Validation of the Undesirable Behavior Strategies Questionnaire: Physical Educators ’ Strategies within the Classroom Ecology

: This study sought to propose the Undesirable Behavior Strategies Questionnaire – Physical Education (UBSQ-PE), based on Doyle, W. (1977). Paradigms for research on teacher e ﬀ ectiveness. Review of Research in Education , 5(1), 163 – 198. classroom ecology paradigm, and validate its use within Physical Education. 289 physical educators participated in the study (80 of them in the pilot study) and completed a survey evaluating the strategies they used across the classroom ecology task systems. An exploratory factor analysis proposed a four-factor solu-tion that measured teachers ’ strategies for Di ﬀ erentiation of Instruction, Interaction for Students, Classroom Management, and Interaction with Students and explained a cumulative 47.45% of the total explained variance (KMO = 0.809, χ 2312 = 1,266, 684, p < 0.001). A con ﬁ rmatory factor analysis indicated a good ﬁ t of the data to the four-factor solution ( χ 2/df ratio = 1.52, comparative ﬁ t index = 0.906, Tucker Lewis index = 0.890, standardized root mean square residual = 0.079, root mean square error of approximation = 0.05), providing evidence for the instrument ’ s internal consistency. Multiple analyses of variance showed that there were di ﬀ er-ences in teachers ’ strategies based on their gender, age, studies, professional knowledge, and school level. The use of UBSQ-PE asserts that teachers use undesirable behavior (UB) strategies according to their personal and professional characteristics. Teacher education programs could use this information for supporting PE teachers to adjust their UB strategies in response to each classroom ’ s interrelated task systems.

Teachers' instructional approach and professional understandings seem to be rather important in terms of preventing or dealing with students' UB.Teachers who can adjust their program of action based on the interactions occurring within the learning environment are more effective when dealing with UBs (Barker & Annerstedt, 2014;Pozo et al., 2018).Teachers' effectiveness in this field may develop along the years of experience (Grube, Ryan, Lowell, & Stringer, 2018), or because of their participation in professional development programs (Morrison & Gleddie, 2019).However, until today, studies have shown that PE teachers learn to deal with students' UB rather empirically, based on their personal characteristics, or contextual influences (Cothran & Kulinna, 2015;Kulinna, Cothran, & Regualos, 2006).
For this reason, many recent studies advocate for the adoption of complexity thinking approaches in teachers' preparation/support to deal with students' UBs (Grube et al., 2018;Grube, Little, & Stringer, 2023;Hovdal et al., 2021;Mercanoglu, 2021).In educational research, complexity thinking focuses on the social nature of teaching and learning, and the multidimensional interactions occurring among human and nonhuman elements (e.g., teachers, students, curriculum, social relationships, objects, spaces, etc.) (Davis & Sumara, 2006).These interactions are responsible for the emergence of patterns or phenomena that are nonlinear and diverse in their nature and may create a disequilibrium (like an UB) to the lesson's flow or effectiveness.It is important therefore, to support teachers to employ a complexity thinking perspective when dealing with UBs, so that they can reflect on the logic underpinning their decisions and handle students' behaviors by adopting a proactive approach (e.g., promote desirable behaviors).

