جهت گیری های علی ورزش، مقررات رفتاری برای ورزش و مراحل تغییر برای ورزش: بررسی روابط آنها

Abstract Objectives The objectives are to (a) explore the relationship between exercise causality orientations and stages of change and in doing so highlight any motivational changes that accompany movement through the stages, and (b) investigate the relative importance of exercise causality orientations and behavioural regulations in discriminating stage of change. Methods One hundred and one female (Mage=28.85±11.21) and 83 male (Mage=33.99±13.86) volunteers completed the Stage of Change for Exercise Ladder, the Exercise Causality Orientations Scale (ECOS) and the Behavioural Regulation in Exercise Questionnaire (BREQ).

Introduction The topic of motivation provides an avenue to understand behavioural choices and decision-making (Biddle & Mutrie, 2001) and to make judgements about the direction, intensity and persistence of behaviour (Kanfer, 1994). In the current climate, numbers participating in exercise are low and individuals seem to find it difficult to adhere to exercise beyond 6 months. As a consequence, research that expands our knowledge of the motivational processes that are used by individuals at different stages of behaviour change may be paramount to help us understand how the trends of inactivity and drop-out can be reversed. Self-determination theory (Deci & Ryan, 1985a) is regarded as an important theoretical perspective for the study of motivation within exercise psychology and by uniting it with the Transtheoretical model of behaviour change (Prochaska & DiClemente, 1984) the forces behind behaviour change can be further understood (Mullan & Markland, 1997). Self-determination theory distinguishes between intrinsic motivation (participation in an activity because of its inherent rewards of interest and enjoyment) and extrinsic motivation (participation in order to gain external rewards or to satisfy an external pressure) and is comprised of three mini-theories: cognitive evaluation theory, organismic integration theory and causality orientations theory. Cognitive evaluation theory details the social factors that promote and undermine the development of intrinsic motivation through their effects on self-determination, perceived competence and relatedness (the three innate needs). Organismic integration theory defines the self-determination continuum and the process by which non-self-determined extrinsic motives can be internalised to become more self-determined. Causality orientations theory describes the individual differences that are present in the interpretation of the characteristics of a specific situation. Using self-determination theory as the theoretical perspective, research has suggested that more extrinsic motives, such as losing weight and increasing fitness, drive the decision to adopt exercise (Ingledew, Markland, & Medley, 1997). However, it is believed that to gain any long-term consistency in exercise behaviour the development of intrinsic motivation and more self-determined motives are necessary (Biddle, 1999, Dishman, 1987, Frederick and Ryan, 1993, Mullan and Markland, 1997 and Wankel, 1993). According to cognitive evaluation theory, every situation or event can be interpreted as being informational, controlling or amotivational and the salience of these aspects (also known as the functional significance) will influence the motivational consequences for the individual. The informational aspect provides the individual with competence information and within a context of self-determination will promote intrinsic motivation. The controlling aspect will undermine self-determination and intrinsic motivation by imparting pressure to behave in a certain manner and to achieve extrinsic rewards. The amotivation aspect will also undermine intrinsic motivation by inducing feelings of incompetence. Importantly, it is not the objective characteristics of the situation that will influence self-determination, perceived competence and ultimately intrinsic motivation it is the individuals' perception of the direction and strength of each of the three aspects that is important. Causality orientations theory suggests that the way a situation is interpreted will differ from person to person. A specific situation can be interpreted as informational by one person and controlling or amotivational by another. It argues that there are personality-based orientations that predispose individuals to seek out an informational, controlling or impersonal aspect within each situation in order to regulate their behaviour. This interpretation will interact with the actual context and characteristics of the situation leading to the final interpretation. Deci and Ryan (1985b) describe three causality orientations: autonomy, control and impersonal. Underlying the autonomy orientation is the experience of choice. Individuals seek out the informational aspect to regulate their own chosen behaviour. They strive to be self-determining and seek out opportunities to achieve this. Behaviour is organised through the pursuit of self-selected goals and interests. The control orientation is underpinned by a search for control and pressure either imposed by others, within