4.2 Path analysis among teacher support, basic needs, and learning engagement. 64
4.2.3 Basic needs and learning engagement
國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
70
4.2.3 Basic needs and learning engagement
Another relationship that present study wanted to explore was basic need and learning engagement. Among three basic needs variables, only need of competence can be negative related to both behavioral (β43
= -.08 , p < .05) and emotional
engagement (β53= -.09 , p < .05) with a significant beta coefficient. However, the
relationship between need of autonomy and learning engagement as well as need of relatedness and learning engagement displayed nonsignificant relationships.Squared multiple correlations (R-square) showed that the variance of basic needs variables could be explained by teacher support are as follows: need of autonomy (R = 0.03), need of relatedness (R= 0.25), and need of competence (R=
0.09). Among the three basic needs, need for relatedness can be explained more variance than the others do. See Figure 4.2.
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
71
4.3 Reduced form of the significant model
In order to explore further the effect of teachers support variables on learning engagement, the reduced form of the significant model was proposed to compare with the significant paths model because basic needs variables showed less mediated role between teacher support and learning engagement. Moreover, teacher support of structure had no significant effect on learning engagement. Therefore, the reduced form of significant paths, which deleted basic need variables and teacher support of structure was proposed to compare the differences. Fit indices were different between two models. RMSEA of the significant model was .064. However, RMSEA of reduced form was .083, which indicated that significant model exhibited better fit compared with reduced form. Parsimonious indices : PNFI and PGFI of reduced significant model and significant model showed that significant model had a better fitness. GFI and AFGI of reduced significant model were a little higher than significant model. The indices are showed in Table 4.6.
In reduced model, teacher support of involvement was positively predictive to behavioral learning engagement (γ41
= .20 , p < .01) and emotional learning
engagement(γ
51= .27 , p < .01). Teacher support of autonomy support had a
significant direct effect on behavioral learning engagement (γ42= .56 , p < .01) and
emotional learning engagement (γ
52= .53 , p < .01) (See figure 4.3).
‧
Figure 4.2 Significant paths of the hypothesized model.
* p < .05 ; ** p < .01
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
73
Table 4.5 The effect of latent variables in SEM model
Latent Variables Direct Indirect Total
Involvement(ξ1) B.N.Autonomy (η1) 0.14 - 0.14
Involvement(ξ1) B.N.Relatedness (η2) 0.36** - 0.36**
Involvement(ξ1) B.N.Competence(η3) 0.05 - 0.05
Involvement(ξ1) Behavioral
engagement(η4) 0.15* 0.03 0.18**
Involvement(ξ1) Emotional
engagement(η5) 0.21** 0.02 0.23**
AutonomySupport(ζ2) B.N. Autonomy (η1) 0.05 - 0.05
AutonomySupport(ζ2) B.N.Relatedness(η2)) 0.12 - 0.12
AutonomySupport(ζ2) B.N.Competence(η3) 0.23* - 0.23*
AutonomySupport(ζ2) Behavioral
engagement(η4) 0.58** -0.01 0.57**
AutonomySupport(ζ2) Emotional
engagement(η5) 0.55** -0.01 0.54**
Structure (ζ3) B.N. Autonomy (η1) -0.21** - -0.21**
Structure (ζ3) B.N.Relatedness (η2) 0.06 - 0.06
Structure (ζ3) B.N.Competence(η3) 0.07 - 0.07
Structure (ζ3) Behavioral
engagement(η4) 0.02 -0.01 0.01
Structure (ζ3) Emotional
engagement(η5) 0.06 -0.01 0.05
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
74
Table 4.5
The effect of latent variables in SEM model (continued)
Latent Variables Direct Indirect Total
B.N.Autonomy (η1) Behavioral
engagement(η4) 0.04 - 0.04
B.N.Autonomy (η1) Emotional
engagement(η5) 0.02 - 0.02
B.N.Relatedness(η2) Behavioral
engagement(η4) 0.07 - 0.07
B.N.Relatedness(η2) Emotional
engagement(η5) 0.05 - 0.05
B.N.Competence(η3) Behavioral
engagement(η4) -0.08* - -0.08*
B.N.Competence(η3) Emotional
engagement(η5) -0.09* - -0.09*
Notes. B.N. indicates basic need *p＜.05 ; **p＜.01
‧
Figure 4.3.Reduced form of the significant relationship between teacher support and learning engagement. * p < .05 ; ** p < .01
Table 4.6 The fit indices of the reduced significant model compared with the significant model
Absolute fit dices Relative fit indices Parsimonious indices X2 df RMSEA GFI AGFI NFI NNFI PNFI PGFI
Notes. Numbers in parentheses are fit indices of the full model
Involvement
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
76
Chapter Five Discussion
This chapter begins with the research findings related to each research question.
