative relationships between Tables 6, 7 present the distribution of the data for the FSSS- Burnout and Perceived performance. As a previous step before Version 2 and CSS-Version 2 items. As can be observed, percentages running the models, we first tested the measurement models of of missing values were £ 0.4% for all the items, not causing Burnout and Perceived stress. Burnout was expected to fit a 3-factor any psychometric consequences on the data analyses (Graham, structure under an exploratory structural equation modeling (ESEM; 2009). Omega reliability coefficients for both scales were excellent1 : Flow = 0.797, Clutch = 0.790. Asparouhov and Muthén, 2009) approach and perceived stress to Subsequently, we tested the measurement models of FSSS – follow a single-factor structure. Then, we estimated the correlations Version 2 and CSS – Version 2 to confirm the structures refined in among the latent factors of Burnout, Perceived performance, and Study 2. On the one hand, the 7-item single-factor structure of FSSS – Clutch/Flow. Finally, we included these factors in the two structural Version 2 presented an acceptable fit to the data: χ2 (df ) = 79.207 models to test the hypothesized relationships. Both structural models were tested using the ML estimator in Mplus software 7.4, and their 1 We also provide evidence of the reliability of our scales by presenting the fit was assessed using the same cut-off criteria indicated for the results of the alpha coefficients: αFlow Study2 = 0.786, αFlow Study3 = 0.791; measurement models. αClutch Study2 = 0.812, αClutch Study3 = 0.786. TABLE 6 Data distribution and factor loadings of FSSS – Version 2 (Study 3). Data distribution (%) l Item 1 2 3 4 5 6 7 8 9 10 MV Item 1. I felt I was competent enough 0.8 0.4 1.3 1.7 4.5 8.3 17.1 32.5 17.6 15.9 0.0 0.618 to meet the high demands of the situation Item 3. I had a strong sense of what I 0.0 0.6 0.0 0.6 2.6 5.1 15.4 18.6 21.2 36.0 0.0 0.637 must do and what I wanted to achieve Item 4. It was really clear to me that I 0.2 0.8 1.7 3.4 6.2 9.4 22.9 25.9 12.4 17.1 0.2 0.710 was doing well Item 5. I was completely focused on 0.2 0.2 0.6 1.5 2.8 6.6 16.9 24.2 20.3 26.8 0.0 0.662 the task at hand Item 6. I felt in total control of my 0.2 0.2 2.6 3.0 4.3 8.8 15.9 32.3 17.1 15.6 0.0 0.636 mind and body Item 7. I was not worried about what 3.0 3.4 5.1 5.6 7.3 11.8 10.3 13.1 13.9 26.5 0.0 0.564 others may have been thinking of me Item 9. I loved the feeling of that 0.2 0.2 0.4 0.8 2.6 5.6 10.1 15.8 17.6 46.7 0.0 0.428 performance and want to capture it again The most selected category for each item is highlighted in bold. All factor loadings were significant at p < 0.001. TABLE 7 Data distribution and factor loadings of CSS – Version 2 (Study 3). Data distribution (%) l Item 1 2 3 4 5 6 7 8 9 10 MV Item 1. I was fully concentrated; I 0.0 0.2 0.6 2.1 4.7 6.0 13.5 24.6 22.1 26.3 0.0 0.645 couldn’t have concentrated more; I was completely committed to what I need to do Item 2. I was playing better because I 0.2 0.4 0.8 2.8 1.7 9.4 13.7 24.8 20.3 26.1 0.0 0.700 was pushing to my limit consciously, giving my maximum Item 3. Being more aware of my 0.0 0.4 0.2 1.3 5.4 7.3 17.3 28.5 22.9 16.7 0.0 0.602 context and its relevance, I was thinking very clearly about me and what I was doing Item 4. There was a full activation, 0.0 0.4 0.4 0.9 2.3 7.1 12.4 22.5 22.7 31.0 0.4 0.618 with a mix of nerves and excitement, that pumped up my energy Item 5. There’s no worrying about 0.6 0.9 1.1 2.6 6.4 8.3 16.1 20.5 17.8 25.5 0.2 0.664 anything else, I was just focusing on the next move The most selected category for each item is highlighted in bold. All factor loadings were significant at p < 0.001. Frontiers in Psychology 09 frontiersin.org Sánchez Vara et al. 10.3389/fpsyg.2023.1066494 (14), p < 0.001, RMSEA [90% CI] = 0.093 [0.074–0.114], CFI = 0.932, across age and sex. On the other hand, however, invariance could not TLI = 0.898. Moreover, all seven items presented satisfactory factor be assured for the FSSS – Version 2, as different indices pointed to loadings > 0.40 (range = 0.428–0.710; see Table 4). On the other different conclusions. hand, the 5-item single-factor model of the CSS – Version 2 showed a good fit to the data: χ2 (df ) = 14.515 (4), p = 0.006, RMSEA [90% CI] = 0.070 [0.034–0.111], CFI = 0.985, TLI = 0.961. Again, all Validity evidence based on relation to items presented satisfactory factor loadings (range = 0.602–0.700; see other variables Table 5). Measurement model for burnout Regarding the Burnout measurement model, the 3-factor ESEM Measurement invariance models of showed a good fit to the data: χ2 (df ) = 22.066 (12), p = 0.037, RMSEA FSSS – Version 2 and CSS – Version 2 [90% CI] = 0.040 [0.010–0.065], CFI = 0.993, TLI = 0.978. Tables 8, 9 present the goodness-of-fit indices for the Structural models involving flow and clutch states measurement invariance models of FSSS – Version 2 and CSS – As the final step of the analysis, we tested two separate structural Version 2. On the one hand, for the CSS – Version 2, factorial equation models to assess the relationships of Flow and Clutch structure, factor loadings, and item intercepts remained invariant states with Burnout and Perceived performance. On the one hand, TABLE 8 Fit Indexes for the measurement invariance models of FSSS – Version 2 (Study 3). Model χ2(df) RMSEA CFI TLI 1χ2(1df) p 1 RMSEA 1CFI 1TLI (90% CI) Sex Configural 132.097* 0.093 0.903 0.865 (40) (0.076–0.111) Metric 139.162* 0.086 0.903 0.885 7.065 (7) 0.422 −0.007 0.000 0.020 (47) (0.070–0.103) Scalar 163.386* 0.087 0.885 0.881 31.288 (14) 0.005 −0.005 −0.018 0.016 (54) (0.072–0.103) Age Configural 136.039* 0.095 0.906 0.868 (40) (0.078–0.113) Metric 154.522* 0.093 0.894 0.874 18.483 (7) 0.010 0.002 −0.012 0.006 (47) (0.077–0.109) Scalar 165.646* 0.088 0.890 0.886 29.607 (14) 0.009 −0.007 −0.016 0.018 (54) (0.073–0.104) The subsamples for the different measurement invariance analyses are Gender: nfemale = 148, nmale = 382; Age: nU 18 = 290, n> 18 = 240;. df = degrees of freedom; RMSEA = root mean square error of approximation; CI = confidence interval; CFI = Comparative Fit Index; TLI = Tucker-Lewis Index. *p < 0.001. TABLE 9 Fit indexes for the measurement invariance models of CSS – Version 2 (Study 3). Model χ2(df) RMSEA CFI TLI 1χ2(1df) p 1 RMSEA 1CFI 1TLI (90% CI) Sex Configural 45.372* (8) 0.133 0.948 0.869 (0.097–0.172) Metric 54.343* (12) 0.115 0.941 0.901 8.970 (4) 0.062 −0.003 0.000 0.004 (0.085–0.147) Scalar 74.094* (16) 0.117 0.919 0.898 28.722 (8) 0.001 0.000 −0.007 −0.001 (0.091–0.145) Age Configural 17.720* (8) 0.068 0.986 0.965 (0.024–0.111) Metric 23.814* (12) 0.061 0.983 0.972 6.093 (4) 0.192 −0.003 0.000 0.005 (0.023–0.097) Scalar 28.386* (16) 0.054 0.982 0.978 10.666 (8) 0.221 −0.005 0.000 0.008 (0.017–0.086) The subsamples for the different measurement invariance analyses are Gender: nfemale = 148, nmale = 382; Age: nU 18 = 290, n> 18 = 240;. df = degrees of freedom; RMSEA = root mean square error of approximation; CI = confidence interval; CFI = Comparative Fit Index; TLI = Tucker-Lewis Index. *p < 0.001. Frontiers in Psychology 10 frontiersin.org Sánchez Vara et al. 10.3389/fpsyg.2023.1066494 the model involving Flow showed a satisfactory fit to the data: anomalies and new insights that scientists are observing. Several χ2 (df ) = 283.404 (116), p < 0.001, RMSEA [90% CI] = 0.052 [0.044– critics can be seen in different systematic reviews about the flow 0.060], CFI = 0.945, TLI = 0.928. As can be observed in Figure 1, topic (Swann et al., 2012; Goddard et al., 2021) where the classical physical and emotional exhaustion (β = -0.163) and Reduced sense nine-dimensions conceptualization have been put in doubt. For of achievement (β = -0.593) negatively predicted players’ experience example, some authors find that only five dimensions of flow were of the Flow state, which in turn was positively related to Perceived experienced at a specific event (Swann et al., 2012), and others performance (β = 0.585). On the other hand, the model including directly say that flow research has not produced a theory, and, Clutch also presented a satisfactory fit to the data: χ2 (df ) = 246.034 for that, lacks explanatory power (Goddard et al., 2021). In this (84), p < 0.001, RMSEA [90% CI] = 0.060 [0.052–0.069], CFI = 0.940, ongoing process of studying flow, the use of different methodologies TLI = 0.915. In this model (see Figure 2), players’ Reduced sense (Swann et al., 2012, 2016a) produced the identification