Study of UB in PE
The teacher's ability to select contextually relevant strategies for preventing or correcting UB is an element that may safeguard lesson flow and effectiveness (Lindsay, 2014;Simonton, Garn, & Solmon, 2017).Thinking of PE as a complex and open system of social relationships (Hastie, Rudisill, & Boyd, 2016;Postholm, 2013;Supaporn, Dodds, & Griffin, 2003), it seems that factors related to the constraints of the physical environment (e.g., open or noisy courts) (Cothran & Kulinna, 2015), the simultaneous movement of bodies and objects (Ryan & Mendel, 2010), or teachers' difficulty to align with students' needs can further trigger the occurrence of UB.Studies show that PE teachers adopt mainly a reactive rather than a proactive approach for dealing with UBs, and this relates to their personal/ professional experiences and characteristics (Barker & Annerstedt, 2014;Hovdal et al., 2021).It seems that female teachers usually face more severe incidents of UB than their male colleagues in primary schools (Kulinna et al., 2006), although men appear to have higher occurrences of UB in high schools compared to women (Kulinna et al., 2006).At the primary level, PE teachers spend a lot of time dealing with off-task behaviors or misbehaviors (James & Collier, 2011), at the expense of facilitating student learning (Hovdal et al., 2021).On the other hand, middle or high school teachers show deficiencies in implementing classroom management or teaching strategies to prevent UBs (Barker & Annerstedt, 2014;Navarro-Paton, Lago-Ballesteros, & Arufe-Giraldez, 2020) and therefore apply strategies that only temporarily reduce UBs (Hovdal et al., 2021).
Students' UB is gaining attention as a topic in relevant PE literature.Studies categorize UB based on their level of seriousness (Goyette, Doré, & Dion, 2000), their type and occurrence within the PE classroom (Kulinna, Cothran, & Regualos, 2003), their different dimensions (Krech, Kulinna, & Cothran, 2010), or the way that they may interfere with peer learning (Lin, Lin, & Wang, 2008).What seems to be common in these studies is that researchers focus on the type of UB or the factors that trigger their occurrence and use theories such as achievement goal theory (Nicholls, 1984) or social cognitive theory (Bandura, 1991), to explain or interpret students' (mis)behavior (Wang, Biddle, & Elliot, 2007;Warburton & Spray, 2008).In general, the emphasis placed on students' UB could be a possible way forward to the study of their nature as well as the mechanisms that justify their occurrence.However, at the level of PE teaching, the determinants of UB need to be identified through theoretical frameworks that adopt a complexity thinking perspective to the study of classroom life.

Doyle's Theory of Classroom Ecology
A theory that may provide a complexity thinking perspective for the study of students' UB is Doyle's ecological paradigm (Doyle, 1977(Doyle, , 1986)).According to Doyle's theory (1986), the establishment of an effective learning environment depends on the dynamic and continuous interaction between three basic systems of activities: (a) classroom management, (b) teaching, and (c) social relations.These systems are complex and interrelated, and any change in one of them will result in a change and/or an adjustment in the others (Doyle, 1977).Classroom ecology is a higher order system that is greater than the sum of its parts.Thus, though at times the structure of each task system may depend more or less on complex subsystems (e.g., routines and protocols, teaching methods, peer interactions, etc.), ultimately it is their intersection that will shape the classroom ecology.The ideas of classroom order (management rules, routines, expectations) and academic work (presentation, practice of subject matter) were inherent in Doyle's initial ecology paradigm (Hastie et al., 2016).The social task system was first emphasized by Allen (1986) and added later (McCaughtry, Tischler, & Flory, 2008), denoting the importance of student social agenda and social interactions in the classroom ecology.According to Hastie and Siedentop (2006, p. 222), "the main benefit of research using the ecological model is that it presents a realistic description of life in the gym."To fully appreciate life in the gym, the PE teacher must be able to understand the interdependent and dynamic nature of each task system.For this reason, Doyle's paradigm offers a rather valuable framework for the development of relevant evaluation models.
By acknowledging the usefulness of the above approaches, we think that it would be useful to develop an instrument that could ecologically evaluate the strategies PE teachers use when dealing with UBs.Such an instrument could help PE teachers understand what type of strategies they use and further reflect on what needs to be done to safeguard lesson effectiveness across each task system of their classroom's ecology.

Purpose and Significance of the Present Study
Based on the above, the current study aimed to investigate the strategies used by teachers to deal with students' UBs within school PE.To achieve this purpose, the Undesirable Behavior Strategies Questionnaire -Physical Education (UBSQ-PE) was developed according to Doyle's classroom ecology paradigm and was used to explore possible differences in PE teachers' strategies based on their gender, age, studies, professional knowledge, and school level.The research objectives were three-fold: (1) the development of the instrument; (2) the evaluation of its psychometric properties through exploratory factor analysis (EFA) and confirmatory factor analysis (CFA); (3) the examination of the extent to which the instrument could be used to trace differences in PE teachers' strategies depending on their personal and professional characteristics.We believe that the results of this study can provide useful information concerning PE teachers' use of UB strategies across the complex and interrelated systems of Doyle's paradigm.Such information can be beneficial for the design of professional learning programs that could support PE teachers to learn how to act proactively towards the nurturing of desirable student behaviors across the task systems of each classroom's ecology.