Next, the discussions of the general points about the findings were outlined next. The limitations of the study and implications for future research concluded this chapter.
5.1 General findings of the present study
The first research question of the study was “Are students’ basic needs affected by teacher support?” The statistical results of the proposed model showed that teacher support of involvement could significantly predict need of relatedness. In addition, teacher autonomy support can affect need of competence with significant coefficient. Moreover, need of autonomy was negatively affected by teacher support of structure. The second research question of the study was “Is learning engagement affected by basic needs?” The result showed that student need of competence was negative related to both behavioral and emotional learning engagement in high school learning context. The third research question of the study was “Can teacher support affect learning engagement directly?” Based on result of the structural equation model, teacher support can predict learning engagement with large variance. Teacher involvement and teacher autonomy support can significantly predict behavioral and emotional learning engagement.
A unique feature of this study was that participants were asked to response to items by choosing a specific teacher as a target subject rather than perceived teachers’
behaviors generally. Therefore, students answered according to a specific teacher’s supporting behaviors (teacher support) and students’ behaviors in that specific class (learning engagement). This could clarify the relationship between teacher behaviors and student engagement.
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
77
5.2 The effect of teacher autonomy support
The conceptualizations of teacher autonomy support in this research included three subcategories: providing choices, fostering relevance, and monitoring.
Providing choices and fostering relevance are essential autonomy support behaviors.
The subcategory was “monitoring”, which originally belonged to teacher support of structure, was incorporated into this autonomy support subscale. In the present study, teacher autonomy support was predictive to both learning engagement and need of competence. However, teacher autonomy support failed to predict need of autonomy.
The first line of relationship would discuss from teacher autonomy support and need of competence then to need of competence and learning engagement.
5.2.1 Teacher autonomy support and need of competence
Need of competence measured the degree to which one feels efficacious about the task. Need of competence has been proven to be beneficial to various learning outcomes including self-determined motivation, well-being, and the success of cooperative learning instruction (Vallerand, Fortier, & Guay, 1997 ; Levesque et al., 2004 ; Haenze & Berger, 2007 ; Marchand & Skinner, 2007). Since need of competence is important to predict learning outcomes, increasing the need of competence among classroom learners becomes an educational goal. In this study, learners’ competence was shaped by teacher autonomy-supportive behaviors, which included giving choices, fostering relevance, and monitoring students’ learning,
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
78
5.2.2 Need of competence and learning engagement
Need of competence negatively predicted both behavioral and emotional learning engagement in the present study. The first possible reason was the importance and challenge of the classroom learning tasks. Based on the flow theory, providing proper challenge and meaningful task promote learners’ intensive engagement (Kowal & Fortier, 1999 ; Mills & Fullagar, 2008). If teachers provide too much or too less challenge, it makes learners have less willingness to participant.
Moreover, if the tasks do not provide proper meanings for learners, students might have no intention to join the activity or they may participate passively in the classes.
