Journal of Teaching in Physical Education

Endorsed by the Curriculum and Instruction Academy of

the NASPE and the AIESEPwww.JTPE-Journal.com

ARTICLE

68

Journal of Teaching in Physical Education, 2014, 33, 68-92 http://dx.doi.org/10.1123/jtpe.2013-0069 © 2014 Human Kinetics, Inc.

Pan is with the Graduate Institute of Physical Education, National Taiwan Sport University, Taojuan, Taiwan.

Relationships Among Teachers’ Self-Efficacy and Students’ Motivation,

Atmosphere, and Satisfaction in Physical Education

Yi-Hsiang PanNational Taiwan Sport University

The purpose of this study was to confirm the relationships among teachers’ self-efficacy, and students’ learning motivation, learning atmosphere, and learning satisfaction in senior high school physical education (PE). A sample of 462 PE teachers and 2681 students was drawn using stratified random sampling and cluster sampling from high schools in Taiwan. The research instruments were the Teachers’ Self-Efficacy Scale, and students’ Learning Motivation Scale, Learning Atmosphere Scale, and Learning Satisfaction Scale, which were designed by the researchers based on theories and existing instruments. Structural equation model-ing was used to evaluate the fit of the hypothetical model. The results revealed that the model had acceptable fit. It was concluded that physical education teachers’ self-efficacy affected students’ learning motivation, learning atmosphere, and learn-ing satisfaction; teachers’ self-efficacy also indirectly and positively influenced learning satisfaction, mediated by learning motivation and learning atmosphere.

Keywords: teachers’ competence, teachers’ beliefs, teaching strategy

“Efficacy beliefs influence how people feel, think, motivate themselves, and behave” (Bandura, 1993, p. 118). These beliefs also influence how much effort people put into a task, how long they persist in the face of obstacles, how resilient they are in dealing with failures, and how much stress or depression they experience in coping with demanding situations (Bandura, 1997). Self-efficacy is a cognitive process in which people construct beliefs about their competence to perform at a given level of attainment (Tschannen-Moran, Woolfolk Hoy, & Hoy, 1998). Self-efficacy is a social cognitive theory, and the theoretical understanding of teachers’ self-efficacy assumes a set of beliefs about teachers’ capacity to have a positive influence on students’ education (Henson, 2002). Therefore, it is important to conduct research regarding teachers’ self-efficacy.

Teachers’ Efficacy and Students’ Satisfaction 69

First, it is necessary to identify the key factors that could influence school effectiveness. Hsieh (1997) established evaluation indicators and a weight system for school effectiveness, and evaluation results showed that the educational input comprised four second-order indicators: goal development and planning (20%), expenditure and equipment (17%), teachers’ competence (43%), and environmental quality (20%). Teachers’ competence was the most important indicator because its weight (43%) was the highest among all the indicators. This suggests that teachers’ self-efficacy, the belief held by teachers regarding their own professional compe-tence, may be able to influence their professional praxis. Therefore, teachers’ self-efficacy also influences their professional behavior and students’ performance in physical education classes. It is important to understand how teachers’ self-efficacy can influence students’ learning processes.

In literature reviews related to teachers’ self-efficacy, most studies adopted the teacher’s viewpoint. For example, Gorozidis and Papaioannou (2011) con-ducted a study of self-efficacy in physical education teachers in a Greek junior high school after new curriculum guidelines were issued. They found that highly self-efficacious teachers had a positive attitude toward the physical education curriculum and intended to continue their current efforts in the future. Pan, Chou, Hsu, Li and Hu (2013) also found that teachers’ self-efficacy could have a direct influence on their commitment to teach health and physical education curricula in elementary schools in Taiwan. However, a more objective understanding of the role of teacher’s self-efficacy in teaching and learning may come when we consider the students’ viewpoint.

Furthermore, teachers have always been concerned about students’ learning satisfaction with the physical education curriculum. Students’ learning experience can have a positive influence on their attitude and behavior in physical education classes. Currently, some researchers are interested in discovering which latent variables affect students’ learning satisfaction (Chen & Stotlar, 2012; Huang, 2007; Lund, 2010; Tessier, Sarrazin, & Ntoumanis, 2010). Because teachers guide student learning, teachers with higher self-efficacy may be able to use a variety of methods to stimulate students’ motivation and create a quality learning atmosphere, both of which could promote students’ learning satisfaction. Ha, Johns, and Shiu (2003) examined students’ perspective in the design and implementation of the physical education curriculum. They found that one third of the participants would not choose physical education if the program were offered as an optional subject, and that more than 50% of the students from grades 10 and 11 disliked the content of most physical education programs.

Given these findings, it is worthwhile to consider why students do not choose physical education and may dislike much of the content of physical education pro-grams. If we provided a quality physical education program to improve students’ learning satisfaction, this would also improve students’ willingness to actively participate in physical education. Positive and satisfying sports participation experiences not only foster students’ physical and mental development but also significantly motivate the establishment of long-term exercise habits (Chen & Stotlar, 2012).

Lund (2010) indicated that physical education curriculum should meet pro-grammatic goals, and participating students should also have some degree of satisfaction. When creating physical education curricula, it is important to realize

70 Pan

that not everyone will enjoy everything that is offered. Students feel satisfaction when they participate in physical education activities during school. These physical education activities can greatly enhance students’ learning experience and posi-tively influence their physical, mental, social and interpersonal wellbeing. Huang (2007) indicated that teaching styles could have a direct and significant impact on learning satisfaction, but that learning motivation and learning strategy do not have a direct and significant impact on learning satisfaction. Chen and Stotlar (2012) indicated that some other factors might influence the relationship between learn-ing motivation and learning satisfaction. These factors include quality of teaching, peer relationships, quality of administrative procedures, and facilities. From a practical perspective, students’ satisfaction in physical education classes could be increased if physical education activities were designed that could enhance their social relationships and create an enjoyable atmosphere. Moreover, based on self-determination theory (Deci & Ryan, 2002), student choice has also been shown to motivate students in physical education classes.

The present study uses an integrated approach to examine the causal relation-ships among teachers’ self-efficacy, students’ learning motivation, learning atmo-sphere and learning satisfaction, to understand the effects of latent variables on students’ learning satisfaction in senior high school physical education programs in Taiwan. The physical education teachers’ self-efficacy is the independent vari-able, with students’ learning motivation and learning atmosphere as mediators, and students’ learning satisfaction as the dependent variable.

Relationship Between Teachers’ Self-Efficacy and Students’ Learning Motivation

Roberts (2001) indicated that “motivation refers to dispositions, social variables, and/or cognitions that come into play when a person undertakes a task at which he or she is evaluated, or enters into competition with others, or attempts to attain some standard of excellence” (p.6). Motivation is a construct developed to explain the degree of attraction the learner has to a particular behavior or learning task. Currently, the most popular theories to explain motivation may be behaviorism, social learning theory, self-determination theory, achievement goal theory, and interest theory (Rink, 2010). Learning motivation is an important issue in the field of education, and therefore, it is worthwhile to explore what factors influence levels of students’ learning motivation. To achieve teaching objectives, a teacher should implement teaching activities, support students’ motivation, and maintain a good learning atmosphere. Caprara, Barbaranelli, Borgogni, and Steca (2003) found that teachers’ self-efficacy had a strong influence on learning motivation. Rutten, Boen, and Seghers (2012) showed that physical education teachers could promote autonomous motivation by satisfying the need for competence and the need for autonomy postulated by self-determination theory. Tessier et al. (2010) also indicated that teachers’ teaching style could have a positive effect on their

Teachers’ Efficacy and Students’ Satisfaction 71

students’ learning motivation. Therefore, the teacher’s professional competence may directly influence the students’ learning motivation. Teachers with higher self-efficacy are good at motivating students by using appropriate teaching strate-gies (Tschannen-Moran & Woolfolk, 2001). This means that teachers with higher self-efficacy would have a more positive influence on students’ learning motivation, which is what we hypothesized in this study.