Participants
A total of 289 PE teachers (120 men and 169 women), aged between 31 and 55, from several regions in Greece participated in this study.Their teaching experience in different grade levels ranged from 5-10 years (45%) to 11-20 years (28.2%) and 26.2% PE teachers had more than 21 years of teaching experience.Regarding their demographic characteristics, participants were heterogeneous only in terms of their professional development in issues concerning UB (Table 1).The criteria for their participation in the study was being a PE teacher with at least 5 years of teaching experience and the delivery of signed informed consent.As described below, 80 PE teachers from the total sample participated in the pilot study and 209 in the main study.

Research Methodology
The research methodology consisted of two parts: (a) the development of UBSQ-PE through a pilot study, and (b) the examination of UBSQ-PE validity and reliability through a main study.Following relevant guidelines for scale development and analysis (Brown, 2006;Tabachnick, Fidell, & Ullman, 2013), item generation and content validity tests were conducted for the newly developed UBSQ-PE during the pilot study.The main study involved item reduction, statistical significance, and reliability testing processes, through UBSQ-PE administration processes.The main goal was to ensure that the newly developed instrument could be adequately used by academic and practitioner audiences to measure the UB construct via an ecological approach.The procedures followed during each part of the study are described here.

Pilot Study: Instrument Development
The UBSQ-PE construction process began with the development of items to assess UB.This process was carried out (a) inductively, generating items from relevant literature and (b) deductively, starting with theoretical definitions based on Doyle's paradigm, from which items were then generated.To ensure content validity, items that evaluated the UB strategies used by PE teachers were selected through a process that involved: (1) literature review; (2) creation of a pool of items; and (3) the conduct of pilot tests.The review of the literature focused on the construct of UB in general educational literature, as well as on studies examining the causes and categories of UB.Furthermore, the ways in which teachers experience and handle UBs within their classrooms were also examined, including the literature examining teacher burnout (Tsouloupas & Carson, 2016), well-being (Aldrup, Klusmann, Ludtke, Gollner, & Trautwein, 2018), appraisal, regulation, coping (Chang, 2013), self-efficacy (Tsouloupas, Carson, & Matthews, 2014), values, and beliefs (Atiles, Gresham, & Washburn, 2017).This literature provided us with a framework for conceptualizing the dimensions of classroom ecology in relation to UB, as well as developing a valid and reliable instrument to measure teachers' strategies for dealing with UB.Subsequently, relevant PE literature was consulted to gain an understanding of teachers' willingness to deal with or prevent UB and their use of instructional models for this purpose from an ecological perspective.In particular, PE studies that focused on the teacher-student relationship (Hastie, 2016;Hastie et al., 2016;Hastie & Siedentop, 2006;Hsu, Shang, & Hsiao, 2021), students' reaction to lesson content (Hsu et al., 2017;Supaporn et al., 2003), or their disruptive behaviors within the classroom (Krech et al., 2010) were also used.Finally, we consulted literature for instrument development generally (e.g., Brown, 2006;Tabachnick et al., 2013), as well as specifically in PE (e.g., Richards, Gaudreault, & Woods, 2017) to inform our process.
Following the literature review, a total pool of 50 items related to the UB of the students were created.These items were modified and then expressively adapted to be compatible with the characteristics of the PE context.Items were shared with 20 PE teachers whose scholarship and academic background focused on PE teaching.The selection of the 20 PE teachers was based on their expertise in the field, since all of them held a PhD degree and had previously participated in professional development programs with a focus on instructional methods, curriculum design, mentoring and supervision, and behavior management.The definition of each system of Doyle's paradigm was provided and PE teachers were asked to assign each item to one of the three systems: classroom management, teaching, and social relations.Teachers evaluated the item matching per system, using a 5-point Likert scale (1 = the item does not at all reflect the system, 5 = the item reflects the system too much).After this process, a total of 25 items (out of the initial 50) were included in the UBSQ-PE instrument.In particular, nine items related to the classroom management system (e.g., when an UB occurs, I deprive the student of the privilege to participate in his/her favorite activity, or I observe the students during the lesson to determine what the UB is and what is causing it, etc.), eight items related to the teaching system (e.g., I apply differentiated instruction techniques to prevent UB from occurring, or I incorporate the teaching of desirable behavior into my classroom's learning objectives), and eight items related to the social relations system (e.g., I encourage communication and cooperation among students within the lesson, or I seek the help of other teachers/colleagues to deal with the unwanted behavior).
The items that were retained were reported with a minimum level of 80% agreement between the 20 PE teachers (Carmines & Zeller, 1979) and are presented in Appendix 1.The 25 items that were not included in the questionnaire (e.g., I use humor as a means of communicating with students to prevent/deal with UB, or, when I plan my lessons I link student assessment to his/her in-class behavior, or, when the UB occurs I raise my tone of voice to the students, etc.) revealed low agreement rates among the PE teachers.Subsequently, a survey was conducted with 60 PE teachers (40 female, 20 male), with the purpose of evaluating the 25 UBSQ-PE items for clarity and comprehension.After the appropriate modifications, the questionnaire was ready to be used in the main study.

