Regular engagement in physical activity is associated with a variety of psychological, social and physical benefits for children and youth [1, 2]. Unfortunately, previous research has consistently demonstrated that physical activity decreases during the transition from childhood to adolescence, with girls in particular showing significant declines in participation [3–5]. There is no widely accepted explanation for this phenomenon; however, several authors have suggested that negative experiences in, and perceptions of, school-based physical education (PE) class may be an important factor in the decline in physical activity, particularly among girls [6]. This decline occurs at the same time PE becomes an optional part of the curriculum in most school systems in North America [7].

School-based PE programs are often touted as an effective means of positively affecting health and physical activity behaviours during childhood [2] and later life [8]. School programs reach the vast majority of children [9], and can affect physical activity both directly and indirectly. PE classes can provide children with activity that directly fosters motor skill development and fitness, and can also stimulate positive perceptions of physical activity - thereby influencing motivation to engage in physical activity [6–10]. Unfortunately, declining physical activity participation during adolescence suggests that there are deficiencies in the current PE programs being implemented. Moreover, it remains unclear whether students' experiences, or perceptions of those experiences, in school PE truly facilitate physical activity over the long term.

One important determinant of physical activity is simply youths' perceived enjoyment of such activity. Data from several studies show that motivation to participate in physical activity in children is influenced by perceptions of physical activity as fun, interesting, and challenging [11, 12]. Enjoyment of physical activity is positively correlated with physical activity participation levels for children and adolescents [13–16]. Enjoyment of PE class specifically has also been identified as a predictor of physical activity [17]. Sallis and colleagues [14] investigated 22 potential correlates of physical activity in school children and found that students' enjoyment of PE class was a robust predictor of physical activity at both younger (grades 4-6) and older (grades 7-12) ages. Given its strong link to physical activity in children and youth, further investigation of enjoyment within the context of PE classes is warranted.

Numerous theories of motivated behavior that have been applied to understanding and predicting physical activity participation (e.g., social cognitive theory, competence motivation theory, self-determination theory) highlight positive affect (e.g., fun, enjoyment) as a proximal motivator of physical activity. Positive affect, in turn, is influenced by more distal perceptions such as self-efficacy or competence, outcome expectations, autonomy, and relatedness. In the literature on children's physical activity, perceptions of competence, autonomy, and perceived success have been shown to predict physical activity enjoyment, as well as to reduce negative perceptions, such as boredom [18, 19]. In the school environment, self-efficacy with respect to leisure-time physical activity was the strongest predictor of PE class enjoyment in a sample of middle school girls [6]. Together, these findings suggest that perceived competence may be an important factor affecting students' enjoyment of their experiences in PE.

The objective of the present study was to investigate associations between perceived athletic competence, PE class enjoyment, and gender in a sample of school-aged children across grades 4 to 6. We conducted a 3-year prospective study, allowing us to extend existing research in this field, which, with one exception [9], has relied exclusively on cross-sectional data [6, 15, 19–21]. Importantly, we were able to examine trends in perceptions of athletic competence and PE enjoyment over time. Based on cognitive theories of motivation and previous research in children's physical activity [e.g., 6-12], we hypothesized that perceived athletic competence would be a significant predictor of children's PE enjoyment.

Although the major objective of the study was to investigate the relationship between perceived competence and enjoyment of PE, we also set out to extend previous work investigating gender differences as they relate to PE enjoyment. In previous studies, girls have been found to enjoy PE less than boys [9, 21, 22]. It is possible that determinants of PE enjoyment vary by gender; however, our understanding of why girls differ from boys in their enjoyment of PE remains limited, primarily because studies have either simply established the presence of gender differences with no further elaboration or test of possible mechanisms [9], or have limited their sample to girls [6]. For example, the study by Barr-Anderson and colleagues [6] identified a number of individual and contextual factors associated with PE enjoyment; but, by excluding boys, the design precluded an examination of the sources of gender differences. In line with previous research, we hypothesized that girls would report lower PE enjoyment than boys. As an exploratory step, gender was examined as a potential moderator of the relationship between perceived athletic competence and PE enjoyment over time.

The average age of children at baseline was 9.6 years (SD = 0.34). Slightly more than half of the students were male (50.9%). Table 1 contains descriptive data for enjoyment of PE and all other time-varying measures included in the analysis. Model 1 (see Table 2) shows the effect of time and gender on reported enjoyment of PE. The presence of a significant effect for gender by time (b = 6.50 × 10^-4, p < 0.001) indicates that there is a divergence of boys' and girls' enjoyment scores. In order to facilitate the interpretation of this result, we have graphed the association between gender, time, and PE using the equation from model 1, illustrated in Figure 1. The Y-axis of this chart, representing predicted values of enjoyment, shows the range from 1 SD below to 1 SD above the total mean. For boys, there is a slight increase (0.18 SDs) in reported enjoyment from grade 4 to grade 6. For girls, however, reported enjoyment of PE slightly decreased (0.08 SDs) over that same period.