Relationship Between Teachers’ Self-Efficacy and Classroom Learning Atmosphere

Management is one of the most difficult functions of teaching for novice teachers. Teachers must elicit student cooperation to support learning engagement and must develop and maintain a learning environment that supports learning (Rink, 2010). Classroom learning atmosphere refers to a learning environment where learners and teachers interact with each other and use a variety of information resources in their learning activities. Evidence indicates that an atmosphere supportive of learning becomes a more prominent influence in that learning achievement and that student attitudes toward learning are closely linked to classroom atmosphere. Bay, Bündoğdu, and Kaya (2010) indicated that a good learning environment could create opportunities that encourage learners to express themselves, and promote a positive social relationship with others during the learning process. Teachers with higher self-efficacy use appropriate teaching strategies to create a quality learning atmosphere. Thus, we hypothesized that teachers’ self-efficacy would influence the learning atmosphere in class.

Relationship Between Teachers’ Self-Efficacy and Students’ Learning Satisfaction

Efficacious teachers employ effective problem solving skills, develop strategies to be more effective teachers, manage their emotions well, and persist in the face of failure (Bandura, 1997). Teachers with high self-efficacy inspire students to work harder to gain greater satisfaction and achievement (Caprara, Barbaranelli, Steca, & Malone, 2006). A number of studies have shown how teachers’ self-efficacy has a positive effect on students’ learning achievement (Chacon, 2005; Martínez-López, Sánchéz, Alvarez, & Cruz, 2010; Shidler, 2009). Tessier et al. (2010) also indicated that teachers’ teaching style could have a positive effect on students’ psychological satisfaction in physical education. Learning achievement is the key reason for students’ learning satisfaction. If students are successful in learning per-formance, they will also be more satisfied with the learning process. Teachers with a stronger self-efficacy possess higher teaching competence, maintain harmonious relationships with colleagues and parents, and are more likely to positively affect students’ performance (Chacon, 2005). Therefore, we hypothesized that teachers’ self-efficacy would influence students’ learning satisfaction.

72 Pan

Relationship Between Students’ Learning Motivation and Learning Atmosphere

Teachers want students to be intrinsically motivated to participate in activities, so they have positive learning experiences. Students with higher motivation try to work hard in their program, which in turn enhances the learning atmosphere. As a result, the level of students’ learning motivation could influence the classroom learning atmosphere. Beggs, Elkins, and Powers (2005) indicated that an individual’s motiva-tion for participating in sports activities may be related to that person’s enjoyment of those activities. Enjoyment is a key factor in the development of the learning atmosphere. Therefore, we hypothesized that students’ learning motivation would influence students’ learning atmosphere.

Relationship Between Students’ Learning Atmosphere and Learning Satisfaction

Teachers with higher self-efficacy could arrange an environment conducive for learning and for developing student-appropriate behavior and content engagement. The teacher’s good management skills are essential for effective teaching in physical education. Learning satisfaction is related to learning atmosphere; an atmosphere with teacher control allows students to experience that learning is serious and an autonomous atmosphere allows students to enjoy the learning process. Previous studies showed that learning atmosphere could influence students’ learning satis-faction (Chan & Chan, 2011; Wang, 2011). For example, Chan and Chan (2011) examined the relationship between classroom atmosphere and learning satisfaction in physical education classes. They found that teachers’ support, expectancy for success, and group cohesion of classroom were the important factors that could influence students’ learning satisfaction in the physical education class. Classroom atmosphere was found to strongly predict learning satisfaction, explaining 57% of the total variance. When students have more teacher support, higher expectations for physical education class, aggressive classroom climate, and stronger confidence in skill performance, they could have higher learning satisfaction in physical educa-tion classes. Therefore, we hypothesized that classroom learning atmosphere would influence students’ learning satisfaction.

Examination of the Overall ModelThe present study sought to confirm the relationships among teachers’ self-efficacy, and students’ learning motivation, learning atmosphere, and learning satisfaction in senior high school physical education classes. The above literature review suggests that teachers’ self-efficacy directly and indirectly influences students’ satisfaction. Teachers’ self-efficacy has a positive effect on learning motivation, learning atmosphere, and learning satisfaction, respectively. On the other hand, the relationship between teachers’ self-efficacy and students’ learning satisfaction is mediated by learning motivation and learning atmosphere. The relationships are presented in Figure 1. In the hypothesized model, teachers’ self-efficacy is an

73

Fig

ure

1 —

Hyp

othe

size

d m

odel

of

the

fact

ors

affe

ctin

g st

uden

ts’

lear

ning

sat

isfa

ctio

n in

phy

sica

l edu

catio

n.

74 Pan

independent variable, and students’ learning motivation, learning atmosphere, and learning satisfaction are dependent variables. The fit of the hypothesized model is evaluated using structural equation modeling (SEM). There are five hypotheses formulated as follows:

Hypothesis 1 (H1): The learning satisfaction model has an acceptable goodness-of-fit.

Hypothesis 2 (H2): Teachers’ self-efficacy has a positive effect on learning motivation, learning atmosphere, and learning satisfaction, respectively.

Hypothesis 3 (H3): Learning motivation has a positive effect on learning atmosphere.

Hypothesis 4 (H4): Learning atmosphere has a positive effect on learning satisfaction.

Hypothesis 5 (H5): The relationship between teachers’ self-efficacy and stu-dents’ learning satisfaction is mediated by learning motivation and learning atmosphere.

MethodParticipants and Settings

Participants were high school physical education teachers and their students throughout Taiwan. According to the data provided by the Ministry of Education, there are 314 public and private high schools in Taiwan. We randomly selected 105 high schools, and administered questionnaires to 2–8 teachers at each school, depending on the size of the school. A total of 630 questionnaires were administered to teachers, 462 completed questionnaires were returned for an acceptable return rate of 73%. Furthermore, six students (those with a student ID number that was 5, 10, 15, 20, 25, and 30) were selected from each class taught by the 630 sampled physical education teachers. Of the 3,780 questionnaires administered to students, 2,681were returned, yielding acceptable return rate of 71%. The final sample com-prised 462 high school physical education teachers and 2,681 students. The age of teachers ranged from 24 to 65 years (male = 259, female = 203, M = 41.30, SD = 3.74); 48.6% had only a bachelor’s degree while 51.4% had master’s or doctorate degrees. The age of the students ranged from 15 to 18 years (male = 1,394, female = 1287, M = 16.43, SD = 0.92), with 40% being in 10th grade, 36% being in 11th grade, and 24% being in 12th grade. The students participated in physical education classes for 45 min twice a week during the school year. The teachers taught 20–22 physical education classes a week; no teacher worked in more than one school. The curriculum taught in the schools is a multiactivity model.

Measures

Teachers’ Self-Efficacy Scale for High School Physical Education Teachers (TSES-HSPET). The TSES-HSPET was designed by Pan (2012) based on Bandura’s (1997) self-efficacy theory and Gibson and Dambo’s (1984) Teacher Efficacy Questionnaire (TEQ). The TSES-HSPET has two dimensions, each

Teachers’ Efficacy and Students’ Satisfaction 75

with seven items: personal teaching efficacy (PTE) and general teaching efficacy (GTE). PTE is the level of confidence a teacher has in his or her own teaching skills and his or her awareness of students’ improvements. An example of a PTE item is “I am good at physical education curriculum design.” PTE is similar to the efficacy expectation of Bandura’s (1997) notion of self-efficacy—the indi-vidual’s conviction that he or she can orchestrate the necessary actions to perform a given task. GTE refers to the teacher’s expectations of how the students can improve their own competence and attitude in physical activity under external environmental influences. An example of a GTE item is “Even though there are not enough sports equipment and facilities, I am able to apply various methods to develop my students’ skills.” GTE is similar to the outcome expectancy of Bandura’s (1997) notion of self-efficacy, which essentially reflects the degree to which teachers believe that the environment can be controlled, that is, the extent to which the student can be taught given such factors as family background, IQ, and school conditions. Each item was rated using a six-point Likert-Like scale that ranged from 6 (strongly agree) to 1 (strongly disagree).