of what of achievement (β = -0.574) negatively predicted their experience can be understood as a new and different optimal psychological of the Clutch state, which positively led to Perceived performance state: clutch. As a new construct, today there are “significant (β = 0.495). definitional, conceptual, and measurement issues” (p. 24; Schweickle et al., 2020) that need to be studied. This was one of the purposes of this work, using the conceptualization for clutch of Discussion Swann et al. (2017a), and the principles of the Integrated Model of Flow and Clutch States (Swann et al., 2017b, 2019), which In Study 3 we used the short version as a questionnaire, observing describe the differences and similarities between both states. For that the use of short versions reduces the missing data (£ 0.4%), this research, we studied both states at the same time, following improving the quality of the results. With good results in invariance the recommendations of Schweickle et al. (2020) applying the testing for clutch, some statistical problems arise from flow, in line theory of The Integrated Model of Flow and Clutch States (Swann with previous research: sex (Elbe et al., 2010; Chang, 2016) and et al., 2017b, 2019). FSSS and CSS were designed as single-factor age (Jackson et al., 2008). As predicted, the experience of optimal instruments to capture the core aspects of each psychological state. psychological states increases the possibility of perceiving the own In this way, evidence from both exploratory and confirmatory factor performance as better. At the same time, the repetition of the optimal approaches supported the single-factor structure of the two short- psychological experiences, decreases the chances of being involved in forms. burnout processes. We are attending to the growing demands for obtaining more functional and practical measurements in the sporting context (Ramis et al., 2010), in line with the idea that reducing application General discussion times in investigation and practice produce benefits to the research process (Jackson et al., 2008; Leo et al., 2015; García-Alba et al., The research field of the optimal psychological states is “reaching 2021). With this objective, our new tools help in developing a better a crisis point” (p.25, Swann et al., 2018) due to the accumulated conceptual understanding and have several advantages related to FIGURE 1 Structural equation model where players’ flow state mediates the relationship between burnout and perceived performance. ∗ p < 0.05; ∗∗ p < 0.001. Frontiers in Psychology 11 frontiersin.org Sánchez Vara et al. 10.3389/fpsyg.2023.1066494 FIGURE 2 Structural equation model where players’ clutch state mediates the relationship between burnout and perceived performance. ∗ p < 0.05; ∗∗ p < 0.001. their short forms: they can be efficient, place less burden on athletes explained these differences based on sports modalities. Comparing and researchers, are economical, and help to improve the data quality swimmers and football players, it seems that for swimmers perceiving (Alcaraz et al., 2020; Rey et al., 2021). For that purpose, we followed that they are experiencing flow, they would need to sense that the MIMO strategy (Alcaraz et al., 2020) that comprises four phases: the time passes more quickly. But football players, in the other I) Initial selection of the items, II) Expert panel, III) Group focus and hand, who must do fast movements to achieve success, in flow IV) Short form evaluation. Our work is respecting the three main state they would be expecting that time passes more slowly (López- goals of the MIMO process that says the new scales are in consonance Torres et al., 2007; García Calvo et al., 2008). Accepting that sport with the actual literature, people that use them will be satisfied with modality will modulate the time expectancy that athletes have, and, the results, and they will be comprehensible for the participants that coming out of the focus groups we organized, reflections of the answer them. athletes in team sport recognize two kinds of time experience while In relation to flow, we found that the multifactorial model fit they are practicing sport. We can divide these differences into better with seven items than with nine, as