Main Study: Procedures and Instrumentation
Following the development of the UBSQ-PE, an email was sent containing an overview of the study and a link to the online survey to 300 PE teachers from various areas of Greece.After providing informed consent, a total of 209 PE teachers completed the survey (see Table 1 for demographics).
The survey included 30 questions in total: a 5-item demographic questionnaire, the 25 items of the UBSQ-PE, and the short form version of the Social Desirability scale (Crowne & Marlowe, 1960).The Social Desirability Scale measures desirable responses through participants' selfreports (true or false), or otherwise their tendency to provide responses that have a social desirability bias and are not reflective of their own opinions.The internal consistency reliability of the Social Desirability Scale was good in the current study (Cronbach's a = 0.87), like other studies that have used it in the past (Campbell & Stein, 2007).
Based on the responses of the participants, EFA was initially applied to identify the UBSQ-PE factor structure.EFA explores relationships and groups the items that are highly correlated (Tabachnick et al., 2013).Principal axis factoring with oblimin rotation was used for factor extraction.The 25-item model was run and factors with eigenvalues greater than 1 were extracted.The items that did not load significantly or cross-loaded on a theoretically unjustifiable factor were removed (i.e., factor loading > 0.30; Brown, 2006).The internal consistency of the instrument was evaluated using Cronbach's alpha indices.Subsequently, a CFA was used to confirm the factor structure that would be suggested through EFA (Loehlin, 2004).As recommended by Brown (2006), goodness-of-fit indicators were used to evaluate the fit of the model in the CFA model.Model fit indices of χ 2 /df, comparative fit index (CFI), Tucker Lewis index (TLI), root mean square error of approximation (RMSEA), and standardized root mean square residual (SRMR) were produced.In general, a chi-square that is non-significant means that the data have a good fit to the model, yet this statistic is sensitive to sample size, thus it was presented out of conversion (Wheaton, Muthen, Alwin, & Summers, 1977) and not as a standalone CFA criterion (West, Taylor, & Wu, 2012).Furthermore, to evaluate goodness of fit, CFI, SRMR, and RMSEA were used with good fit thresholds of CFI > 0.90, SRMR < 0.08, and RMSEA < 0.08 (Dagnall, Denovan, Parker, Drinkwater, & Walsh, 2018).Modification indices (MI) were inspected with values higher than 20 (Dagnall et al., 2018), while only errors from items within the same subfactors were covaried to improve model fit (Byrne, 2010).Multiple Analysis of Variance (MANOVA) was also used to examine differences in the factors extracted from the UBSQ-PE based on participants' gender, age, studies, professional development seminars, and school level.SPPS 21.0 software was used for data cleaning, descriptive statistics, and EFA, while CFA analysis was performed using JAMOVI software.All research activities were approved by the research ethics board of the university of the first author.