Based on evaluation criteria of fit indices (Kline, 1998), confirmatory factor analysis showed that the TSES-HSPET had acceptable fit (χ2 (75) = 167.94, p < .05, RMSEA = 0.07, GFI = 0.92, CFI = 0.98) Composite reliability (CR) values were .78 and .90 for PTE and GTE, respectively. Convergent validity was consid-ered acceptable based on the average variance extracted (AVE) values of .43 and .53. Discriminant validity was also acceptable, with a confidence interval of this parameter’s correlation coefficient being between .28 and .40. Therefore, it was concluded that the TSES-HSPET is both reliable and valid in the current sample, including two factors and 14 items.

Learning Motivation Scale in Physical Education (LMSPE). This scale was based on Bandura’s conceptualization of self-efficacy in social cognitive theory and the Motivated Strategies for Learning Questionnaire (MSLQ, Pintrich, Smith, Garcia, & McKeachie, 1993). The LMSPE was modified on the basis of previous scales (Huang, 2007; Pintrich et al., 1993) and focused on motivation factors of high school students in physical education classes. This learning motivation scale has four factors: value, expectation, affective, and self-efficacy. Expectancy component refers to students’ beliefs regarding whether they can accomplish a task in physical education. An example of an expectancy item is “I wish to improve my move-ment skill in physical education courses”. Value component refers to students’ viewpoints concerning why they engage in a physical education curriculum. An example of a value item is ‘Promoting physical and mental health is a critical goal of physical education courses’. Affective component refers to students’ feelings regarding what is required to exhibit a positive attitude in physical education. An example of an affective item is “I like to participate in various learning activities in physical education courses”. Self-efficacy component refers to individual beliefs concerning the ability to successfully perform physical activities. An example of a self-efficacy item is “I can learn skills well in physical education courses”. Each item was rated using a six-point Likert-Like scale that ranged from 6 (strongly agree) to 1 (strongly disagree).

The examination of LMSPE’S reliability and validity was carried out in two stages. At stage 1 of the study, 314 students completed the scales. The exploratory

76 Pan

factor analysis that was used in the first stage established the construct validity of this instrument (74.50% of total variance explained). The Cronbach’s alpha coef-ficient was 0.93 for the total scale, and 0.89, 0.84, 0.80, and 0.91 for the subscales, indicating satisfactory internal consistency. At stage 2, 327 students completed the LMSPE, and confirmatory factor analysis was used to test construct validity. The result showed acceptable fit (χ2 (98) = 298, p < .05; RMSEA = 0.07, GFI = 0.90, CFI = 0.97). The average variance extracted of the latent variables were 0.64, 0.58, 0.73, and 0.72, with composite reliability of 0.88, 0.84, 0.86, and 0.91, respectively. We conclude that the LMSPE is psychometrically sound with adequate reliability and validity in the current sample, including four factors and 16 items.

Learning Atmosphere Scale in Physical Education (LASPE). This study used the LASPE designed by Pan and Cheen (2010), who modified it based on previous scales (Aldridge, Fraser, & Ntuli, 2009) so as to focus on learning atmosphere factors for high school students. This LASPE has four factors with 17 items, with components that include group cohesion, learning enjoyment, teacher support, and peer relationships. Group cohesion refers to students’ behavior regarding what they can cooperate to accomplish tasks in physical education. An example of a group cohesion item is “My classmates work hard together to accomplish common goals”. Learning enjoyment refers to students’ feelings to experience happy learning in physical education. An example of a learning enjoyment item is “Physical education is an interesting course”. Teacher support refers to teachers’ behavior to help students’ learning in physical education courses. An example of a teacher support item is “The physical education teacher is always concerned about the learning performance of the students”. Peer relationship refers to students’ feeling to develop positive relationships in physical education. An example of a peer relationship item is “Classmates like to share learning experiences in physical education classes”. Each item was rated using a six-point Likert-Like scale that ranged from 6 (strongly agree) to 1 (strongly disagree).

In the first stage, exploratory factor analysis showed that this scale had accept-able construct validity (67.89% of total variance explained). The Cronbach’s alpha coefficient was 0.93 for the total scale, and 0.82, 0.86, 0.91, and 0.73 for the sub-scales, indicating satisfactory internal consistency. In the second stage, confirmatory factor analysis was used to test construct validity. The results showed acceptable fit (χ2 (98) =232, p < .05; RMSEA = 0.06, GFI = 0.92, CFI = 0.98). Average vari-ance extracted of latent variables were 0.55, 0.68, 0.74, and 0.52, and composite reliability scores were 0.83, 0.89, 0.92, 0.81, respectively. We thus can conclude that the LASPE is a psychometrically sound instrument with adequate validity and reliability in the current sample, including four factors and 17 items.

Learning Satisfaction Scale in Physical Education (LSSPE). The LSSPE was modified based on previous scales (Tsai, 2006; Xiang, Lee, & Solmon, 1997) and focused on satisfaction factors of high school students in physical education classes. The LASPE has two factors including teaching implementation and learning effect. Teaching implementation refers to the teachers’ teaching skills and attitudes. An example of a teaching implementation item is “The physical education teacher is adept at using various strategies to enhance student learning effects”. Learning effect refers to students’ learning performance. An example

Teachers’ Efficacy and Students’ Satisfaction 77

of a learning effect item is “I feel satisfied regarding the learning effect of motor skills in physical education”. Each item was rated using a six-point Likert-Like scale that ranged from 6 (strongly agree) to 1 (strongly disagree).

In stage 1, we sampled 314 students who completed the questionnaire. The exploratory factor analysis was used in the first stage, and the result was 73.78% of total variance explained, supporting the construct validity of the instrument for the current sample. The Cronbach’s alpha coefficient of internal consistency was 0.93 for the sum scale, and subscales were 0.90 and 0.93. In stage 2, 327 students completed the LSSPE and confirmatory factor analysis was used to test construct validity. The result showed acceptable fit (χ2 (43) = 110, p < .05; RMSEA = 0.07, GFI = 0.94, CFI = 0.99). The average variance extracted from the latent variables were 0.79 and 0.65, with composite reliability of 0.81 and 0.70. It is concluded that the LSSPE is a good measurement instrument with adequate validity and reliability, including two factors and 11 items.

Procedure

The study was approved by the institutional review board of the National Taiwan Sport University. The researcher communicated with the heads of school PE offices before conducting the questionnaire survey. The researcher mailed the question-naires to the heads and requested their cooperation and assistance in conducting the questionnaire survey for teachers and students. A consent form accompanied the questionnaires, explaining the nature of the study and assuring the teachers and students that their answers would be kept confidential, and that answers were neither right nor wrong and should only reflect their own opinions. The partici-pants completed the questionnaire in an average of 15 minutes, and put them in a cardboard box upon completion. Similarly to a former study (Standage, Duda, & Ntoumanis, 2006), data were collected toward the end of the academic term so as to give students enough time to form perceptions of the class. The data were col-lected for eight months by the academic team. To this end, collecting data at the end of the term allowed teachers and students to provide more accurate responses.

Data Analyses

Structural equation modeling (SEM) tested the hypothesized model of a linear structural relationship among teachers’ self-efficacy, learning motivation, learning atmosphere, and learning satisfaction. For the latent variables, the current study created composite scores and used scores of 12 subscales as indicator variables.