ourselves (e.g. guilt) or by the environment (reward contingencies). Individuals search out the controlling aspect (internal and/or external) to regulate their behaviour and they find themselves doing things because ‘they are told to’, ‘they should’, ‘they have to’ or ‘they must’. The sense of self-determination is missing. Finally, the impersonal orientation is based on the individuals' feeling that there is independence between behaviour and outcomes. They feel unable to regulate their behaviour to be able to achieve desired outcomes. In this case, the amotivating aspect is salient. Causality orientations theory suggests that each individual possesses a certain degree of each orientation. However, it is likely that one orientation will predominate in a given situation. It seems then that causality orientations theory can further understanding on how people interpret external events and thus complement the proposals of cognitive evaluation theory. Cognitive evaluation theory suggests that to experience human growth and development, the motivation for our behaviour should come from intrinsic sources. This occurs by satisfying our innate needs of self-determination, perceived competence and relatedness. This would suggest that there is an innate human tendency to want to experience intrinsic motivation. However, causality orientations theory suggests that for some individuals, the motivation for behaviour comes from external sources and they prefer to seek out situations of non-self-determined control to regulate their behaviour. It may be important to consider this situation when trying to understand the motivational processes behind the adoption and maintenance of exercise and indeed, when in a client consultation situation trying to foster behaviour change. Theoretically, high levels of the autonomy orientation would be the most conducive for the maintenance of exercise behaviour as self-determination is supported and intrinsic motivation should be cultivated. A high level of the control orientation may be detrimental to long term participation in exercise as self-determination is being constantly undermined and intrinsic motivation will not be developed. (Unless of course those individuals can always seek out the control and pressure they require to remain motivated.) Similarly, for those that are impersonally oriented it is unlikely that they will be exercising at all as they do not feel competent in being able to exercise and to achieve the desired outcomes. Alternatively, it may be that these motivational orientations are not fixed in personality as Deci and Ryan (1985b) suggest. It may be that the control orientation dominates in the early stages of exercise adoption but with regular participation in exercise the control orientation becomes less dominant and the autonomy orientation is strengthened. Due to the scarcity of literature in the area of causality orientations and exercise these proposals are highly speculative. Therefore, as a starting point, to show whether there is any substance to the proposals a cross-sectional study investigating the pattern of causality orientations within individuals at different stages of behaviour change seems appropriate. This methodology was used in early studies investigating the relationship between behaviour change and the processes of change (Gorely & Gordon, 1995) and self-efficacy (Marcus, Selby, Niaura, & Rossi, 1992). The Transtheoretical model (Prochaska & DiClemente, 1984) and in particular the temporal stages of change aspect of the model is commonly used to conceptualise the different stages in the exercise behaviour change process. This dynamic model of exercise behaviour change describes the five discrete stages that individuals pass through when making the decision to move from being sedentary to become active. It gives an indication of the individuals current exercise status along with their intention to maintain or change their behaviour (Prochaska & Velicer, 1997). Individuals in the precontemplation stage are sedentary and have no intention to become active within the next 6 months. Those in contemplation are still sedentary but are actively weighing up the pros and cons of becoming active and are intending to do so in the near future. It is not until the preparation stage that individuals have made an active commitment towards behaviour change and may have begun exercising albeit irregularly. Once in the action stage, individuals have fully engaged in behaviour change and are exercising regularly (typically classified as three exercise sessions per week for a minimum of 20 min). Individuals are classified as being in the maintenance stage when they have been exercising regularly for 6 months. The model conceptualises behaviour change through the stages as a cyclical process rather than a linear progression. Individuals can relapse out of any stage and regress back to a previous stage. Despite criticism on conceptual and empirical grounds (see Whitelaw, Baldwin, Bunton, & Flynn, 2000 for review) the model does provide an easy method on which to classify the current exercise behaviour and intention of individuals. It is important when classifying individuals into stage that a