Abstract
The aim of this study was to assess the reliability and validity of the Amotivation Inventory in Physical Education (AI-PE). In addition, the study sought to identify the relationships between students’ amotivation, physical self-concept, and teacher ratings of National Curriculum attainment levels in PE. Students (N = 510) from a secondary school in South-East England participated in the study by completing a questionnaire during their scheduled PE lesson. Confirmatory factor analysis provided support for the factorial validity of the four-dimensional model of amotivation. Hierarchical multiple regression analyses were conducted on a sub-sample of students with moderate to high levels of amotivation in PE (n = 164). The results showed that both physical self-concept (negatively) and attainment (positively) were predicted by students’ deficient ability beliefs. Overall, the findings support the multidimensionality of amotivation and its association with important outcomes in PE.
Physical education (PE) can play an important part in a child’s perception of sport, exercise and physical activity. In England, PE is compulsory from 5 to 16 years of age, and is regarded as extremely important in schools as it provides children with the knowledge and skills to sustain physical activity in the future (Sas-Nowosielski, 2008). Obesity levels are rising in the UK, particularly among children, and therefore schools need to accept that they play a large part in a child’s health (Department of Health, 2010). The Childhood Obesity National Support Team (CO NST) is concerned that even if schools meet the government target of two hours of PE per week, pupils are not sufficiently physically active in those two hours (Department of Health UK, 2010).
Participation levels in PE also decline as children progress through secondary school (DfE: PE and Sport Survey, 2009/10; Vallerand and Bissonnette, 1992). Evidence has shown that in the school year 2009/10, 87% of pupils in years 7–9 participated in 120 minutes of curriculum PE each week, 64% in years 10–11 and only 23% in years 12–13 (DfE: PE and Sport Survey, 2009/10). By the time students reach year 11 (16 years of age), interest for some in physical activity is non-existent, they do not want to participate, and are ‘amotivated’ (Ntoumanis et al., 2004). It is important to note that participating in PE at school is not an option. It is compulsory for every child to participate in the lesson unless they have a legitimate reason not to do so. Classes may contain a significant minority of amotivated students; however, it may not be easy to identify these students by observation alone. Understanding the psychological processes that students experience whilst participating in PE is an interesting and complex area that needs to be researched further. Therefore the purpose of the present investigation was to examine amotivation in secondary school PE and to identify important correlates of this understudied construct.
Amotivation
Self-determination theory has been a prominent framework for understanding human behaviour (Deci and Ryan, 1985, 2000) and has underpinned recent research in physical activity and youth (Standage and Treasure, 2002; Standage et al., 2003, 2005). Deci and Ryan (1985, 2000) theorised that motivation is a multidimensional construct and encompasses three global types of motivation – intrinsic motivation, extrinsic motivation and amotivation, that all function on a self-determination continuum.
Amotivation refers to the absence of motivation. Individuals do not want to participate in an activity and make no effort to engage in the behaviour (Deci and Ryan, 2000). In PE, amotivated individuals may lack enthusiasm in lessons, which might signify boredom. They might show ‘avoidance’ behaviours such as continually forgetting to bring their PE kit as an excuse to not participate, and their attendance may be low (Ntoumanis et al., 2004). They may also show signs of incompetence (Bandura, 1986) and learned helplessness (Ntoumanis et al., 2004) and a failure to value the activity (Ryan, 1995). Performance goals and social approval goals have also been linked closely with amotivation (Papaiannou et al., 2007).
The majority of research to date has focused on amotivation as a unidimensional construct. However, Legault et al. (2006: p580) proposed that ‘amotivation is itself an entity, a complex and multifaceted process, which is not so much an absence as a broad effect of unmet needs’. Pelletier and colleagues (Pelletier et al., 1999) proposed a model of amotivation and investigated reasons for individuals’ lack of motivation toward environmental protective behaviours. Pelletier and co-workers referred to amotivation as ‘global helplessness beliefs’ (p2485) but argued that an individual is amotivated for more specific reasons. Furthermore, they identified three additional dimensions of amotivation: strategy beliefs, capacity beliefs and effort beliefs. Legault et al. (2006) further developed the multidimensional model of amotivation in an academic school setting, developing a taxonomy of amotivation which entailed four dimensions: deficient ability beliefs, deficient effort beliefs, insufficient task values, and unappealing task characteristics.