ResultsTable 1 provides the means, standard deviations, correlation coefficients and skew-ness and kurtosis values for all observed variables. The skewness (-0.30 to -0.96) and kurtosis (-0.56–1.39) of the 12 observed variables were in the acceptable range of ± 2 SD. Bollen and Long (1993) suggested that the skewness and kurtosis indices of all the items were lower than 2.0, which indicates univariate normal distribu-tion of the data. Data were collected using random sampling and were normally

78

Tab

le 1

D

escr

ipti

ve S

tati

stic

s an

d C

orr

elat

ion

s A

mo

ng

Var

iab

les

(N =

462

)

12

34

56

78

910

1112

1. P

TE

-

2. G

TE

.77*

-

3. V

A.4

2*.4

0*-

4. E

X.4

6*.4

3*.5

7*-

5. S

E.3

0*.3

8*.4

6*.4

8*-

6. A

F.4

8*.4

9*.6

3*.6

3*.6

8*-

7. G

C.3

7*.3

9*.2

5*.2

4*.2

7*.3

1*-

8. L

E.5

6*.5

2*.6

6*.5

9*.5

5*.7

5*.3

9*-

9. T

S.6

7*.6

1*.4

1*.4

3*.2

8*.4

6*.3

7*.5

9*-

10. P

R.4

8*.4

6*.3

1*.3

2*.2

8*.3

2*.5

1*.4

2*.4

5*-

11. I

T.7

5*.6

6*.4

5*.4

3*.2

9*.4

6*.3

5*.5

9*.7

4*.4

7*-

12. E

L.6

4*.6

3*.5

4*.6

1*.5

2*.6

8*.3

5*.7

3*.5

7*.4

7*.6

7*-

Mea

n4.

864.

775.

355.

144.

445.

024.

615.

024.

994.

865.

035.

02

SD.4

7.6

5.6

6.7

1.9

6.7

9.8

3.8

0.8

5.7

2.7

6.7

3

Skew

ness

-.51

-.30

-.96

-.58

-.56

-.72

-.82

-.72

-.88

-.68

-.62

-.38

Kut

osis

.13

-.19

.68

-.15

.30

.19

1.39

.49

1.17

1.31

.39

-.56

* p

< .0

5

Not

es. 1

.PT

E =

per

sona

l te

achi

ng e

ffica

cy;

2.G

TE

= g

ener

al t

each

ing

effic

acy;

3.V

A =

val

ue;

4. E

X =

exp

ecta

tion;

5. S

E =

sel

f-ef

ficac

y; 6

. AF

= a

ffec

tive;

7. G

C =

gr

oup

cohe

sion

; 8.

LE

= l

earn

ing

enjo

ymen

t; 9.

TS

= t

each

ers’

sup

port

; 10

.PR

= p

eer

rela

tions

hip;

11.

TI

= t

each

ing

impl

emen

tatio

n; 1

2. L

E =

lea

rnin

g ef

fect

; α=

C

ronb

ach’

s al

pha

coef

ficie

nt; S

D =

sta

ndar

d de

viat

ion

Teachers’ Efficacy and Students’ Satisfaction 79

distributed; therefore, the maximum likelihood method was appropriate for testing this model. The three categories used as evaluation criteria were preliminary fit, overall model fit, and fit of the internal structure of the model (Bagozzi & Yi, 1988).

Examination of Model Fit

Preliminary Model Fit Criteria. The parameter estimates of the model for physi-cal education teachers’ self-efficacy and students’ learning satisfaction are shown in Table 2. The analysis showed that none of the standardized error variance estimates were negative. These standardized error variances were between 0.13 and 0.62 (p < .05). None of the related absolute values among the parameters was close to 1. The factor loadings for each observed variable were between 0.48 and 0.92. Only one factor loading (0.48) did not reach 0.50, but was considered close enough to meet the preliminary fit criteria.

Overall Model Fit. Evaluation criteria for the overall model fit comprised absolute fit measures, incremental fit measures, and parsimonious fit measures (Hair, Black, Babin, & Anderson, 2010). Although the whole model (χ2 (49) = 516, p < .05) did not meet the fit criteria, results of statistical output indicated that errors of measure-ment of some variables were correlated, including self-efficacy and affective, group cohesion and peer relationship, and teaching implementation and teacher support. Both self-efficacy and affective component can influence behavior (Bandura, 1986; Pintrich, Smith, Garcia, & McKeachie, 1993). Students who exhibit high self-efficacy levels in sport demonstrate positive attitudes toward physical education, indicating that both variables are related to sport participation. Both group cohesion and peer relationship have common concepts in social interaction. Students who exhibit high group cohesion have strong peer relationships, indicating that both variables are related to interpersonal relationships. Both teaching implementation and teacher support have common concepts for teaching behavior. Teachers with better quality of teaching implementation would have more teacher support for their students, indicating that both variables are related to teaching behavior. This study set the errors of measurement correlation between these variables; when accounted for, the χ2 value declined by 227 (χ2 (46) = 289, p < .05). Other fit indices showed that the data adequately fit the hypothetical model (see Table 3). Hu and Bentler (1999) indicated that χ2 is the most commonly applied statistical indicator. Because the value of χ2 is affected by the sample size, a large number of participants can cause it to be inflated when assessing model fit. Therefore, others measurement indices have been applied to test model fit (Hair et al., 2010; Jöreskog & Sörborn, 2003). As shown in Table 3, other indices showed that the data adequately fit the revised model. Specifically, SRMR (0.05) was the same as the ideal value of 0.05; RMSEA (0.09) was smaller than 0.10; GFI (0.91) was greater than the required threshold of 0.90; NFI (0.96) was greater than 0.90; RFI (0.96) was greater than 0.90; IFI (0.97) was greater than 0.90; NNFI (0.96) was greater than 0.90; CFI (0.97) was greater than 0.90; AGFI (0.84) was greater than 0.80. Generally speak-ing, the measurement indices of overall model fit were acceptable, and thus, the model had acceptable external quality.

Fit of Internal Structure of the Model. The fit of the internal structure of the model is used to examine whether the model met the criteria for acceptable fit

80

Tab

le 2

P

aram

eter

Est

imat

es f

rom

th

e S

tan

dar

diz

ed S

olu

tio

n

Theo

retic

alm

odel

Rev

ised

mod

el

para

met

erpa

ram

eter

es

timat

est v

alue

Com

plet

e st

anda

rdiz

ed

para

met

er e

stim

ates

para

met

er

estim

ates

t val

ueC

ompl

ete

stan

dard

ized

pa

ram

eter

est

imat

es

λ x 1

1.8

825

.64*

.93

.88

25.2

2*.9

2

λ x 2

1.8

124

.16*

.89

.82

24.2

6*.9

0

λ y11

.53

-.7

7.5

4-

.78

λ y21

.56

17.2

4*.7

7.5

717

.75*

.78

λ y31

.70

15.6

6*.7

1.6

414

.19*

.66

l y41

.75

20.4

7*.9

0.7

320

.22*

.88

λ y52

.46

-.5

1.4

3-

.48

λ y62

.73

11.4

6*.8

5.7

510

.77*

.87

λ y72

.75

11.1

3*.8

0.7

110

.17*

.75

λ y82

.48

9.72

*.6

1.4

611

.49*

.58

λ y93

.73

-.8

5.7

0-

.84

λ y10

3.6

723

.47*

.84

.69

23.5

4*.8

7

γ 11

.68

12.9

8*.6

8.7

013

.39*

.70

γ 21

.51

8.45

*.5

1.3

56.

44*

.35

γ 31

.37

5.60

*.3

7.3

56.

35*

.35

β 21

.53

8.45

*.5

3.7

18.