it is in the classical two levels: level 1 (micro), related to the moment where they are Csikszentmihalyi definition. From the FSSS, Item 2, related to the executing motor actions (usually a short period of time) and, level union of action and awareness, and Item 8, related to the perception 2 (macro), a larger time related to the game competition or training of time transformation, did not fit with the other dimensions. session. With this double level of time perception, possibly one item Differences in the perception of the subjective experience of flow have cannot express the full experience of time perception. Thus, our 7- been shown for different authors and has been related to different item scale seems to be more universally applicable, avoiding biases causes as the diversity in the sports analyzed (Swann et al., 2012), caused by differences in the type of sport and regional or cultural gender (Elbe et al., 2010; Chang, 2016), or resources (Stander et al., characteristics. 2015). Specifically for Item 2, in previous research in the Spanish In the development of the CSS, we found some problems with context, poor results related to the union of attention and awareness Item 6, where its correlation with other factors was poor. This Item have been seen. For example, López-Torres et al. (2007) found in 6 of CSS together with Item 2 of FSSS are theoretically related swimmers that in their best performances only 20% of the athlete’s in the term that both try to identify the sense of automaticity in sample spoke about the experience of that dimension, and, on their the execution of the action. And, as seen before, some difficulties bad days, no athlete mentioned that dimension. It seems that in the in previous research in Spanish context have been observed. Spanish context, and more related to team sports, athletes do not And for the content of this item, it seems that there could identify their actions during the game as automatic. On the other be a relation with the same structural issue that shows time hand, Item 8 of the flow scale appears to not completely identify perception, based on a double level reality: a first level, where the kind of perceptions athletes have related to the time experience. the individual focus is on the own action, and a second level, Problems in that dimension were indicated in previous research related to the actions of the team (individual actions in relation (Jackson and Marsh, 1996; García Calvo et al., 2008), and authors with the general coordination with the mates) during a specific Frontiers in Psychology 12 frontiersin.org Sánchez Vara et al. 10.3389/fpsyg.2023.1066494 event. Deleting that item, the results indicate that this new scale perception of burnout processes and increase the performance for studying clutch presents fit index values, optimal values of perception of the players. internal consistency, appropriate concurrent validity, and is invariant for gender and age. Our results showed that these two states exist separately, and they are correlated. Moreover, the relation of flow and clutch with Data availability statement other psychological variables were in the expected direction, based on previous research (Bakker et al., 2011): positive for the subjective The raw data supporting the conclusions of this perception of players performance, and negative for burnout. Saying article will be made available by the authors, without that, the flow short scale with seven items (FSSS – Version 2) undue reservation. has very good reliability, as is happening with the 5-item scale related to clutch (CSS – Version 2). On that point, these scales can be good tools to continue investigating the complex subjective experience of the optimal psychological states. The research in Ethics statement this field (reviewed in Schweickle et al., 2020; Goddard et al., 2021) brought us into a point where new scales were needed The studies involving human participants were reviewed and together with new insights into a better definition of flow and approved by Comité de Ética de la Investigación. Written informed its relationship with what is already an optimal psychological consent to participate in this study was provided by the participants’ state: clutch. legal guardian/next of