The first dimension of amotivation, ‘deficient ability beliefs’, encompasses the idea that people have an expectation of their ability to accomplish certain tasks. In the present study, deficient ability beliefs represent students’ views that they are unable to achieve or ‘do well’ in PE. This dimension can be likened to self-efficacy, as both constructs can be defined as the ‘beliefs in one’s capabilities to organise and execute the causes of action required to produce given attainments’ (Bandura, 1997: p3). If an individual possesses deficient ability beliefs, they may know that certain behaviours will achieve a specific outcome, but will believe they do not have the ability to successfully carry out the required behaviour (Shen et al., 2010).
The second dimension of amotivation is ‘deficient effort beliefs’ (Pelletier et al., 1999). This dimension looks at whether a student is willing to invest effort into the activity. Effort represents a more easily visible behavioural component of amotivation, for example jogging as opposed to sprinting, and standing still during game situations. The third dimension is ‘insufficient value placed on a task’, or in other words, whether students value the activity they are doing in PE. Do they understand why they are practising a certain skill for example? Do they value the importance of having to participate in physical activity? Ntoumanis (2001) found that amotivation was evident when children felt they were ‘wasting their time’ in PE lessons. The last dimension of amotivation is ‘unappealing characteristics of the task’. The task itself must be interesting for the student so that they are more likely to engage in the activity. If they deem the task as boring and tedious they are more likely to be amotivated (Ntoumanis 2001; Ntoumanis et al., 2004). It has also been found that tasks which are too challenging for students lead to amotivation as they lack the belief that they can accomplish the challenge, and do not want to take risks for fear of failing (Standage et al., 2005).
In the United States, Shen and colleagues (2010) initially assessed amotivation in PE by adapting and validating the Academic Amotivation Inventory (AMI). The AMI was devised to measure the multidimensional construct of academic amotivation. Shen et al. (2010) provided preliminary evidence that the Amotivation Inventory in Physical Education (AI-PE) was an appropriate measurement tool to assess the four dimensions of amotivation.
Correlates of amotivation
Physical self-concept
The self can be viewed as multidimensional and hierarchical. A multidimensional model of the self describes the self at global and domain levels. The domain level includes the physical, social and academic self and is related strongly to motivation (Crocker et al., 2008: p217). Physical self-concept is an important mediator in physical activity. The self-concept is a person’s evaluation of his or her own qualities in general or specific domains (Marsh, 1987), and is helpful for understanding an individual’s outlook toward physical activity. For example, experiences of low physical attractiveness and social physique anxiety influence students’ motivation to engage in exercise (Crocker et al., 2008).
Adolescents may exercise more in order to achieve weight loss (Sabiston et al., 2007), and therefore may have high extrinsic motivation. However, some children may avoid physical activity if possible because of the anxiety associated with their negative physical appearance. This can be termed as ‘social physique anxiety’ that occurs as a result of ‘interpersonal evaluation of one’s physique’ (Hart et al., 1989: p96). Amotivation in compulsory PE classes is likely to be linked with lower physical self-concept.
Attainment
There have been no studies to date that have directly examined the relationship between amotivation and students’ attainment in PE following the National Curriculum guidelines in England. There has, however, been research conducted on the motivational profiles and their relationship with performance in PE. Boiche et al., (2008) conducted research into the relationship between different motivational profiles and performance in gymnastics. Findings suggested that the more self-determined profile, that is, higher intrinsic motivation and higher extrinsic motivation (introjected regulation), achieved higher grades than moderate profiles or less self-determined profiles including amotivation. Unfortunately these findings are not able to be generalised to students in the UK, as students’ performance in PE is not solely graded on their ability to execute certain skills. Students’ grades encompass the execution, selection and application of skills and evaluation of performance.
Sun and Chen (2010) have recently examined self-determination components and learning in middle school PE and found that amotivation negatively related to knowledge improvement, but no relationship was found with motor skill achievement. In educational contexts, self-determined motivational regulations have been associated with a range of outcomes. Intrinsic motivation has been found to contribute to better academic performance (Deci and Ryan, 1985). However, less research attention has been directed to the role of amotivation in determining educational outcomes.