85*

.71

(con

tinu

ed)

81

Theo

retic

alm

odel

Rev

ised

mod

el

para

met

erpa

ram

eter

es

timat

est v

alue

Com

plet

e st

anda

rdiz

ed

para

met

er e

stim

ates

para

met

er

estim

ates

t val

ueC

ompl

ete

stan

dard

ized

pa

ram

eter

est

imat

es

β 32

.67

7.81

*.6

7.6

68.

16*

.66

δ 1.1

37.

55*

.15

.13

7.35

*.1

5

δ 2

.16

10.0

5*.2

1.1

69.

01*

.19

ε 1.2

012

.99*

.41

.19

12.7

2*.3

9

ε 2.2

212

.94*

.41

.21

12.6

8*.3

9

ε 3.4

713

.65*

.50

.54

13.5

4*.5

7

ε 4.1

38.

51*

.19

.16

9.68

*.2

3

ε 5.6

014

.81*

.74

.62

14.8

5*.7

7

ε 6.2

011

.96*

.28

.17

10.8

5*.2

3

ε 7.3

213

.19*

.36

.39

13.6

6*.4

3

ε 8.3

814

.52*

.63

.40

14.6

1*.6

6

ε 9.2

011

.58*

.28

.21

12.1

7*.3

0

δ 10

.19

12.2

2*.2

9.1

510

.73*

.24

θε3e

4.0

94.

58*

.09

θε5e

8.2

07.

74*

.20

θε7e

9.1

48.

09*

.14

*p <

.05

Tab

le 2

(c

on

tin

ued

)

82 Pan

Table 3 Measurement Index Analysis of Overall Model Fit

Measurement index

Ideal numerical

value

Value of theoretical

model

Value of revised

modelMeasurement

criteria

Absolute fit measures

χ2 p>.05 χ2 (49)=516, p<.05 χ2 (46)=289, p<.05 No

SRMR †.05 0.06 0.05 YesRMSEA <.10 0.14 0.09 YesGFI >.90 0.84 0.91 YesAGFI >.80 0.75 0.84 Yes

Incremental fit measures

NFI >.90 0.95 0.97 YesRFI >.90 0.93 0.96 YesIFI >.90 0.95 0.97 YesNNFI >.90 0.94 0.96 YesCFI >.90 0.95 0.97 Yes

Parsimonious fit measures

PNFI ≥ .50 0.71 0.68 Yes

PGFI ≥ .50 0.53 0.53 Yes

Note. Standardized root mean-square residual (SRMR), Root mean square error of approximation (RMSEA), Goodness of fit index (GFI), Adjusted goodness of fit index (AGFI), Normal fit index (NFI), Relative fit index (RFI), Incremental fit index (IFI), Nonnormed fit index (NNFI), Comparative fit index (CFI), Parsimonious normed fit index (PNFI)

(Bagozzi & Yi, 1988). Hair et al. (2010) recommended that the individual reliabil-ity of observed variables and the standardized factor loading of observed variables be over 0.50. The results of this study showed that three of twelve observation variables were not over 0.50, but others met these criteria. Bagozzi and Yi (1988) suggested that the composite reliability (CR) of latent variables should reach 0.60. As noted in Table 4, all latent variables fulfilled this criterion. Bagozzi and Yi also suggested that average variance extracted of latent variables should reach 0.50. Most indices met this criterion except the learning atmosphere variable (0.47), which was close to the acceptable level. Consequently, it was concluded that the revised model had an acceptable fit to the empirical data supporting H1.

Effects on the Latent Variables in the Theoretical Model

Direct Effects. First, according to Figure 2, the direct effects of teachers’ self-efficacy on students’ learning motivation, learning atmosphere, and learning satisfaction were 0.70, 0.35, and 0.35, respectively. Second, physical education teachers’ self-efficacy had the highest effect on learning motivation (0.70), thus

Teachers’ Efficacy and Students’ Satisfaction 83

Table 4 Measurement Indices for Fit of Internal Structure of Model

Latent variables

Composite reliability of latent variables

Average variance extracted

from latent variables Observation variables

Individual reliability

of observation variables

Teacher self-efficacy

0.91 0.83 personal teaching efficacy 0.92general teaching efficacy 0.90

Learning motivation

0.86 0.61 value, 0.78expectation, 0.78self-efficacy 0.66affective 0.88

Learning atmosphere

0.77 0.47 group cohesion, 0.48learning enjoyment, 0.87teacher support, 0.75peer relationship 0.58

Learning satisfaction

0.84 0.73 teaching implementation 0.84learning effect 0.87

supporting H2. Third, the direct effect of learning motivation on learning atmo-sphere was 0.71. Thus, H3 was supported. Moreover, the direct effect of learning atmosphere on learning satisfaction was 0.66. Thus, H4 was supported.

Indirect Effects. First, the indirect effect of teachers’ self-efficacy on learning atmosphere was mediated by learning motivation. According to Table 5, the indirect effect was 0.50, and it was greater than direct effect of teachers’ self-efficacy on learning atmosphere (0.35). This means that the mediator, learning motivation, can improve the influence of teachers’ self-efficacy on the learning atmosphere. It also shows that the relationship between teachers’ self-efficacy and learning atmosphere was mediated by learning motivation. Second, the indi-rect effect of teachers’ self-efficacy on learning satisfaction was also mediated by learning atmosphere. The indirect effect was 0.23. This means that teachers’ self-efficacy had an indirect effect on learning satisfaction mediated by learning atmosphere, which effect was smaller than direct effect of teachers’ self-efficacy on learning satisfaction (0.35). It also shows that the relationship between teach-ers’ self-efficacy and learning satisfaction was mediated by learning atmosphere. Moreover, the indirect effect of teachers’ self-efficacy and learning satisfaction was mediated by learning motivation and learning atmosphere. The indirect effect was 0.33, and it was close to direct effect of teachers’ self-efficacy on learning satisfaction (0.35). The analysis of indirect effects, therefore, supports H5. This means that the relationship between teachers’ self-efficacy and students’ learn-ing satisfaction was mediated by learning motivation and learning atmosphere.

84 Pan

Figure 2 — Parameter estimates of the revised model for teachers’ self-efficacy, learning motivation, learning atmosphere, and learning satisfaction.

Total Effects. First, the total effect of teachers’ self-efficacy on learning motiva-tion is described in Table 5. Teachers’ self-efficacy had a strong positive effect (0.70) on learning motivation, explaining 49% of the total variance. Second, with regard to the total effect of teachers’ self-efficacy on the learning atmosphere, according to Table 5, there were direct effects (0.35) and indirect effects (0.50). The total effect of teachers’ self-efficacy on learning atmosphere, then, was 0.85. Learning satisfaction was explained much more by indirect than by direct effects. Teachers’ self-efficacy had a direct effect on learning atmosphere and an indirect effect on learning atmosphere mediated by learning motivation, explaining 95% of the total variance. Moreover, in terms of the total effect of teachers’ self-efficacy on learning satisfaction, as noted in Table 5, the direct effects (0.35) and indirect effects (0.23 and 0.33) resulted in a total effect of teachers’ self-efficacy on learn-ing satisfaction of 0.91. Teachers’ self-efficacy had a direct effect on learning satisfaction and an indirect effect on learning satisfaction mediated by learning motivation and learning atmosphere, explaining 95% of the total variance.