	Model 1 Estimate (SE)	Model 2 Estimate (SE)	Model 3 Estimate (SE)
Intercept	10.7 (0.05)***	8.41 (0.15)***	9.02 (0.2)***
Time (days)	-1.6 × 10-4 (8.5 × 10-5)	-2.5 × 10-4 (8.2 × 10-5)**	-3.0 × 10-3 (3.5 × 10-4)***
Age (baseline)	0.03 (0.08)	0.03 (0.06)	0.04 (0.06)
Gender
Male	0.2 (0.07)**	0.01 (0.06)	-0.05 (0.25)
Female	-	-	-
Male by Time	5.8 × 10-4 (1.2 × 10-4)***	5.7 × 10-4 (1.1 × 10-4)***	2.7 × 10-3 (5.4 × 10-4)***
BMI		-0.03 (0.01)***	-0.03 (0.01)***
Parental Education
High School or Less		-0.16 (0.09)	-0.16 (0.09)
College/Technical School		-0.001 (0.06)	-5.0 × 10-5 (5.5 × 10-2)
Undergraduate Degree		-0.23 (0.09)**	-0.22 (0.09)**
Professional/Graduate Degree		-0.09 (0.11)	-0.08 (0.1)
Education missing		-	-
Organized Activity Participation		1.5 × 10-2 (3.6 × 10-3)***	1.3 × 10-2 (3.5 × 10-3)***
Perceived Athletic Competence (PAC)		0.15 (4.6 × 10-3)***	0.12 (0.01)***
PAC by Time			1.5 × 10-4 (1.9 × 10-5)***
PAC by Male			4.9 × 10-3 (1.3 × 10-2)
PAC by Time by Male			1.2 × 10-4 (2.8 × 10-5)***

** p < .01; *** p < .001

In Model 2, the effect of PAC on reported enjoyment was assessed in concert with other potential covariates. Consistent with the hypothesis, there was a significant effect for PAC, with higher scores associated with greater PE enjoyment, net of the other factors (see Model 2 in Table 2). The coefficient for the interaction between time and gender remained largely unchanged after adjusting for BMI, participation in organized sports, parental education, and PAC. After adjusting for age, gender and time, however, these variables did exert independent effects on enjoyment. Higher BMIs were associated with lower levels of reported PE enjoyment (b = -0.03, p < 0.001), whereas greater participation in sports and organized activities was associated with higher levels of reported enjoyment (b = 0.02, p < 0.001). There was also a weak effect of parental education, with children of parents who hold undergraduate university degrees reporting lower levels of enjoyment of PE than other groups (b = -0.21, p < 0.001).

As shown in Table 2, the introduction of PAC and the other covariates to the equation in Model 2 had very little effect on the coefficient for time by gender (b = 6.30 × 10^-4, p < 0.001 compared with b = 6.50 × 10^-4, p < 0.001 from Model 1). This suggests that the effect of PAC on enjoyment over time may differ by gender. In order to test for this, a set of 2-way and 3-way interactions with PAC, time, and gender were added to the model. Complete results from this analysis are shown in Model 3 in Table 2. Most interestingly, a significant 3-way interaction emerged between gender, time, and PAC (b = -1.00 × 10^-4, p < 0.001). Again, to facilitate the interpretation of this result, we used the full equation in Model 3 to graph the association between PAC, time, and perceived enjoyment for both boys and girls (see Figure 2). In Figure 2, "lower" PAC is represented by a score 1 SD below the mean, and "higher" PAC by a score 1 SD above the mean. The results indicate that gender has no effect on enjoyment of PE for children who perceived themselves as having higher competence with regard to sports and athletics. In fact, over time, there is a slight trend in both boys and girls in this group toward increasing levels of enjoyment. Gender difference did, however, emerge among children with lower perceived competence. Specifically, although children with lower perceived competence persistently had lower levels of enjoyment of PE compared to those with higher perceived competence, the enjoyment levels among boys remained relatively stable over time. Among girls, however, lower perceived competence was marked by decreases in enjoyment of PE over time.

The purpose of the present study was to investigate the association between perceived athletic competence and enjoyment of PE. Drawing on cognitive theories of motivation [24, 26], we hypothesized that perceived athletic competence would be positively associated with PE class enjoyment. Generally, the results of this study supported this hypothesis, but the relationship between perceived competence and PE enjoyment was qualified by a 3-way interaction with gender and time. Consistent with previous cross-sectional research [6, 15, 18–20, 26], we found that girls reported lower levels of PE enjoyment than boys. Importantly, however, our results indicate that these differences increased over time. Although previous research has found significant associations between PAC and enjoyment of PE [19], this was the first study to test the association using longitudinal data.