The reasons why children are amotivated in PE classes need to be understood so that effective interventions can be developed. A deeper understanding of amotivation in PE is required in order to ultimately increase participation levels in sport and exercise at school and beyond. There is a dearth of literature on amotivation in youth physical activity. Therefore, the first aim of the present study was to validate the AI-PE that has been adapted from Legault and colleagues (2006) and initially tested by Shen and colleagues (2010), and to confirm whether the four dimensions of amotivation can be supported in PE. Students’ self-reported emotional and behavioural engagement and disaffection were also measured to examine the concurrent validity of the AI-PE. The second aim of this study was to test the predictive utility of the four amotivation beliefs in understanding students’ physical self-concept and attainment in PE.
Method
Participants
Participants (N = 510; 217 boys, 293 girls) were recruited from a mixed comprehensive secondary school in the southeast region of England. Ages ranged from 11 to 16 years (16–17 years n = 72; 15 years n = 86; 14 years n = 146; 13 years n = 101; and 12 years n = 105; boys’ mean age = 13.6 years; girls’ mean age = 14.0 years). The majority of students were of middle socioeconomic status from a range of ethnic backgrounds including Caucasian, African American, Nigerian, Asian and Greek.
Students participated in PE classes twice per week from 11–14 years of age (key stage 3). Between 15 and 16 years (key stage 4), students participated in one PE class per week. Classes were of the same sex in most cases, and were taught by a teacher of the same sex as the students. Two classes were mixed sex and were taught by a male teacher. PE classes were taught in groups of 25 or more and were compulsory for students to attend.
Measures
Amotivation in physical education
The AI-PE (Shen et al., 2010) was used to measure students’ amotivation. The AI-PE consists of 16 items measuring the four dimensions of amotivation: deficient ability beliefs (e.g. ‘I don’t have what it takes to do well in PE’); ‘Deficient effort beliefs (e.g. ‘I’m not energetic enough’); Unappealing characteristics of the task (e.g. ‘My PE lessons are not stimulating’); and Insufficient value of the task (e.g. ‘PE is not valuable to me’). At the start of the questionnaire, students were asked to rate on a Likert scale from 1–7 (1 = never; 7 = always), how often they experienced a lack of motivation to do PE lessons (cf. Legault et al., 2006). They were then asked to rate on a Likert scale from 1–7 (1 = does not correspond at all; 7 = corresponds exactly) each statement that corresponded with their reasons for not wanting to do PE. Preliminary evidence for the reliability of the AI-PE has been provided by Shen et al. (2010).
Physical self-concept
Six items were taken from the ‘global physical’ scale on the original ‘Physical Self Description Questionnaire’ (PSDQ) (Marsh, 1994). Participants reported on a 6-point Likert scale ranging from 1–6 (1= not at all true; 6 = very true). Example items were, ‘Physically I am happy with myself’ and ‘Physically I feel good about myself’. Evidence for the reliability and validity of the PSDQ has been provided by Marsh (1994).
Attainment in PE
To determine the PE teachers' ratings of attainment, a sub-sample of students (n = 230; 16–17 years n = 23; 15 years n = 18; 14 years n = 69; 13 years n = 60; and 12 years n = 60) was assessed at the end of the school year and awarded a National Curriculum (NC) level ranging from 3a–8a based on their overall performance, effort and behaviour, covering a range of different activities and skills. The National Curriculum for Physical Education (NCPE) recommends that students are assessed in PE using ‘levels of attainment’. These levels range from 1–8, with level 8 being the highest. Secondary schools in the UK can opt to add sub-levels to the assessment criteria to monitor students’ progress more easily by recommending that students improve their attainment grade by two sub-levels per academic year. For example, each level (e.g. level 3) consists of three sub-levels (3a, 3b and 3c). A student in year 7 (aged 11–12 years) should be achieving an average level 3a–4c in the majority of activities. By year 8 (aged 12–13 years), a student who achieved 3a in year 7 should be achieving a 4b by the end of year 8. It is, therefore, expected that attainment will show a high correlation with age. NC sub-levels were recoded using a Likert-type scale from 1–12, 1 equalling the lowest level 3c and 12 corresponding to the highest sub-level 6a, as this was the highest sub-level that was achieved in the current sample.