clear definition of what is meant by exercise is given (Richards Reed, Velicer, Prochaska, Rossi, & Marcus, 1997). Obviously, the stage classification of an individual may differ depending on whether health-related physical activity or structured exercise is being measured. It was the intention within this study to study planned and structured exercise/sport as defined by Loughlan and Mutrie (1995) as opposed to physical activity as defined by American College of Sports Medicine (1998). Mullan and Markland (1997) used the stages of change model to investigate the role of behavioural regulations (as described in the organismic integration theory) in the behaviour change process. Behavioural regulations are arranged along a continuum of self-determination. External regulation is when behaviour is undertaken due to external pressure from significant others and reflects non-self-determination. Behaviour undertaken due to internal pressure (such as guilt) is a reflection of introjected regulation (IJ) and is the next step along the self-determination continuum. Once an individual begins to value the outcomes/benefits of exercise and see exercise as being important to them they are described as being regulated by identified means. Identified regulation is characterized by high levels of self-determination. The next step on the continuum is to integrated regulation, which is indicative of behaviour being undertaken willingly and without a sense of pressure. Integrated regulation is analogous to intrinsic regulation, which also represents fully self-determined behaviour. However, the two are slightly different. Intrinsic regulation is characterized by the interest and enjoyment gained from the activity itself, whereas integrated regulation is still motivated by the importance attached to the outcome and is therefore still characterized as extrinsic motivation. Mullan and Markland (1997) showed that those individuals in action and maintenance were distinguished from those in prepreparation (a combination of those in precontemplation and contemplation) and preparation by their use of the more self-determined behavioural regulations (intrinsic and identified regulations). Furthermore, they showed that self-determined motivation (assessed by the Relative Autonomy Index (RAI), a unidimensional measure of the degree of self-determination) increased from prepreparation to preparation through to action and maintenance. They concluded that there was an apparent shift in motivational focus during the behaviour change process from being regulated by non-self-determined motivation to being regulated by self-determined forms of motivation. In uniting the causality orientations theory with the stages of change model, it can be ascertained whether this shift also occurs with respect to the predominance of the control, autonomy and impersonal orientations. Rose, Markland, and Parfitt (2001) developed and provided support for the factorial and convergent validity and reliability of the exercise causality orientation scale (ECOS). It measures the strength of each of the three orientations within the exercise environment. Within their study, relationships between the orientations and behavioural regulations for exercise were provided as a means of validating the scale. It was shown that the more self-determined behavioural regulations (intrinsic and identified regulations) were positively related to the autonomy orientation and negatively correlated with the impersonal orientation. External regulation (completely non-self-determined) showed a positive relationship with both the control and impersonal orientations. Therefore, having a predisposition to extract an informational or controlling significance from the exercise situation is associated with the behavioural regulations used to motivate exercise participation. Previous cross-sectional research has shown the link (but not direction of causality) between behavioural regulations in exercise and stages of change (Mullan & Markland, 1997) and the relationships between exercise causality orientations and behavioural regulations (Rose et al., 2001). In order to complete the conceptual link the association between exercise causality orientations and stage of change should be explored. This is the focus of the present study. It is proposed that those in the initial stages of behaviour change (precontemplation, contemplation and preparation) will exhibit high levels of the control and impersonal orientations and low levels of the autonomy orientation. While for those in the later stages of change (action and maintenance) the situation will be reversed and high levels of the autonomy orientation and low levels of the control and impersonal orientations will be shown. This study also seeks to investigate the relative importance of exercise causality orientations and behavioural regulations in discriminating stage of change in order to ascertain which asserts most influence. In accordance with previous research that has shown gender differences in motivational orientations (Fortier et al., 1995, Pelletier et al., 1995, Mullan and Markland, 1997 and Rose et al., 2001) all analysis will be conducted separately for males and females.