EFA and CFA
The instrument validation process started with the use of EFA and initially extracted five factors based on the Scree plot test (Cattell, 1966) In the five-factor solution, the item "I take time to resolve conflicts within the class" was loaded in two factors and was thus removed.This was a confounding Validation of the Undesirable Behavior Strategies Questionnaire  5 item since in relative literature, conflict resolution may be handled with various strategies or styles (Hovdal et al., 2021).The items "I use physical exercise as punishment (e.g., push-ups, running) to deal with unwanted behavior" and "I rigorously apply pre-agreed consequences in the event of deviation from the rules/routines/expectations I have set" were removed due to low loading values and conceptual misalignment with other factor items.The final solution extracted four factors based on the Scree plot (Figure 1).According to Field (2016), when the communalities are low (<0.6) and less than 250 individuals participate in a study, the final number of extracted factors in EFA is more accurately determined by the Scree plot (pp.740-742).In our case, the four-factor solution explained a cumulative 47.45% of the total explained variance (KMO = 0.809, χ 231 2 = 1,266,684, p < 0.001) (Table 2).
The first factor included seven items (Cronbach a = 0.80) to measure Differentiation of Instruction Strategies (24.5% of the variance), the second factor consisted of four items (Cronbach a = 0.66) to measure Interaction for Students Strategies (8.75% of the variance), the third factor included five items (Cronbach a = 0.54), to measure Classroom Management Strategies (7.84% of the variance), and the fourth factor included six items (Cronbach a = 0.75) to measure Interaction with Students Strategies (6.26% of the variance).Based on Doyle's paradigm, the first factor related to the teaching task system, the third factor to the classroom management task system, and the second and fourth factors to the social relations task system (i.e., teachers' interactions for and with students).CFA indicated a good fit of the data to the four-factor solution (χ 2 /df ratio = 1.52,CFI = 0.906, TLI = 0.890, SRMR = 0.079, RMSEA = 0.05).Model MI did not recommend any justifiable changes to the model, and all factor loadings were significant (Table 3).This supported the fact that the items were good indicators of the underlying factors.

Sample Differences
Using the same sample that was used to conduct CFA, we examined the bivariate correlations among the study constructs and the Social Desirability Scale.The results showed that Social Desirability was weakly correlated (r = -0.161,p < 0.05) only with the factor of Classroom Management Strategies of UBSQ-PE.This finding could be attributed to the fact that some of the items in this factor referred to strategies that were mainly reactive in nature (e.g., to gain control of students' UB). MANOVA