Engagement and disaffection
A 20-item questionnaire, adapted from a self-report measure in education (Skinner et al., 2009; Wellborn, 1991), was devised to assess students’ behavioural and emotional engagement and disaffection in their PE lesson. Five items were used to measure behavioural engagement, focusing on the students’ effort, attention and persistence whilst participating in PE (e.g. ‘I try hard to do well in PE’). Behavioural disaffection was measured using five items that focused on students’ lack of effort and withdrawal from activities during a PE lesson (e.g. ‘When I am in PE, I think about other things’). Emotional engagement was measured using five items that tapped into students’ emotions to indicate whether they were motivationally involved in the PE lesson (e.g. ‘When I am in PE I feel good’). Emotional disaffection was also measured using five items to focus on the emotions that may indicate students’ motivational withdrawal from activities in their PE lesson (e.g. ‘When we work on something in PE I feel bored’). Students were asked to rate on a Likert scale from 1–4 (1 = not at all true; 4 = very true) what statement best represented their opinions.
Procedures
Having attained ethical approval for the study and informed consent from the participants, the questionnaires were administered at the beginning of a PE lesson by the PE teacher. Students were informed that their responses would be confidential and that they would have the right to withdraw at any time. Participants were asked not to confer with other participants and to complete the questionnaire honestly. Each participant took approximately 10 minutes to complete the questionnaire.
Results
Factorial validity of AI-PE
Confirmatory factor analysis (CFA) using EQS software (version 6.1; Bentler, 2003) was employed to determine the factor structure of the scales utilised. The comparative fit index (CFI) was chosen as an incremental fit index, and the standardised root mean square residual (SRMR) and the root mean square error of approximation (RMSEA) represented absolute fit indices. Hu and Bentler (1999) proposed that acceptable fit of a hypothesised model to the data is indicated when the CFI is close to .95, the SRMR is close to .08, and the RMSEA is close to .06. In this study, the AI-PE was conceptualised to consist of four correlated factors (Legault et al., 2006; Shen et al., 2010). These factors were deficient ability beliefs, unappealing task characteristics, insufficient value of the task, and deficient effort beliefs. Examination of the fit indices indicated that the data adequately fit the four-factor model (CFI = .95; NNFI (Non-normed fit index) = .94, SRMSR = .05, RMSEA = .06).
Descriptive statistics, scale reliabilities and correlations
The means, standard deviations, and Cronbach’s alpha coefficients were calculated for each subscale and are presented in Table 1. All subscales demonstrated good internal reliability (α >.82) (Kline, 1986). Coefficients for the four dimensions of amotivation ranged from .82 to.92. The four dimensions of the AI-PE each had mean scores below 3.0.
Descriptive statistics for the dimensions of amotivation, physical self-concept, emotional and behavioural engagement and disaffection, and attainment in physical education (Total Sample: N = 510).
Note. AB:Ability beliefs; EB: Effort beliefs; TV: Task value; TC: Task characteristics; PSC: Physical self-concept.
To explore the relationship between amotivation scores and the other variables, correlations were assessed among the dimensions of amotivation (using the total sample), emotional and behavioural engagement and disaffection, physical self-concept, and attainment (see Table 2). As hypothesised, all four dimensions of amotivation were associated positively with emotional and behavioural disaffection (correlation coefficients ranging from .36 to .59) and negatively with emotional and behavioural engagement (correlation coefficients ranging from -.29 to -.64) demonstrating the concurrent validity of the AI-PE. Physical self-concept showed a negative relationship with all four dimensions of amotivation, with correlation coefficients ranging from -.15 to -.37. Attainment was found to positively correlate with the four amotivation dimensions. Correlation coefficients ranged from .15 to .22.
Correlations among amotivation, emotional and behavioural engagement and disaffection, physical self-concept and attainment scores (Total Sample: N = 510).
Note. ** p < .01.
AB: Ability beliefs; EB: Effort beliefs; TV: Task value; TC: Task characteristics; PSC: Physical self-concept.
Hierarchical multiple regression analyses
A series of hierarchical regression analyses were conducted on a sub-sample of participants (n = 164) who scored at or above the scale midpoint (4 or above) on the question ‘How often do you experience a lack of motivation to do PE lessons’ (see Table 3). This sample was deemed to consist of students with moderate to high levels of amotivation (Legault et al., 2006).