3. Results The results of the Mahalanobis distance analysis showed that two cases were potential outliers. The Mahalanobis distance values for these two cases were 17.18 and 17.13, respectively (χ2 critical value was 14.07). Therefore, these two cases (1 male and 1 female) were removed from the analysis. One female was also removed from the analysis due to having missing values. 3.1. Classification into stage of change Four (2.2%) individuals reported being in precontemplation (2 male, 2 female), 29 (15.8%) in contemplation (13 male, 16 female), 48 (26.4%) in preparation (23 male, 25 female), 27 (14.9%) in action (16 male, 11 female) and 73 (40.3%) in maintenance (29 male and 44 female). Due to the low number of participants in precontemplation, this stage was combined with contemplation and participants were described as being in the prepreparation stage. This procedure has been carried out previously by Mullan and Markland (1997). Male and female means and standard deviations for the ECOS and BREQ subscales at each stage of change are shown in Table 1. Table 1. Means and standard deviations (in parentheses) for the ECOS and BREQ subscales by stage of change Autonomy Control Impersonal External Introjected Identified Intrinsic Prepreparation Males 33.93 (7.15) 26.86 (9.94) 19.29 (6.53) 2.64 (1.98) 4.86 (2.18) 4.14 (4.38) 8.43 (5.24) Females 32.28 (6.12) 32.28 (7.49) 22.22 (5.58) 3.83 (2.94) 4.72 (2.32) 5.39 (4.17) 7.89 (4.64) Preparation Males 34.48 (5.01) 27.78 (8.17) 21.26 (5.94) 3.57 (2.04) 6.48 (1.16) 6.09 (2.45) 10.61 (2.44) Females 32.92 (6.86) 28.12 (10.36) 20.16 (4.29) 3.28 (2.70) 5.97 (1.88) 6.96 (2.96) 10.04 (3.62) Action Males 39.38 (5.43) 26.13 (7.68) 17.31 (5.57) 3.25 (2.38) 7.81 (1.42) 8.00 (2.68) 13.56 (2.00) Females 35.00 (4.11) 30.58 (6.35) 19.83 (7.38) 3.67 (2.35) 6.17 (1.20) 9.75 (1.96) 13.17 (1.59) Maintenance Males 41.41 (6.15) 24.31 (8.32) 15.79 (6.49) 1.93 (1.53) 7.55 (1.33) 9.59 (3.98) 14.00 (2.71) Females 39.36 (5.28) 26.32 (7.25) 17.70 (6.45) 2.78 (2.59) 7.80 (1.39) 10.96 (3.00) 14.24 (2.10) Table options 3.2. Discriminant function analysis Ecos. Results from the discriminant function analysis of the ECOS subscales showed that one function was significant for males (canonical r=0.52; Wilks' Λ=0.70, df=9, P<0.001) and accounted for 89.9% of the variance. As shown in Table 2, the structure correlations indicated that higher levels of the autonomy orientation (0.96) and a lack of the impersonal orientation (−0.52) defined the function. The function at the group centroid showed that males in action and maintenance scored positively on this function (0.28,0.67) while those in prepreparation and preparation scored negatively (−0.67, −0.60). The function correctly classified 51.8% of cases. Table 2. Discriminant function analysis using the ECOS or BREQ subscales separately to differentiate between the stages of change ECOS discriminating variable Structure matrix Standardised coefficients Males Females Males Females Autonomy 0.96 0.87 0.91 0.89 Control −0.21 −0.39 −0.18 −0.48 Impersonal −0.52 −0.45 −0.16 −0.09 Stage of change Value of the discriminant function at the group centroid Males Females Prepreparation −0.67 −0.85 Preparation −0.60 −0.47 Action 0.28 −0.22 Maintenance 0.67 0.67 BREQ discriminating variable Structure matrix Standardised coefficients Males Females (function 1) Males Females (function 1) Function 1 Function 2 Function 1 Function 2 Intrinsic regulation 0.77 0.25 0.81 0.19 −0.31 0.26 Identified regulation 0.68 0.02 0.70 0.54 −0.67 0.59 Introjected regulation 0.75 0.66 0.71 0.45 1.12 0.44 External regulation −0.25 0.60 −0.13 −0.59 0.59 −0.57 Stage of change Value of the discriminant function at the group centroid Males Females (function 1) Function 1 Function 2 Prepreparation −1.32 −0.56 −1.62 Preparation −0.62 0.37 −0.71 Action 0.37 0.62 0.47 Maintenance 0.97 −0.34 0.95 Table options One discriminant function was significant for females (canonical r=0.54; Wilks' Λ=0.68, df=9, P<0.001) and accounted for 91.7% of the variance. As shown in Table 2, the structure correlations indicated that the presence of the autonomy orientation (0.87) and a lack of the control orientation (−0.39) and impersonal orientation (−0.45) defined the function. However, the standardised coefficient showed that the impersonal construct was redundant (−0.09). The function at the group centroid showed that females in maintenance scored positively on this function (0.67) while those in prepreparation, preparation and action scored negatively (−0.85, −0.47, −0.22). The function correctly classified 50% of cases. BREQ. Results from the discriminant function analysis of the BREQ subscales showed that two functions were significant for males (canonical r=0.67; Wilks' Λ=0.45, df=12, P<0.001; canonical r=0.42; Wilks' Λ=0.81, df=6, P<0.02). The first function accounted for 78.1% of the variance and the second 20.8%. As shown in Table 2, the structure correlations indicate the first function was defined by intrinsic regulation (0.77), introjected regulation (0.75) and identified regulation (0.68). However, the standardised coefficients showed that intrinsic regulation was redundant (0.19). The second function was defined by external regulation (0.60) and introjected regulation (0.66). The function at the group centroid showed that males in action and maintenance scored positively on the first function (0.37,0.97) and those in prepreparation and preparation scored negatively (−1.32, −0.62). For the second function, males in preparation and action scored positively on the function (0.37,0.62) while those in prepreparation and maintenance scored negatively on the function. The functions correctly classified 56.6% of cases. One discriminant function was significant for females (canonical r=0.72; Wilks' Λ=0.46, df=12, P<0.001) and accounted for 95.7% of the variance. As shown in Table 2, the structure correlations indicated that function was defined by intrinsic regulation (0.81), introjected regulation (0.71) and identified regulation (0.70). However, the standardised correlations suggested that intrinsic regulation was redundant (0.26). The function at the group centroid showed that females in action and maintenance scored positively on this function (0.47,0.95) while those in prepreparation and preparation scored negatively (−1.62, −0.71). The function correctly classified 60.2% of cases. ECOS and BREQ combined. Table 3 shows that when all subscales from the ECOS and BREQ are included one discriminant function was significant for males (canonical r=0.68; Wilks' Λ=0.42, df=21, P<0.001) and accounted for 75.2% of the variance. The structure correlations indicated that intrinsic regulation (0.75); introjected regulation (0.73), identified regulation (0.66) and autonomy (0.63) defined the function. However, examination of the standardised coefficients showed that intrinsic regulation (0.06) and autonomy (0.23) were redundant as their coefficients did not reach the cut-off criteria of >0.30. The function at the group centroid showed that males in action and maintenance scored positively on this function (0.38,1.00) while those in prepreparation and preparation scored negatively (−1.33, −0.66). The function correctly classified 63.9% of cases. Table 3. Discriminant function analysis using the ECOS and BREQ subscales together to differentiate between the stages of change ECOS & BREQ discriminating variable Structure matrix Standardised coefficients Males Females Males Females Intrinsic regulation 0.75 0.73 0.06 0.27 Identified regulation 0.66 0.63 0.53 0.52 Introjected regulation 0.73 0.64 0.45 0.46 External regulation −0.25 −0.12 −0.49 −0.29 Autonomy 0.63 0.46 0.23 0.06 Control −0.11 −0.21 −0.14 −0.50 Impersonal −0.30 −0.26 0.02 −0.07 Stage of change Value of the discriminant function at the group centroid Males Females Prepreparation −1.33 −1.84 Preparation −0.66 −0.71 Action 0.38 0.39 Maintenance 1.00 1.06 Table options One discriminant function was significant for females (canonical r=0.76; Wilks' Λ=0.38, df=21, P 0.001) and accounted for 91.8% of the variance. As shown in Table 3, examination of the structure matrix showed the function was defined by intrinsic regulation (0.73), introjected regulation (0.64), identified regulation (0.63) and autonomy (0.46). The standardised coefficients indicated that intrinsic regulation (0.27) and autonomy (0.06) were redundant as their coefficients did not reach the >0.30 cut-off criteria. The function at the group centroid showed that those in action and maintenance scored positively on the function (0.39,1.06) and those in prepreparation and preparation scored negatively (−1.84, −0.71). The function correctly classified 66.3% of cases. 