Discussion
The aim of this study is to develop a UBSQ-PE, an ecologically relevant instrument to examine UB strategies of teachers and provide initial validity and reliability evidence for its use in PE.Following a two-part methodological approach, using EFA and CFA, as well as MANOVA tests, the results of the present study provided support to the factorial validity and reliability of UBSQ-PE.The four factors of UBSQ-PE, namely, Differentiation of instruction, Interaction for students, Classroom management, and Interaction with students allow us to investigate UB ecologically.As an application of Doyle's classroom ecology paradigm (Doyle, 1977), UBSQ-PE asserts that teachers use UB strategies in different systems of classroom ecology (e.g., classroom management, teaching, and social relations) and in regard to their personal and professional characteristics.The results are discussed here along with Doyle's systems of classroom ecology.
As noted in our study, early career and primary school teachers were concerned about classroom management, more than PE teachers working in secondary education or teachers who had spent more years in schools.Similar findings are reported in relevant literature (Garrahy, Cothran, & Kulinna, 2005;Goyette et al., 2000;Oliveira & Graca, 2013;O'Neill & Stephenson, 2014;Tsouloupas & Carson, 2016;Zhang, 2021).With the change in grade level or with the advent of teacher experience, the classroom ecology changes.This may happen either because students lose their interest in lesson activities as they grow older (Cothran & Kulinna, 2007;Kulinna et al., 2006), or because teachers adopt loose accountability systems for keeping students engaged (James & Collier, 2011).Based on Doyle's paradigm, managerial classroom systems that focus on cooperation rather than compliance produce a positive classroom climate that helps students behave with respect and responsibility during classroom activities.Thus, if PE teachers can engage students actively in the classroom managerial system (e.g., establish together rules for appropriate behavior, allow students to co-decide classroom routines and protocols), instead of reducing demands in their instructional system, as it usually happens in secondary education (Siedentop, 2002), then a more positive classroom climate will be established.Relevant studies show that UBs may be the result of a "casual" learning environment, or a consequence of the teacher's inability to accommodate the student social system within the instructional system (Hastie, 1995;Supaporn et al., 2003).That is the reason why classroom management needs to shift to complexity thinking understandings of the classroom as a "social system" (Postholm, 2013), within which teachers may adopt either a teacher-as-friend stance (Tilga, Hein, Koka, & Hagger, 2020) or student-centered social task systems (Hastie & Siedentop, 2006), so that students are encouraged to experience personal relevance and value in classroom activities (Hovdal et al., 2021).Social interaction strategies such as counseling, recreational enrichment, expressing and recognizing emotions, etc., may be used as effective tools for keeping students on task and avoiding overly regulated management systems (Oliveira & Graca, 2013).
Concerning classroom instruction, it was critical to note that the teachers in this study dealt differently with UBs to increase their instructional time.Female PE teachers were more willing to differentiate their instruction than their male colleagues.Similarly, older PE teachers or teachers who had attended postgraduate studies or professional development seminars differentiated their instruction more, compared to younger colleagues or colleagues with limited knowledge in this field.These findings are supported by literature showing that female teachers, as well as teachers who have more years of teaching or professional experience, seek to adopt a variety of coping strategies within their classrooms by reflecting on what is happening during instruction (Levy & Khoury-Kassabri, 2021;Tsouloupas et al., 2014).Based on Doyle's paradigm  Validation of the Undesirable Behavior Strategies Questionnaire  9 (1986), teaching is effective when teachers can locate key resources or events in the dynamics of the classroom environment and develop/modify their program of action accordingly.The four factors of the UBSQ-PE instrument can provide explicit evidence on the UBs that may occur across different task systems as a part of the adoption of different instructional strategies within PE.As it is evident from many studies, the effectiveness of an instructional strategyas initially designed or exhibited by the PE teacherdepends on the negotiations that will take place during its implementation (Barney & Christenson, 2013;Barney & Prusak, 2016;Barney & Strand, 2006;Strand & Bender, 2011).These negotiations may be either verbal, motor, or social.Thus, it is important for the teacher to observe what is happening and afterwards respond with the selection of the appropriate strategy.Doyle (1986) suggested that teachers need to keep a balance between the primary vectors of classroom activities (e.g., action, order) and student-initiated responses (e.g., changes to lesson content, engagement, etc.).Studies show that female teachers adopt affective curriculum goals and participatory processes as primary vectors (Burke & Collins, 2001;Pounder & Coleman, 2002), while male teachers are more conservative and prefer to adopt procedural leadership teaching styles (Eagly, Johannesen-Schmidt, & Van Engen, 2003).Overall, the more teaching or professional experience that teachers gain, the more confident they feel to implement their program of action, despite the presence of UBs (James & Collier, 2011;Zee, de Jong, & Koomen, 2016).The framework adopted in the development of the UBSQ-PE instrument allows us to study teachers' effectiveness in dealing with UBs in relation to the complexity of each classroom's ecology.Since effective PE teaching is closely related to the purposefulness of instruction (Hovdal et al., 2021), the selection of authentic and context-relevant instructional tasks will ultimately produce a task-oriented learning ecology within which the occurrence of UB will be minimized (Cervelló, Jiménez, del Villar, Ramos, & Santos-Rosa, 2004;Claver, Martínez-Aranda, Conejero, & Gil-Arias, 2020;Moreno-Murcia, Sicilia, Cervelló, Huéscar, & Dumitru, 2011).
From an ecological point of view, over-regulatory teaching styles bring short-term reductions to UB, since they fragment student-generated responses to the classroom social system (Doyle, 1986).The development of cautious thinking by the teachers is necessary to improve their ability to understand the consequences of their instructional decisions and choose possible alternatives (Hovdal et al., 2021;Pozo, Grao-Cruces, & Pérez-Ordás, 2018).Instructional activities that hold students socially accountable while caring for their needs can maintain the flow of the lesson and the commitment of students to learning (Hastie et al., 2016;Hovdal et al., 2021;Lindsay, 2014;Supaporn et al., 2003).When lesson flow is protected, teachers are more able to meaningfully reflect on the factors that influence student learning (Hastie, 2016).In the long run, such an ability can increase teachers' confidence in their teaching and sense of mattering (Richards et al., 2017).It therefore seems that PE teacher education, or professional development programs should be designed with a focus on preparing future practitioners with the skills and knowledge needed to keep the flow of classroom ecology balanced, despite potential student pressures.The use of models such as Teaching Personal and Social Responsibility, Adventure Education, or Cooperative Learning has been suggested in ecological studies as an effective means for this purpose (Carlson & Hastie, 1997;Dyson & Strachan, 2004;Hastie, 1995Hastie, , 2000;;Sinelnikov & Hastie, 2008).The different factors of the UBSQ-PE instrument can allow researchers to trace differences in UB strategies based on the structure and activities of each curriculum model.
Overall, regardless of their age and professional knowledge, all participants in this study recognized the importance of the social task system as a way to deal with UB.As reported in similar studies (Dyson et al., 2010;McCaughtry et al., 2008), social relation strategiesadopted as part of the teacher's interaction with students or for students (i.e., with colleagues or parents)can improve lesson quality, by changing the lesson's focus from managing students UB to encouraging their appropriate behavior (Barker & Annerstedt, 2014;Shingjergji, 2016), or fostering supportive behavior (Ahmadi et al., 2023).Researchers favorably report on the use of social relation strategies, as a way of strengthening teacher-student-parent relations within the school context (e.g., one-on-one discussions, developing mutual respect, getting to know others) (Garrahy et al., 2005;Goyette et al., 2000), or within the wider school community (e.g., family discussions, teacher-student communication, respect for the rights of all students) (Hovdal et al., 2021).The above strategies establish a climate of moral engagement and mutual support (Hsu et al., 2017), which can afterwards be transferred from the PE classroom to everyday life situations (Sullivan, Johnson, Owens, & Conway, 2014).This is rather important if we consider that ecological analyses within a PE classroom may extend beyond the PE subject to the wider school ecology.Towards this direction, the newly developed instrument can bring more caution concerning the co-dependency between PE courses and school administration support programs.
Finally, male teachers in this study had lower scores than female teachers regarding their social relation strategies with and for students.While there is no single explanation for this difference, it is worth taking it into account, so as to sufficiently support all teachers to understand and deal with the interrelationships taking place within the classroom's ecology.Based on our results, the lower scores of male PE teachers in their social relation strategies may relate to the reported conservatism of male teachers during the design of instruction (Eagly et al., 2003).Given the importance of socioemotional learning processes for every student's holistic development (Richards, Ivy, Wright, & Jerris, 2019), it would be worth implementing professional development programs for all PE teachers, within which they could be introduced to the principles of social connectedness and belonging (Klemola, Heikinaro-Johansson, & O'Sullivan, 2013), so that they can integrate them during instruction.In essence, we call for an ecological integration of socio-emotional learning in future pedagogical work in the context of PE, so that several aspects of classroom dynamics (e.g., task presentation, student-teacher interactions, student negotiation of content and tasks, etc.) can be addressed in their interdependence and with a focus on student equity.