kin. The study of the optimal psychological states brought researchers to the situation of describing two different states that represent similar experiences or that can be alternately experienced inside events but with specific characteristics. First, as validation of Author contributions questionnaires is an ongoing process, future research should continue to assess and improve the short forms presented in this ASV, JC, and SA contributed to conception and design of the study to facilitate a better and deeper understanding of these study. ASV collected the data and organized the database and optimal states. Specifically, further assessment of the invariance wrote the first draft of the manuscript. JC and RM refined the is required, as the FSSS – Version 2 failed to provide clear design of the instruments. SA performed the statistical analysis. evidence supporting its invariance across age. In this line, maybe ASV and SA wrote sections of the manuscript. JC, SA, and the experience of flow is conceptualized depending on the maturity RM contributed to manuscript revision, read, and approved the level of the athlete. In addition, in this study, we focused on submitted version. All authors contributed to the article and developing two efficient single-factor short forms that capture approved the submitted version. the core aspects of each optimal state. However, we understand that, as future research on those states evolves, new and more complex (i.e., multi-factorial) questionnaires would be required. Second, in our research, one limitation of the study is that the Conflict of interest sample included mostly team sports. Another limitation is related to the lack of a priori sample power analysis. In our study, The authors declare that the research was conducted in the we relied on rules-of-thumb (e.g., Nunnally, 1967), but those absence of any commercial or financial relationships that could be rules could be outdated and thus we encourage future research construed as a potential conflict of interest. to conduct an analysis of the sample size requirements before collecting the data (Wolf et al., 2013). And third, it would be interesting to use this short scale in different cultures to compare possible differences from regional contextualization about the two Publisher’s note states. This study shows the benefits of using short forms, relates to All claims expressed in this article are solely those of the the percentage of missing data, and provides evidence supporting authors and do not necessarily represent those of their affiliated their psychometrical merit, obtained with a vast sporting sample. organizations, or those of the publisher, the editors and the reviewers. Together with that, we are continuing with the development of Any product that may be evaluated in this article, or claim that may a better understanding of the subjective experiences related to be made by its manufacturer, is not guaranteed or endorsed by the optimal psychological states. In the breaking point that we are at publisher. in the field of flow research, new pathways are opened to continue researching the differences and similarities between the two optimal states: flow and clutch. These two instruments contribute a series of improvements in the understanding of the two optimal states, Supplementary material and, at the same time, these short forms versions also facilitate its uses in the professional sphere and its administration at different The Supplementary Material for this article can be found online moments over the competition season. These two states are relevant at: https://www.frontiersin.org/articles/10.3389/fpsyg.2023.1066494/ for better sport practices as we have seen that they diminish the full#supplementary-material Frontiers in Psychology 13 frontiersin.org Sánchez Vara et al. 10.3389/fpsyg.2023.1066494 References Alcaraz, S., Jordana, A., Pons, J., Borrueco, M., Ramis, Y., and Torregorsa, M. (2020). Goddard, S., Jackman, P., Stevens, C., and Swann, C. (2021). A systematic review Máxima Información, Mínima Molestia (MIMO): Reducir cuestionarios para cuidar de of flow interventions in sport and exercise. Int. Rev. Sport Exerc. Psychol. 