The current findings support previous calls for the targeting of perceived competence in intervention work aimed at increasing PE enjoyment among school-aged children [6, 20]. Although the hypothesized PAC and PE enjoyment relationship was supported in this study, it is important to acknowledge that the relationship was complicated by an interaction with gender and time. Specifically, the interaction showed that children of both genders who had higher levels of perceived competence reported higher and increasing levels of PE enjoyment. However, differences emerged among children with lower levels of perceived competence, where boys reported low but stable levels of PE enjoyment over time, while girls began with lower levels of PE enjoyment that continued to decline further over time. These findings offer evidence that gender differences in PE enjoyment are effectively non-existent among children with high levels of perceived competence, and suggest greater efforts should be made to understand why some children have lower perceptions of athletic competence. Improving children's views of their physical abilities may be an important goal, particularly among girls with low perceived competence. The interaction between gender, competence and enjoyment should be a key consideration in planning the content and delivery of PE.

These findings support perceived competence as a focus for interventions designed to increase enjoyment of PE. It is important, however, to understand that competence is just one of many determinants of enjoyment of and participation in physical activity. Cognitive theories of intrinsic motivation, for example, emphasize not only competence, but also perceptions of choice or autonomy [24, 27]--constructs that were not measured in this study. A PE program that not only targets positive perceptions of competence, but that also provides a variety of different activities--allowing participants to choose those most of interest to them--is likely to produce greater levels of intrinsic motivation and enjoyment [9, 28].

Recent work has also emphasized the importance of understanding sport ability beliefs [29], which underlie feelings of competence because they reflect children's perceptions of the modifiability of their performance [30]. If children perceive their physical abilities to be fixed and not amenable to improvement though practice, it is unlikely that perceptions of enjoyment will be increased in PE settings, where the emphasis is often on raw skill acquisition. One way to improve engagement and perceived competence would be to expose children to a wider spectrum of activities. This would create more opportunities for individuals to discover activities for which they have a particular aptitude or that they find appealing, interesting, and challenging. In fact, this is consistent with the stated policy of PE in the jurisdiction in which this study was conducted. The Ontario Ministry of Education, the governmental body responsible for education curriculums for the schools included in this study, outlines that a major learning goal to be helping students achieve a greater understanding of achieving lifelong health through the promotion of physical and health literacy. Physical literacy, defined as the ability to move, with confidence, across a wide array of physical activities that benefit the whole person (mind and body), being particularly relevant [31]. A caveat, however, was that schools (and teachers) in our study were ultimately responsible for designing and implementing PE classes; which could have varied considerably across school districts depending on available resources (e.g., some schools had state of the art gymnasiums, while other schools did not even have on-site facilities). The idea is to expose children to a greater range of activities, enabling all children to find confidence in different movements suited to individual athletic abilities. Therefore, it was not particularly surprising that higher levels of organized physical activity, as measured by the PQ, were associated with greater reported enjoyment of PE. Overall, focusing on developing competence may not be enough, and understanding and changing negative core beliefs with regard to abilities may also need to be part of the strategy. For example, focusing on individual goal setting, striving for self-set or negotiated standards of performance, and providing greater autonomy and choices in physical activity may be necessary.

Although not a focus of the present study, many of the control variables included in the models were also significantly related to perceived enjoyment of PE. In particular, higher BMI scores were associated with lower level of enjoyment throughout this time period. BMI is of particular interest, as other work has shown that conditions that influence motoric competence are also associated with lower perceived enjoyment of PE [32, 33]. Unpacking the temporal associations between these factors and PE enjoyment remains to be done.

The present study has a number of limitations. Although our analysis took correlation within and variation among schools into account with the use of random effects, we were unable to also consider class or school-level factors associated with the delivery of PE (e.g., teacher competence) or the perceived climate in PE class (e.g., supportive, competitive, etc.). While our focus was on individual-level factors, the inclusion of contextual effects would provide a more complete picture; this remains an area for future research. In particular, intervention studies that manipulate teaching environment could shed light on the importance of context in perceived enjoyment.

Enjoyment of PE makes continued engagement in physical activity more likely; if children do not enjoy PE, they may be less likely to lead an active lifestyle, and to realize the associated health benefits. The results of our study show that girls, in particular, are less likely to enjoy physical education when compared to boys and that this gap in enjoyment grows over time. This may be due to the fact that girls have, on average, lower perceived physical ability. Of course, other factors, not measured in this study (e.g., maturation) may also contribute to observed gender differences. This is the first study to show that the association between perceived enjoyment of PE and PAC is contingent upon gender as well as time. If enjoyment of PE is to be increased, physical educators should attend not only to the development of skills, but also to increasing feelings of competence, particularly competence in PA that the child is able to perceive with relevance to activities inside and outside of the school. In this sense, the mind, as well as the body, must be the focus of instruction in PE class, and should be part of the broader curriculum in physical education for all children; which may be especially the case for young girls.