Conclusion
In this study, Doyle's ecological paradigm was a useful framework for developing an instrument for evaluating the UB strategies of PE teachers.We found that PE teachers applied UB strategies across three main task systems of the classroom ecology: management (maintaining order), instruction (promoting learning), and social relations (encouraging interactions), based on their personal and professional characteristics.Older PE teachers or teachers with higher levels of professional knowledge and experience differentiated their instruction to prevent UB.For this reason, they did not use classroom management strategies as much as young and early career PE teachers did.Female PE teachers were more willing than males to embrace differentiated teaching and social relations strategies.For all PE teachers, social interaction was an important element of classroom ecology.Given the theoretical paradigm adopted in this study, the above findings may have practical implications for the professional development of PE teachers.As evidence indicates a relationship between teachers' personal/professional characteristics and the way they deal with UB, the newly developed instrument could provide valuable information on teachers' decision to choose a strategy "as life in the gym unfolds."Furthermore, the stability or effectiveness of PE teachers' strategies could also be studied in relation to parameters such as their responses to student pressures, the flow of the lesson, the pace of learning activities, or their capacity to maintain student participation.According to Hastie (2016), these concepts are understudied fields of the ecology paradigm.

Limitations
This study is not without limitations.Initially, the validity and reliability evidence of the UBSQ-PE instrument are still to be accumulated for the field of PE teaching.Given the fact that the sample in this study came from only one geographical area, more UBSQ-PE validation tests are needed in the future.Second, relevant studies are needed to combine UBSQ-PE data with systematic observation data to better understand the effects of different UB strategies on PE teaching/learning goals and events related to the content covered in the classroom.Overall, an ecological examination of PE teachers' strategies could help us develop anthropological understandings of the ways they handle UBs across the interacting task systems, so that they get classroom work done.

Practical Implications
The different dimensions of Doyle's paradigm in relation to the four factors of the UBSQ-PE instrument presented in this study can help teachers and teacher educators reframe discussions on effective instruction within PE.Based on insights gained during the process of conducting the present study, our intention is not to suggest the UBSQ-PE as an instrument for recording PE teachers' technical skills in dealing with students' UB.From our own experience as PE teachers and teacher educators, we acknowledge that students need opportunities and time to respond to teachers' program of action.For this reason, teachers need to be able to acknowledge the strategies they adopt during instruction and adapt these in relation to the influences, properties, and realities of their classrooms.Our experience shows that some PE classes are over-instructed and under-practiced, in the sense that more time is spent on organizational issues, compared to instructional or social ones.Therefore, the question of what strategies work best, in our opinion, needs to focus on teachers' finding ways of crafting programs of action that help students navigate effectively within the dynamics of the classroom ecology.The enactment of instructional, managerial, and social strategies that are in the best interest to the classroom's "common good" may afford more learning to occur.Thus, with the employment of the UBSQinstrument we hope to strengthen the arguments for the wider adoption of complexity thinking approaches in PE, as approaches that understand classroom life as dynamic and unpredictable.Such understandings can afford PE teachers and researchers more opportunities to reflect why students behave as they do and what should be done from the teachers' part so that the accumulation of positive classroom experiences is achieved.

Figure 1 :
Figure 1: The Scree plot test for the four-factor solution.

Table 2 :
Factor loadings and communalities for the four-factor solution for the UBSQ-PEItemsFactor loadings Communalities Cronbach's alpha % Variance explained Differentiation of Instruction I adjust the lesson's challenge to handle UB 0lesson content and give students a variety of choices in pursuit of desired occurs, I deprive the student of a privilege or the right to participate in occurs, I deny the entire class the opportunity to participate in a students to interact within the class and create groups on their own 0 0.05, **p < 0.01, ***p < 0.001.For every bold value there are either *, **, *** assigned to the F index, suggesting their significance level.

Table 1 :
Demographic information of participants Interaction for Students Strategies (p < 0.05), and Interaction with Students Strategies (p < 0.01).Significant differences were also found between age groups for Differentiation of Instruction Strategies (p < 0.05), and Classroom Management Strategies (p < 0.001).The results further showed that PE teachers who had studied at the post-graduate level (p < 0.01) or attended professional development seminars (p < 0.001) were more willing to differentiate their instruction.Finally, a significant difference was found between PE teachers of different grade levels for Classroom Management Strategies (p < 0.01).The results of the univariate ANOVA analyses are shown in detail in Table 4.

Table 4 :
Results of univariate comparisons per UBSQ-PE factor