3.3. ECOS and stage of change The two factor (gender×stage) MANOVA conducted on the ECOS subscales revealed significant main effects for stage of change (F(9,416.32)=6.35, Wilks' Λ=0.73, P<0.001, η2=0.10) and for gender (F(3,171)=4.14, Wilks' Λ=0.93, P<0.01, η2=0.07). For autonomy, follow-up univariate ANOVAs revealed significant main effects for stage of change (F(3,173)=17.64, P<0.001, η2=0.23) and for gender (F(1,173)=6.62, P<0.02, η2=0.04). Tukey post-hoc tests showed that levels of autonomy were greater in action and maintenance compared to prepreparation and preparation. There were no significant differences between prepreparation and preparation or between action and maintenance. Overall, males showed greater levels of autonomy than the females. Within the control orientation, only a significant main effect for gender exists (F(1,173)=5.29, P<0.05, η2=0.03) which showed that females had greater levels of control than males. The analysis of the impersonal subscale revealed only a significant main effect for stage of change (F(3,173)=5.38, P<0.01, η2=0.09) which showed that levels of the impersonal orientation were significantly lower in maintenance than in prepreparation and preparation (see Fig. 1). Levels of the autonomy, control and impersonal orientations across the stages of ... Fig. 1. Levels of the autonomy, control and impersonal orientations across the stages of change. Figure options 3.4. BREQ and stage of change The two factor (gender×stage) MANOVA conducted on the BREQ subscales revealed significant main effects for stage of change (F(12,458.01)=12.29, Wilks' Λ=0.48, P<0.0001, η2=0.22) and for gender (F(4,173.00)=3.00, Wilks' Λ=0.94, P<0.02, η2=0.07). The follow-up univariate ANOVAs revealed a significant main effect for stage of change for external regulation (F(3,174)=2.71, P<0.05, η2=0.05), introjected regulation (F(3,174)=30.56, P<0.001, η2=0.35), identified regulation (F(3,174)=25.26, P<0.001, η2=0.30) and intrinsic regulation (F(3,174)=33.84, P<0.001, η2=0.37). There was also a gender main effect for identified regulation (F(1,174)=6.16, P<0.02, η2=0.03). Tukey post-hoc analysis showed that external regulation was greater at prepreparation than at maintenance and levels of introjected regulation were greater in action compared to preparation and in preparation compared to prepreparation. There were no differences between action and maintenance. With respect to identified regulation, there were significant differences between prepreparation and preparation compared to action and maintenance. Overall, levels of identified regulation were greater in females than males. Finally, levels of intrinsic regulation were greater in action compared to preparation and in preparation compared to prepreparation. There were no significant differences between action and maintenance. 3.5. Relationships between the ECOS and BREQ subscales Table 4 shows the results of the Pearson's correlations between stage of change and the ECOS and BREQ subscales. Results showed that the autonomy subscale was significantly positively correlated with introjected regulation (0.49), identified regulation (0.38), intrinsic regulation (0.69) and RAI (0.65). The control subscale was significantly positively correlated with external regulation (0.47) and introjected regulation (0.15). The impersonal orientation was significantly positively correlated with external regulation (0.29) and significantly negatively correlated with introjected regulation (−0.21) and intrinsic regulation (−0.32). Table 4. Pearson's correlations between the subscales of the ECOS and BREQ External Introjected Identified Intrinsic Autonomy −0.15* 0.49** 0.38** 0.69** Control 0.47** 0.15* 0.11 0.09 Impersonal 0.29** −0.21** −0.07 −0.32** * significant at P<0.05; ** significant at P<0.01