1–36. ∗ v. las personas participantes en psicología del deporte [Maximum Information, Minimum doi: 10.1080/1750984X.2021.1923055 Discomfort (MIMO). Shortening questionnaires to take care of participants in sport Graham, J. W. (2009). Missing data analysis: Making it work in the real world. Ann. psychology]. Inform. Psicol. 119, 49–64. Rev. Psychol. 60, 549–576. doi: 10.1146/annurev.psych.58.110405.085530 Alcaraz, S., Viladrich, C., and Torregrosa, M. (2013). Less time, better quality. Hibbs, D. (2010). A conceptual analysis of clutch performances in competitive sports. shortening questionnaires to assess team environment and goal orientation. Spanish J. J. Philos. Sport 37, 47–59. doi: 10.1080/00948705.2010.9714765 Psychol. 16, 1–14. doi: 10.1017/sjp.2013.74 Arce, C., de Francisco, C., Andrade, E., Arce, I., and Raedeke, T. (2010). Adaptación Hu, L., and Bentler, P. M. (1999). Cutoff criteria for fit indices in covariance structure española del athlete burnout questionnaire (ABQ) para la medida del burnout analysis: Conventional versus new alternatives. Struct. Equ. Model. 6, 1–55. doi: 10.1080/ en futbolistas [Spanish version of Athlete Burnout Questionnaire (ABQ) for the 10705519909540118 measurement of burnout in soccer players]. Psicothema 22, 250–255. Jackson, S. (1992). Athletes in flow: A qualitative investigation of flow states in elite Asparouhov, T., and Muthén, B. (2009). Exploratory structural equation modeling. figure skaters. J. Appl. Sport Psychol. 4, 161–180. doi: 10.1080/10413209208406459 Struct. Equ. Model. 16, 397–438. doi: 10.1080/10705510903008204 Jackson, S. (1995). Factors influencing the occurrence of flow state in elite athletes. Bakker, A., Oerlemans, W., Demerouti, E., Bruins Slot, B., and Karamat, D. A. (2011). J. Appl. Sport Psychol. 7, 138–166. doi: 10.1080/10413209508406962 Flow and performance: A study among talented dutch soccer players. Psychol. Sport Jackson, S., and Eklund, R. C. (2002). assessing flow in physical activity: The flow Exercise 12, 442–450. doi: 10.1016/j.psychsport.2011.02.003 state scale-2 and dispositional flow scale-2. J. Sport Exerc. Psychol. 24, 133–150. doi: Bentler, P. M. (1990). Comparative fit indexes in structural models. Psychol. Bull. 107, 10.1123/jsep.24.2.133 238–246. doi: 10.1037/0033-2909.107.2.238 Jackson, S., and Marsh, H. (1996). Development and validation of a scale to measure Boateng, G. O., Neilands, T. B., Frongillo, E. A., Melgar-Quiñonez, H. R., and Young, optimal experience: The flow state scale. J. Sport Exerc. Psychol. 18, 17–35. doi: 10.1123/ S. L. (2018). Best practices for developing and validating scales for health, social, and jsep.18.1.17 behavioral research: A primer. Front. Public Health 6:149. doi: 10.3389/fpubh.2018.00149 Jackson, S., and Roberts, G. (1992). Positive performance states of athletes: Toward a Browne, M. W., and Cudeck, R. (1993). “Alternative ways of assessing model fit,” in conceptual understanding of peak performance. Sport Psychol. 6, 156–171. doi: 10.1123/ Testing structural equation models, eds K. A. Bollen and J. S. Long (Beverly Hills, CA: tsp.6.2.156 Sage), 136–162. Jackson, S., Ford, S., Kimiecik, J., and Marsh, H. (1998). Psychological correlates of flow Chalip, L., Csikszentmihalyi, M., Kleiber, D., and Larson, R. (1984). Variations of in sport. J. Sport Exerc. Psychol. 20, 358–378. doi: 10.1123/jsep.20.4.358 experience in formal and informal sport. Res. Quart. Exerc. Sport 55, 109–166. doi: Jackson, S., Martin, A., and Eklund, R. (2008). Long and short measure of flow: 10.1080/02701367.1984.10608385 Examining construct validity of the FSS-2, DFS-2, and new brief counterparts. J. Sport Chang, H. (2016). Gender differences in leisure involvement and flow experience Exerc. Psychol. 30, 561–587. doi: 10.1123/jsep.30.5.561 in professional extreme sport activities. World Leisure J. 59, 124–139. doi: 10.1080/ Kimiecik, J. C., and Stein, G. L. (1992). Examining flow experiences in sport contexts: 16078055.2016.1166152 Conceptual issues and methodological concerns. J. Appl. Sport Psychol. 4, 144–160. doi: Chen, F. F. (2007). Sensitivity of goodness of fit indexes to lack of measurement 10.1080/10413209208406458 invariance. Struct. Equ. Model. 