85

Tab

le 5

E

ffec

ts o

f Tea

cher

s’ S

elf-

effic

acy,

Lea

rnin

g M

oti

vati

on

, Lea

rnin

g A

tmo

sph

ere,

an

d L

earn

ing

Sat

isfa

ctio

n

Teac

hers

’ sel

f-ef

ficac

y>

lear

ning

mot

ivat

ion

Teac

hers

’ sel

f-ef

ficac

y >

lear

ning

atm

osph

ere

Teac

hers

’ sel

f-ef

ficac

y >

lear

ning

sat

isfa

ctio

nle

arni

ng m

otiv

atio

n >l

earn

ing

atm

osph

ere

lear

ning

atm

osph

ere

>lea

rnin

g sa

tisfa

ctio

n

Dire

ct e

ffect

0.70

0.35

0.35

0.71

0.66

Teac

hers

’ sel

f-effi

cacy

>

lear

ning

mot

ivat

ion

> le

arni

ng at

mos

pher

e

Teac

hers

’ sel

f-effi

cacy

>

lear

ning

atm

osph

ere

> le

arni

ng sa

tisfa

ctio

nIn

dire

ct e

ffect

0.50

0.23

Teac

hers

’ sel

f-effi

cacy

>

lear

ning

mot

ivat

ion

> le

arni

ng at

mos

pher

e >

lear

ning

satis

fact

ion

Indi

rect

effe

ct0.

33

Tota

l effe

ct0.

700.

850.

910.

710.

66

86 Pan

DiscussionThe results indicated that the revised model had acceptable fit. This finding empiri-cally substantiates the positive causal relationship among teachers’ self-efficacy, and students’ learning motivation, learning atmosphere, and learning satisfaction in senior high school physical education. In sum, the total effect of learning sat-isfaction reached 0.91, and the results support the first hypothesis (H1). That is, physical education teachers’ self-efficacy had a positive direct effect on learning motivation, learning atmosphere, and learning satisfaction, and it had a positive indirect effect on learning satisfaction mediated by learning motivation and learning atmosphere. The causal relationship model of teachers’ self-efficacy and students’ learning satisfaction had a high explained variance (95% of the total variance). With a direct effect 0.35, teachers’ self-efficacy positively affected the students’ learning satisfaction; however, teachers’ self-efficacy also had an indirect effect (0.23, 0.33) on learning satisfaction mediated by learning motivation and learning atmosphere, which increased the total effect of teachers’ self-efficacy on satisfaction (0.91). This means that physical education teachers’ self-efficacy was just one of the factors that influenced students’ learning satisfaction, and it also had an indirect impact on learning satisfaction mediated by students’ learning motivation and learning atmosphere, which could increase the impact on learning satisfaction. The impli-cation is that physical education teachers should increase their self-efficacy, use good curriculum design and instruction strategies to stimulate students’ motivation, and create a strong atmosphere for learning. Ultimately, putting these factors into practice will increase students’ learning satisfaction.

The second hypothesis (H2) was that teachers’ self-efficacy would have a positive effect on students’ learning motivation, learning atmosphere, and learn-ing satisfaction, respectively. Figure 2 shows that teachers’ self-efficacy had a significant positive correlation with students’ learning motivation, atmosphere, and satisfaction, supporting the second hypothesis. These results are similar to those of previous studies. Caprara et al. (2003) indicated that teachers’ self-efficacy could influence students’ learning motivation. Bay et al. (2010) demonstrated that teachers with high self-efficacy could create a democratic learning atmosphere and a friendly learning environment. Tessier et al. (2010) also showed that teachers’ teaching style could influence students’ psychological need satisfaction in physical education class. Based on Bandura’s social cognitive theory, Martin, McCaughtry, Hodges-Kulinna and Cothran (2008) indicated that teachers’ self-efficacy play a key role in how physical educators think about their curriculum design and teach-ing strategies. As reported in Figure 2, teachers’ self-efficacy had the highest effect on motivation (0.70), and had a lower effect on learning atmosphere (0.35) and satisfaction (0.35), respectively, indicating that teachers’ self-efficacy may be more effective in promoting motivation than learning atmosphere and satisfaction. Therefore, if students feel their teachers have higher professional competence, they will be more motivated to participate in physical education activities. In contrast, teachers with lower self-efficacy may contribute to lower learning participation. Learning atmosphere and satisfaction are influenced by other factors as well, includ-ing teaching styles, teacher–student relationship, and learning performance. Thus, teachers with high self-efficacy can influence motivation more than they affect learning atmosphere and satisfaction.

Teachers’ Efficacy and Students’ Satisfaction 87

Based on the findings of the current study, teacher’s self-efficacy could posi-tively influence learning motivation, learning atmosphere, and learning satisfaction for students. This finding explicitly shows that teachers’ self-efficacy can play a key role in influencing students’ learning processes. Physical education teachers should participate in professional growth programs and teaching workshops to enhance their professional competence in the areas of curriculum design, teaching strategies, and learning assessments, which could improve their personal teaching efficacy (Gibson & Dambo, 1984). On the other hand, teachers may also want to strengthen their teaching practices to be more effective, thereby enhancing their general teaching efficacy. In educational settings, teachers with high self-efficacy consider the needs of the students, respect pupils’ opinions, and design interest-ing and challenging programs to motivate students. Moreover, physical education teachers should develop friendly relationship with their students, including provid-ing positive encouragement and feedback, and creating harmonious interactions, which could create a better learning atmosphere for students’ learning processes. Furthermore, a teacher with higher efficacy should also set different teaching objec-tives and adjust the difficulty level of the curriculum for students. Students with higher abilities should be given more challenging tasks, while the students with lower abilities also have the opportunity to achieve success with fewer challenging tasks. In sum, teachers’ self-efficacy can directly and positively influence students’ learning motivation, learning atmosphere and learning satisfaction. Thus, physi-cal education teachers must enhance their own teachers’ self-efficacy to improve students’ learning performance.

The third hypothesis (H3) was that students’ learning motivations would have a positive effect on learning atmosphere. Figure 2 shows that students’ learning motivation had a significant positive influence on learning atmosphere, supporting the third hypothesis. Beggs et al. (2005) showed that an individual’s motivation in sport participation was related to enjoyment of sport activity. Motivation guides human behavior, inspiring people to actively engage in an activity. As illustrated in Table 5, learning motivation had a higher direct effect (0.71) than the direct effect of teachers’ self-efficacy on learning atmosphere (0.35). On the other hand, teachers’ self-efficacy did have an indirect effect (0.50) on learning atmosphere mediated by learning motivation; this effect was also higher than the direct effect of teachers’ self-efficacy on learning motivation (0.35). This indicates that teach-ers’ self-efficacy mediated by students’ learning motivation had a higher effect on learning atmosphere than the direct effect of teachers’ self-efficacy. Rink (2010) indicated that motivated students approach a learning task positively and with more intensity, but unmotivated students may not want to participate in learning activities and spend time avoiding the task. Generally speaking, the students who have higher motivation to engage in physical education programs can create a more positive learning atmosphere. Thus, it can be seen that students’ learning motivation can have a positive influence on the learning atmosphere of physical education classes. This finding explicitly shows that inspiring students’ learning motivation could increase the learning atmosphere in class. In other words, teachers should design attractive programs, use direct and indirect teaching strategies to improve students’ learning value, expectancy, achievement and affective components in physical education.

As reported in Figure 2, students’ learning atmosphere had a significant positive influence on learning satisfaction, supporting the fourth hypothesis. These results

88 Pan

are similar to those of previous studies (Chan & Chan, 2011; Wang, 2011). Learning atmosphere had four factors: group cohesion, learning enjoyment, teacher support, and peer relationship. Physical education curriculum should meet educational goals and the students should also get more learning satisfaction in physical education curriculum (Lund, 2010). Zhang, Solmon and Gu (2012) also found that a support-ive environment is positively associated with students’ achievement outcomes in physical education. Students’ achievement outcomes are the same as the effects of learning for students’ learning satisfaction in this study. Physical education teachers can use positive and effective teaching strategies to enhance the learning atmosphere in class. For example, Hellison (1995, 2011) developed the teaching personal and social responsibility (TPSR) model, which is a good teaching strategy for improv-ing students’ responsible behavior. TPSR could enhance pupils’ level of personal and social responsibility (Gordon, 2010), including respect, effort, self-direction and caring. Students with stronger sense of responsibility could exhibit responsible personal and social behavior, such as exhibiting self-respect and respect for others in physical activity settings, which helps create a good learning atmosphere. If teachers create a good learning atmosphere that improves group cohesion, enhances student enjoyment, supports student learning, and builds positive peer relationship in class, students will gain greater learning satisfaction.