Hierarchical multiple regressions predicting physical self-concept and attainment (Amotivated Sample: n = 164).
Note. *** p < .001, ** p < .01, * p < .05.
AB: Ability Beliefs; EB: Effort Beliefs; TV: Task Value; TC: Task Characteristics.
Predicting physical self-concept from age, gender, and amotivation dimensions
The results showed that the amotivation model was significant in explaining 9% of the variance of physical self-concept scores once age and gender were taken into account. Deficient ability beliefs was the only dimension of amotivation to significantly predict physical self-concept (β = −.28, p < .001) (see Table 3).
Predicting attainment from age, gender, and amotivation dimensions
The amotivation dimensions accounted for an increase in explained variance of 5.2% for attainment after controlling for age and gender. Age (β = .67, p < .001) and deficient ability beliefs (β = .21, p < .05) were the only significant predictors of attainment scores (see Table 3). The final model accounted for 44% of variance in students’ attainment.
Discussion
Factorial validity of the AI-PE
Amotivation can be described as the feelings of alienation and helplessness in PE (Ntoumanis et al., 2004). The first aim of the present study was to validate the existing AI-PE (Shen et al., 2010) and to provide empirical support for the four dimensions of amotivation. The findings supported the conceptualisation of amotivation being a multifaceted construct (Shen et al., 2010), identifying multiple sources that lead to students’ feelings of alienation. The results of the CFA supported the factorial validity of the four dimensions of amotivation suggesting that the AI-PE is a reliable and valid measurement tool. Results also supported the concurrent validity of the AI-PE, whereby emotional and behavioural engagement and disaffection showed significant correlations with all four dimensions of amotivation in expected directions.
The second aim of the study was to explore the relationships between physical self-concept and attainment with amotivation and whether the four amotivation dimensions significantly predicted physical self-concept and attainment scores after controlling for age and gender.
Physical self-concept
Results from the present study indicated that certain dimensions of amotivation negatively predicted physical self-concept. This finding is in line with the initial hypothesis, and is supported by Deci and Ryan (2000) who mention the ‘struggle for body control’ (referring to social physique anxiety) as being an outcome of deficiencies in self-determination. Therefore, the low feelings towards the physical self in PE may be a consequence of amotivation, in particular ‘deficiency in ability beliefs’, which the present results show to be highly significant in predicting physical self-concept.
In the present study, ‘deficiency in ability beliefs’ represented whether a student felt they were unable to ‘do well’ in PE, and therefore a student who believed they did not have the ability to achieve in their PE lesson would be likely to have a low physical self-concept, and subsequently will have a negative evaluation of themselves physically in different contexts and situations. A consequence of this may be a reluctance to get involved in an activity. Thus, the relationship between amotivation and sense of physical self is likely to be reciprocal over time. Future work using a longitudinal design should attempt to identify the relative strength of these bidirectional associations.
A student’s perception of their ability may be influenced by the ability of others in their PE class, such that physical self-competence decreases if other students are more able physically (Trautwein et al., 2008). Teachers therefore need to be aware that mixed-ability classes may serve to lower students’ sense of physical competence.
Attainment
In the present study, all four dimensions of amotivation positively correlated with teacher-rated attainment. However, ‘deficiency in ability beliefs’ was the only dimension that positively predicted attainment in the regression analysis. This implies that students who perceive deficiencies in their ability are, in fact, receiving positive judgements of attainment from their PE teacher. The secondary school that was used in the present study divided students into mixed-ability groups for their PE lessons. Some of the students who were awarded a high attainment level from their teacher may self-report low motivation due to not feeling as able as others in the class. Some students may feel that performing better than their classmates is an improbable task. The student, however, may still meet the criterion assessment criteria and therefore achieve a high NC level. In addition, students’ perception of their ability in class may be inconsistent with their teachers’ judgements (Furrer and Skinner, 2003).