14, 464–504. doi: 10.1080/10705510701301834 Leo, F., González-Ponce, I., Sánchez-Oliva, D., Pulido, J., and García-Calvo, T. (2015). Cheung, G. W., and Rensvold, R. B. (2002). Evaluating goodness-of-fit indexes Adaptation and validation in spanish of the group environment questionnaire (GEQ) for testing measurement invariance. Struct. Equ. Model. 9, 233–255. doi: 10.1207/ with professional football players. Psicothema 27, 261–268. doi: 10.7334/psicothema2014. S15328007SEM0902_5 247 Coste, J., Guillemin, F., Pouchot, J., and Fermanian, J. (1997). Methodological López-Torres, M., Torregrosa, M., and Roca, J. (2007). Características del “flow”, approaches to shortening composite measurement scales. J. Clin. Epidemiol. 50, 247–252. ansiedad y estado emocional, en relación con el rendimiento de deportistas de élite doi: 10.1016/S0895-4356(96)00363-0 [Characteristics of "Flow", anxiety and emotional state in relation to the performance of elite athletes]. Cuadernos Psicol. Dep. 7, 25–44. Csikszentmihalyi, M. (1975a). Play and intrinsic rewards. J. Human. Psychol. 15, 41–63. doi: 10.1177/002216787501500306 Marsh, H. W., and Jackson, S. A. (1999). Flow experiences in sport: Construct validation of multidimensional, hierarchical state and trait responses. Struct. Equ. Model. Csikszentmihalyi, M. (1975b). Beyond boredom and anxiety. San Francisco, CA: 6, 343–371. doi: 10.1080/10705519909540140 Jossey-Bass Publishers. Marsh, H. W., Muthén, B., Asparouhov, A., Lüdtke, O., Robitzsch, A., Morin, A. J. S., Csikszentmihalyi, M. (1981). Leisure and socialization. Soc. For. 60, 332–340. doi: et al. (2009). Exploratory structural equation modeling, integrating CFA and EFA: 10.2307/2578438 Application to students’ evaluations of university teaching. Struct. Equ. Model. 16, Csikszentmihalyi, M. (1986). Reply to the commentaries by Diaz-Guerrero, freedman, 439–476. doi: 10.1080/10705510903008220 Gergen, Milgram and Martin, Ryan and Deci, and Stettner. New Ideas Psychol. 4, 53–59. McDonald, R. P. (1970). The theoretical foundations of principal factor analysis and doi: 10.1016/0732-118X(86)90058-9 alpha factor analysis. Br. J. Math. Stat. Psychol. 23, 1–21. doi: 10.1111/j.2044-8317.1970. Csikszentmihalyi, M. (1990). Flow. The psychology of optimal experience. New York, tb00432.x NY: Harper & Row. Muthén, L. K., and Muthén, B. O. (2015). Mplus editor (versión 7.4) [Computer Csikszentmihalyi, M. (1992). A response to the Kimiecik & Stein and Jackson papers. Software]. Los Angeles, CA: Author. J. Appl. Sport Psychol. 4, 181–183. doi: 10.1080/10413209208406460 Nunnally, J. C. (1967). Psychometric theory. New York, NY: McGraw-Hill. Csikszentmihalyi, M. (1999). If we are so rich, why aren’t we happy? Am. Psychol. 54, 821–827. doi: 10.1037/0003-066X.54.10.821 Ramis, Y., Torregrosa, M., Viladrich, C., and Cruz, J. (2010). Adaptación y validación de la versión española de la escala de ansiedad competitiva SAS-2 para deportistas de De Francisco, C. (2015). Versión reducida del athlete burnout questionnaire (ABQ): iniciación [Adaptation and validation of the Spanish version of the sport anxiety scale Propiedades psicométricas preliminares. Revista Psicol. Dep. 24, 177–183. SAS-2 for young athletes]. Psicothema 22, 1004–1009. de Nakamura, J., and Csikszentmihalyi, M. (2014). “The concept of flow,” in Flow and Rey, R., Ojeda, M., and Casa, J. (2021). Validation of the Sexting behavior and the foundations of positive psychology: The collected works of mihaly csikszentmihalyi, ed. motives questionnaire (SBM-Q). Psicothema 33, 287–295. doi: 10.7334/psicothema M. Csikszentmihalyi (Berlin: Springer), 239–263. 2020.207 Elbe, A. M., Strahler, K., Krustrup, P., Wikman, J., and Stelter, R. (2010). Experiencing Rheinberg, F., Vollmeyer, R., and Engeser, S. (2003). “Die Erfassung des Flow-Erlebens flow in different types of physical activity intervention programs: Three randomized [The assessment of Flow Experience],” in Diagnostik von Selbstkonzept. Lernmotivation studies. Scandin. J. Med. Sci. Sports 20, 111–117. doi: 10.1111/j.1600-0838.2010.01112.x und Selbstregulation, eds J. Stiensmeier-Pelster and F. Rheinberg (Gottingen: Hogrefe), Engeser, S., and Rheinberg, F. (2008). Flow, performance and moderators of challenge- 261–279. skill balance. Motivat. Emot. 32, 158–172. doi: 10.1007/s11031-008-9102-4 Schweickle, M. J., Swann, C., Jackman, P. C., and Vella, S. A. (2020). Clutch Fernández, M. A., Godoy-Izquierdo, D., Jaenes, J. C., Bohórquez, M. R., and Vélez, M. performance in sport and exercise: A systematic review. Int. Rev. Sport Exerc. Psychol. (2015). Flow y rendimiento en corredores de maratón. Revista Psicol. Dep. 24, 9–19. 