The fifth hypothesis (H5) was that the relationship between teacher self-efficacy and student learning satisfaction would be mediated by learning motivation and learning atmosphere. Figure 2 shows that H5 was supported. On the one hand, as illustrated in Table 5, teachers’ self-efficacy and learning satisfaction was medi-ated by learning atmosphere, with an indirect effect of 0.23. On the other hand, learning satisfaction was mediated by learning motivation and learning atmosphere, with an indirect effect of 0.32. In this case, teachers’ self-efficacy had a higher effect mediated by learning motivation and atmosphere than mediated by learn-ing atmosphere alone. Strand and Scantling (1994) indicated that contemporary physical education programs could ultimately encourage more students to actively participate in physical activities if the programs were more enjoyable and varied, and if they more closely reflected the needs of students. If teachers can stimulate student motivation for learning, this could improve classroom learning atmosphere and later influence learning satisfaction.

Teachers’ self-efficacy has a positive direct influence on learning motivation, learning atmosphere, and learning satisfaction, and a positive indirect influence on learning satisfaction mediated by learning motivation and learning atmosphere. The Social Cognitive Theory (Bandura, 1986) notes that efficacy beliefs influence teachers’ decision-making, efforts, and perseverance as they perform their tasks. Teachers’ self-efficacy plays a key role in motivating the students to participate in learning. Bandura (1997) also proposed that teachers’ beliefs in their instructional efficacy partly determine how they structure academic activities in their classrooms and shape students’ evaluations of their intellectual capabilities. Teachers with high self-efficacy can use appropriate teaching strategies to implement teaching activities. However, increasing students’ motivation and creating a better learning atmosphere could also further enhance learning satisfaction. The development of an appropriate learning atmosphere is the most important factor for influencing learning satisfac-tion. In physical education teaching practices, teachers could adopt Mosston and

Teachers’ Efficacy and Students’ Satisfaction 89

Ashworth’s (2002) reciprocal teaching style to improve peer relationship. Teachers may also use cooperative learning to enhance group cohesion (Dyson, Linehan, & Hastie, 2010). Moreover, the sport enjoyment model (Scanlan & Lewthwaite, 1986) indicated that the sources of sport enjoyment include two dimensions—extrinsic-intrinsic and achievement-nonachievement, and thus teachers can create a conducive learning environment to maximize the outcomes in each dimension and enhance enjoyment. The physical education curriculum should foster a good learning atmosphere to contribute to student learning satisfaction.

This study used a logical and clear research framework. Its sampling procedure was rigorous and the research tools were psychometrically sound for the current population. However, it is worth considering the measurement of teachers’ self-efficacy to determine whether the phenomenon of social expectations influence this construct. Some teachers may have higher social expectations that may cause bias in their questionnaire responses. Therefore, strategies to avoid the problems of social expectation should be considered. Finally, emphasizing research anonymity, future studies might administer a social expectation scale to understand and avoid the influence of social expectation on teachers’ perceived self-efficacy. Another limitation of the study is the manner in which the questionnaires were distributed as this may relate to potential elements of coercion. The researcher made an effort to alleviate pressure given the number of schools selected. But since the heads of school physical education offices handed out the questionnaires in this study, there remains a possibility some students may have felt pressured to participate. In future research projects of this nature, one strategy to deal with potential coercion in broad survey designs is to select proxies not affiliated with the study to administer the questionnaires. With teachers/administrators absent, participants may feel less pres-sure and the researcher can then ensure consistency in the data gathering protocol.

The present study not only provides a comprehensive model regarding teachers’ self-efficacy, and students’ learning motivation, classroom learning atmosphere, and learning satisfaction in high school physical education classes, but also shows direct and indirect effects among the latent variables. However, it looks only at effects with high school students as participants. Future studies can include middle school and elementary school students for cross-validation of the theoretical model. Because the sample was drawn only from Taiwan, we make no attempt to generalize beyond the immediate population. To obtain evidence of cross-culture validity, we strongly recommend that researchers examine the linear structure relationship of the model in different countries. We also recommend that further studies explore the factors that influence the self-efficacy of physical education teachers. The use of qualitative measures to complement quantitative data may be of value to obtain comprehensive and in-depth findings. Furthermore, it will be beneficial to establish a more definite causal relationship among physical education teachers’ self-efficacy and students’ learning process.

ReferencesAldridge, J., Fraser, B., & Ntuli, S. (2009). Utilizing learning environment assessments to

improve teaching practices among in-service teachers undertaking a distance-education program. South African Journal of Education, 29, 147–170.

90 Pan

Bagozzi, R.P., & Yi, Y. (1988). On the evaluation of structural equation models. Journal of the Academy of Marketing Science, 16(1), 74–94 10.1007/BF02723327. doi:10.1007/BF02723327

Bandura, A. (1986). Social foundations of thought and action: A social cognitive theory. New York, NY: Prentice Hall.

Bandura, A. (1993). Perceived self-efficacy in cognitive development and functioning. Educational Psychologist, 28(2), 117–148 10.1207/s15326985ep2802_3. doi:10.1207/s15326985ep2802_3

Bandura, A. (1997). Self-efficacy: The exercise of control. New York, NY: Freeman.Bay, E., Bündoğdu, K., & Kaya, H.I. (2010). The perceptions of prospective teachers on

the democratic aspects of the constructivist learning environment. Electronic Journal of Research in Educational Psychology, 8, 617–642.

Beggs, B.A., Elkins, K.J., & Powers, S. (2005). Overcoming barriers to participation in campus recreational sports. Recreational Sports Journal, 29(2), 143–155.

Bollen, K.A., & Long, J.S. (1993). Testing structural equation models. Newbury Park, CA: Sage Publications.

Caprara, C., Barbaranelli, L., Borgogni, L., & Steca, P. (2003). Efficacy beliefs as determi-nants of teachers’ job satisfaction. Journal of Educational Psychology, 92, 821–832 10.1037/0022-0663.95.4.821. doi:10.1037/0022-0663.95.4.821

Caprara, C., Barbaranelli, L., Steca, P., & Malone, P.S. (2006). Teachers’ self-efficacy beliefs as determinants of job satisfaction and students’ academic achievement: A study at the school level. Journal of School Psychology, 44, 473–490. doi:10.1016/j .jsp.2006.09.001

Chacon, C.T. (2005). Teachers’ perceived efficacy among English as a foreign language teachers in middle schools in Venezuela. Teaching and Teacher Education, 21, 257–272 10.1016/j.tate.2005.01.001. doi:10.1016/j.tate.2005.01.001

Chan, C.C., & Chan, S.C. (2011). An investigation of classroom climate, learning motiva-tion and learning satisfaction in physical education class of Taipei city senior high and vocational school students [Abstract]. Physical Education Journal, 44, 437-458. Abstract retrieved from http://www.airitilibrary.com

Chen, H.C., & Stotlar, D.K. (2012). An examination of the motivation and satisfaction of college students enrolled in physical education courses. Sport Science Review, 1(2), 43–63 10.2478/v10237-012-0003-9.

Deci, E.L., & Ryan, R.M. (2002). Handbook of self-determination research. Rochester, NY: University of Rochester Press.

Dyson, B.P., Linehan, N.R., & Hastie, P.A. (2010). The ecology of cooperative learning in elementary physical education classes. Journal of Teaching in Physical Education, 29, 113–130.