There may be several moderators of the relationship between amotivation and outcomes in PE. For example, achievement goal theory (Nicholls, 1989) interprets what drives student-motivated learning behaviour in the academic setting and may help further explain why deficient ability beliefs account for significant variance in physical self-concept. Research has found that students in secondary schools (Walling and Duda, 1995) can be identified as task-and ego-oriented. Walling and Duda (1995) found that students with a high ego-orientation avoided difficult learning tasks and attributed their successes or failures to their ability.
In addition, Sarrazin and colleagues (2002) focused on the effects of task-involving versus ego-involving motivational climate created by the coaches. A task-involving climate emphasised learning and team co-operation in contrast to an ego-involving climate that focused on winning and rivalry between team members (Ames, 1992). The study found that athletes who dropped out from training perceived their coach to be less task involving and more ego involving. They also addressed how the motivational climate related to the three psychological needs of autonomy, competence and relatedness, which need to be supported in the environment for an individual to display self-determined motivation (Deci and Ryan, 2000). Sarrazin et al. (2002) found that the more the athletes perceived the climate as ego involving, the less competent they felt. This may be due to an ego-oriented climate failing to focus on athletes’ skills and progress, thus reducing the chance for them to feel competent. It is therefore important in future research to explore the psychosocial antecedents of amotivation in more detail by focusing on the motivational climate fostered by the PE teacher, and students’ perceptions of the support provided by teachers.
Self-theories of ability may provide another interpretation of students’ perceptions of ability. Dweck (1999) stated that incremental beliefs (the attribute is viewed as changeable and malleable) and entity beliefs (the attribute is viewed as fixed and stable) determine the adoption of achievement goals. Children who hold incremental beliefs are more likely to adopt a task goal orientation (Hong et al., 1999), and those children who hold entity beliefs are more likely to adopt an ego-oriented achievement goal (Mueller and Dweck, 1998).
These findings were supported by Wang et al. (2002), who found that task orientation was associated with positive forms of self-determination and a belief that ability could be changed through learning. Biddle et al. (2003) looked at the relationship between amotivation, goal orientations and sport ability, and found that entity beliefs directly predicted self-reported amotivation but no relationship was found with ego goals and amotivation. In conclusion, if students believe their ability is not stable they then will appreciate that if they invest effort into the activity they will improve and subsequently will enjoy PE more. If they believe their ability is unchangeable then they may become frustrated and amotivated.
The finding that ‘deficiency in ability beliefs’ positively predicted attainment may be explained by the assessment criteria that were used by the PE teachers in the study. The attainment levels that the teachers awarded students represented an average score determined across a number of different activities over the academic year. Therefore, attainment scores may not have produced a reliable correlation with amotivation, as students reported their amotivation levels at one point during the summer term, rather than at different points across a number of activities over the year. In future studies, teachers should be asked to award the current level at which the student is performing, at the time of completing the amotivation questionnaire.
Limitations, future research and conclusion
Limitations of the present study need to be recognised when drawing conclusions. First, the study used a cross-sectional design and future longitudinal research needs to be conducted to examine changes in amotivation over time with concomitant changes in outcomes in order to determine the direction and strength of influence. It is possible that changes in physical self-concept and attainment impact on students’ amotivation and the bidirectional relationships need to be explored further using a larger amotivated sample.
Second, although the NCPE (2000) requires teachers to adopt a conscious focus on assessment and to make judgements on students’ overall performance, there is no statutory requirement for teachers to collect evidence of pupils’ attainment in any particular way. Carroll (1994) indicated that the teacher must be systematic in their assessment, be clearer about the criteria being used, and observe children carefully. Current findings raise the question as to whether teachers are all assessing using the same criteria and/or are carefully observing the students’ performance.
In conclusion, the present study advances the current literature in a number of ways. To date, there has been one study that has validated the AI-PE (Shen et al., 2010), but the authors did not further explore the amotivation dimensions as possible predictors of key variables such as physical self-concept and teacher ratings of attainment. The findings, therefore, facilitate our understanding of the importance of amotivation as a multidimensional construct and the effects of sub-dimensions of amotivation on potentially important outcomes. Overall, the findings point towards the importance of the deficient ability beliefs component of amotivation, but larger studies of amotivated students are required to ascertain the significance of ‘deficient effort beliefs, insufficient task value, and unappealing task characteristics’. The results also highlight a potential disparity between a teacher’s views of students’ attainment and students’ perception of their motivation for the subject.