14, 102–129. doi: 10.1080/1750984X.2020.1771747 García Calvo, T., Jiménez, R., Santos-Rosa, F. F., Reina, R., and Cervelló, E. (2008). Sinnett, S., Jäger, J., Singer, S. M., and Antonini, R. (2020). Flow states and associated Psychometric properties of the spanish version of the flow state scale. Span. J. Psychol. 11, changes in spatial and temporal processing. Front. Psychol. 11:381. doi: 10.3389/fpsyg. 660–669. doi: 10.1017/S1138741600004662 2020.00381 García-Alba, L., Postigo, A., Gullo, F., Muñiz, J., and Del Valle, J. (2021). PLANEA Stander, F. W., Rothmann, S., and Botha, E. (2015). The role of teammate relationships, Independent life skills scale: Development and validation. Psicothema 33, 268–278. doi: communication and self-efficacy in predicting athlete flow experience. J. Psychol Afr. 25, 10.7334/psicothema2020.450 494–503. doi: 10.1080/14330237.2015.1124600 Frontiers in Psychology 14 frontiersin.org Sánchez Vara et al. 10.3389/fpsyg.2023.1066494 Steiger, J. H., and Lind, J. C. (1980). “Statistically-based tests for the number of common Swann, C., Keegan, R., Piggott, D., and Crust, L. (2012). A systematic review of the factors,” in Paper presented at the annual spring meeting of the psychometric society, Iowa experience, occurrence, and controllability of flow states in elite sport. Psychol. Sport City, IA. Exerc. 13, 807–819. doi: 10.1016/j.psychsport.2012.05.006 Swann, C., Crust, L., and Vella, S. A. (2017a). New directions in the psychology of Swann, C., Piggott, D., Schweickle, M., and Vella, S. A. (2018). A Review of scientific optimal performance in sport: Flow and clutch. Curr. Opin. Psychol. 16, 48–53. doi: progress in flow in sport and exercise: Normal science, crisis, and a progressive shift. 10.1016/j.copsyc.2017.03.032 J. Appl. Sport Psychol. 30, 249–271. doi: 10.1080/10413200.2018.1443525 Swann, C., Crust, L., Jackman, P. C., Vella, S. A., Allen, M. S., and Keegan, R. Tse, D. C., Nakamura, J., and Csikszentmihalyi, M. (2019). Beyond challenge-seeking (2017b). Psychological states underlying excellent performance in sport: Towards an and skill-building: Toward the lifespan developmental perspective on flow theory. integrated model of flow and clutch states. J. Appl. Sport Psychol. 29, 375–401. doi: J. Positive Psychol. 15, 171–182. doi: 10.1080/17439760.2019.1579362 10.1080/10413200.2016.1272650 Tucker, L. R., and Lewis, C. (1973). A reliability coefficient for maximum likelihood factor analysis. Psychometrika 38, 1–10. doi: 10.1007/BF02291170 Swann, C., Crust, L., Jackman, P., Vella, S. A., Allen, M. S., and Keegan, R. (2016a). Performing under pressure: Exploring the psychological state underlying clutch Verner-Filion, J., and Vallerand, R. J. (2018). A longitudinal examination of elite performance in sport. J. Sports Sci. 35, 2272–2280. doi: 10.1080/02640414.2016.126 youth soccer players: The role of passion and basic needs satisfaction in athletes 5661 optimal functioning. Psychol. Sports Exerc. 39, 20–28. doi: 10.1016/j.psychsport.2018. 07.005 Swann, C., Jackman, P. C., Schweickle, M. J., and Vella, S. A. (2019). Optimal experiences in exercise: A qualitative investigation of flow and clutch states. Psychol. Sport Vlachopoulos, S., Karageorghis, C., and Terry, P. (2000). Hierarchical confirmatory Exerc. 40, 87–98. doi: 10.1016/j.psychsport.2018.09.007 factor analysis of the flow state scale in an exercise setting. Depart. Phys. Educ. Sports Sci. 18, 815–823. doi: 10.1080/026404100419874 Swann, C., Keegan, R. J., Crust, L., and Piggott, D. (2016b). Psychological states underlying excellent performance in professional golfers: “Letting it happen” vs. Wolf, E. J., Harrington, K. M., Clark, S. L., and Miller, M. W. (2013). Sample size “making it happen”. Psychol. Sport Exerc. 23, 101–113. doi: 10.1016/j.psychsport.2015. requirements for structural equation models: An evaluation of power, bias, and solution 10.008 propriety. Educ. Psychol. Meas. 76, 913–934. doi: 10.1177/0013164413495237 Frontiers in Psychology 15 frontiersin.org