Gibson, S., & Dambo, M.H. (1984). Teacher efficacy: A construct validation. Journal of Edu-cational Psychology, 76(4), 569–582 10.1037/0022-0663.76.4.569. doi:10.1037/0022-0663.76.4.569

Gordon, B. (2010). An examination of the responsibility model in a New Zealand second-ary school physical education program. Journal of Teaching in Physical Education, 29, 21–37.

Gorozidis, G., & Papaioannou, A. (2011). Teachers’ self-efficacy, achievement goals, attitudes and intentions to implement the new Greek physical education curriculum. European Physical Education Review, 17, 231–254 10.1177/1356336X11413654. doi:10.1177/1356336X11413654

Ha, A.S., Johns, D.P., & Shiu, E.W. (2003). Students’ perspective in the design and imple-mentation of the physical education curriculum. Physical Educator, 60(4), 194–207.

Hair, J.F., Black, W.C., Babin, B.J., & Anderson, R.E. (2010). Multivariate data analysis (7th ed.). Englewood Cliffs, NJ: Prentice Hall.

Hellison, D. (1995). Teaching responsibility through physical activity. Champaign, IL: Human Kinetics.

Teachers’ Efficacy and Students’ Satisfaction 91

Hellison, D. (2011). Teaching personal and social responsibility through physical activity (3rd ed.). Champaign, IL: Human Kinetics.

Henson, R. (2002). From adolescent angst to adulthood: Substantive implications and measurement dilemmas in the development of teacher efficacy research. Educa-tional Psychologist, 37(3), 137–150 10.1207/S15326985EP3703_1. doi:10.1207/S15326985EP3703_1

Hsieh, K.C. (1997). The establishment of the evaluation indicators and the weight system of school effectiveness in elementary education (Unpublished doctoral dissertation). National Chegchi Univeristy, Tawian.

Hu, L., & Bentler, P.M. (1999). Cutoff criteria in fix indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling, 6(1), 1–55 10.1080/10705519909540118. doi:10.1080/10705519909540118

Huang, R.Y. (2007). The effects of teaching styles, learning motivation, and learning strate-gies on college students’ learning satisfaction in table tennis course [Abstract]. Bulletin of Sport and Exercise Psychology of Taiwan, 10, 15–36 Abstract retrieved from http://www.airitilibrary.com.

Jöreskog. K. G., & Sörborn, D. (2003). Lisrel 8.54: User’s reference guide. Chicago, IL: Scientific Software International.

Kline, R.B. (1998). Principles and practice of structural equation modeling. New York, NY: Guilford Press.

Lund, J. (2010). Evaluating your physical education curriculum. In J. Lund & D. Tannehill (Eds.), Standards-based physical education curriculum (2nd ed., pp. 57–84). Sudbury, MA: Jones and Bartlett publishers.

Martin, J.J., McCaughtry, N., Hodges-Kulinna, P., & Cothran, D. (2008). The influences of professional development on teachers’ self-efficacy toward educational change. Physical Education and Sport Pedagogy, 13(2), 171–190 10.1080/17408980701345683. doi:10.1080/17408980701345683

Martínez-López, E., Sánchéz, M.Z., Alvarez, M.R., & Cruz, M.D. (2010). Self-efficacy expectations in teacher trainees and the perceived role of schools and their physi-cal education department in the educational treatment of overweight students. European Physical Education Review, 16, 251–267 10.1177/1356336X10385044. doi:10.1177/1356336X10385044

Mosston, M., & Ashworth, S. (2002). Teaching physical education (5th ed.). Boston, MA: Benjamin Cummings.

Pan, Y.H. (2012). The development of a teacher’s self-efficacy instrument for high school physical education teachers. World Academy of Science. Engineering and Technology, 66, 1152–1157.

Pan, Y.H., & Cheen, J.R. (2010). Evaluating validity and reliability of high school students’ learning atmosphere scale in physical education [Abstract]. Book of Abstracts (p. 232). National Institute of Education, Singapore: The 3rd International Conference of Physical Education and Sports Science.

Pan, Y.H., Chou, H.S., Hsu, W.T., Li, C.H., & Hu, Y.L. (2013). Teacher self-efficacy and teaching practices in the health and physical education curriculum in Taiwan. Social Behavior and Personality, 41(2), 241–250. doi:10.2224/sbp.2013.41.2.241

Pintrich, P.R., Smith, D.A.F., Garcia, T., & McKeachie, W.J. (1993). Reliability and predictive validity of the motivated strategies for learning questionnaire (MSLQ). Educational and Psychological Measurement, 53, 801–814 10.1177/0013164493053003024. doi:10.1177/0013164493053003024

Rink, J.E. (2010). Teaching physical education for learning (6th ed.). New York, NY: McGraw-Hill Companies, Inc.

Roberts, G.C. (2001). Understanding the dynamics of motivation in physical activity: The influence of achievement goals on motivational processes. In G.C. Roberts (Ed.), Advances in motivation in sport and exercise (pp. 1–50). Champaign, IL: Human Kinetics.

92 Pan

Rutten, C., Boen, F., & Seghers, J. (2012). How school social and physical environments relate to autonomous motivation in physical education: The mediating role of need satisfaction. Journal of Teaching in Physical Education, 31, 216–230.

Scanlan, T.K., & Lewthwaite, R. (1986). Social psychological aspects of competition for male youth sport participants: IV. Predictors of enjoyment. Journal of Sport Psychol-ogy, 8, 25–35.

Shidler, L. (2009). The impact of time spent coaching for teacher efficacy on student achievement. Early Childhood Education, 36, 453–460 10.1007/s10643-008-0298-4. doi:10.1007/s10643-008-0298-4

Standage, M., Duda, J., & Ntoumanis, N. (2006). Students’ motivational processes and their relationship to teacher ratings in school physical education: A self-determination theory approach. Research Quarterly for Exercise and Sport, 77(1), 100–110.

Strand, B., & Scantling, E. (1994). An analysis of secondary student preferences toward physical education. Physical Educator, 51(3), 119–129.

Tessier, D., Sarrazin, P., & Ntoumanis, N. (2010). The effect of an intervention to improve newly qualified teachers’ interpersonal style, students’ motivation and psychological need satisfaction in sport-based physical education. Contemporary Educational Psychol-ogy, 35, 242–253 10.1016/j.cedpsych.2010.05.005. doi:10.1016/j.cedpsych.2010.05.005

Tsai, Y.Y. (2006). Confirmatory factor analysis of the hearing impaired students’ learning satisfaction with physical education inventory. The Journal of Special Education, 23, 185–200.

Tschannen-Moran, M., & Woolfolk, H.A. (2001). Teacher efficacy: Capturing an elusive con-struct. Teaching and Teacher Education, 17, 783–805 10.1016/S0742-051X(01)00036-1. doi:10.1016/S0742-051X(01)00036-1

Tschannen-Moran, M., Woolfolk Hoy, A., & Hoy, W.K. (1998). Teacher efficacy: Its meaning and measure. Review of Educational Research, 68(2), 202–248 10.3102/00346543068002202. doi:10.3102/00346543068002202

Wang, M.Y. (2011). Table tennis coach’s leadership behavior on athlete’s satisfaction-organizational atmosphere as an example. Journal of Sport and Recreation Research, 5(3), 83–95.

Xiang, P., Lee, A.M., & Solmon, M.A. (1997). Achievement goals and their correlates among American and Chinese students in physical education: A cross-cultural analy-sis. Journal of Cross-Cultural Psychology, 28, 645–660 10.1177/0022022197286001. doi:10.1177/0022022197286001

Zhang, T., Solmon, M.A., & Gu, X. (2012). The role of teachers’ support in predicting students’ motivation and achievement outcomes in physical education. Journal of Teaching in Physical Education, 31, 329–343.