They may engage in the classes to obey or comply classroom rules. Therefore, in Taiwanese high school classrooms, students with either high degree of competence or lower degree of competence may regard traditional didactic or lecture as boring and then pay less attention. Providing meaningful tasks and help learners value their leanings as well as providing optimal challenge could promote high engaging classes.
The second plausible explanation was that classroom goal structure or teaching methods may affect learners’ engagement in the classes (e.g., Gonida, Voulala, and Kiosseoglou, 2009 ; Greene, Miller, Crowson, Duke, & Akey, 2004). Constructive instruction, cooperative learning, or traditional didactic method may produce different outcomes for different types of learners. Giving a speech and lecturing are popular teaching methods in high school classrooms. Different methods of instruction which connected to distinct classroom goal structures may produce different results in basic needs. Different types of instructions could meet learners’
needs to a different degree. A research based on the experimental design showed jigsaw cooperative learning instruction benefited more three basic needs as well as
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
79
learning engagement compared with direct instruction condition (Haenze & Berger, 2007). In addition, Gonida et al showed that parents and school mastery goal structure are both critical in predicting student personal mastery goal and behavioral engagement.
Consequently, different teaching instructional methods which connected to classroom goal structures may affect students’ engagement.
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
80
5.2.3 Teacher autonomy support and need of autonomy
The next path relationship discussed first from teacher autonomy support and need of autonomy and then to need of autonomy and learning engagement.
Need of autonomy assessed students’ beliefs of how they wanted and wished to choose as well as control their learning activities. (e.g., I want to decide for myself how to live my life”). Teacher support of autonomy had no significant effect on need of autonomy. The possible explanation was that in this study need of autonomy indicated that students had the desires as well as whishes to be autonomous and independent. These students were less satisfied with their learning environments.
Therefore, they tended to perceive teachers’ behaviors less positively.
Need of autonomy was not predictive to learning engagement in this study. It indicated that students who wanted to be autonomous in learning would not engage in class. Students who had incorporated the feeling of dependency and autonomy in a classroom were different form those students who would like to be independent and autonomous. Wishing for independent and autonomous (conscious wants) suggested that they were less satisfied in the need of autonomy compared with learners who do not perceive this needs. For example, some studies investigated the three basic needs of Taiwanese junior high students. They used “I can” rather than
“I want” (e.g., “I can decide how to do my schoolwork”). These studies found that basic needs could predict learning outcomes including autonomous motivation (Chien, 2009), learning interest (Hsueh, 2009), and learning goals (Hsieh, 2008).
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
81
5.2.4 Teacher autonomy support and learning engagement
The general prediction has been made that students and teachers in collectivistic high school context were under the pressure of passing college entrance exam, so classrooms with accountability could be achieved by controlling behaviors in the classrooms. Under this proposition, the application of autonomy-supportive teaching is doubted because it may have less effectiveness in collectivist cultures.
However, this study showed that perceiving autonomy support from teachers was advantageous to learning engagement. Some research studies, which investigated Asian elementary school students, junior high school students, and college students, have been proven the effect of autonomy support on leaning engagement and motivation (d’Ailly, 2003; Vansteenkiste et al., 2005 ; Shih, 2008;
Jang et al., 2009). This study supported these findings and further extended the application of teachers’ autonomy support in high school level in an East Asian country Taiwan. This finding is also consistent with studies in the methods of both survey methods and experimental designs (Reeve, Bolt, & Cai, 1999 ; Reeve et al., 2004 ; Oliver, Markland, Hardy, & Petherick, 2008). Based on the findings of this study, some practical implications could be proposed as follows.
First, providing choices is an effective way to enhance students’ perceived competence, and intrinsic motivation (Patall, Cooper, & Robinson, 2008). Since providing choices have its function and usefulness, the suggestions could be proposed that providing choices should be implemented under autonomy-supportive context, should connect learners’ intrinsic motivation and volition (Reeve, Nix &
Hamm, 2003), and should consider other instructional situations including when and how to provide choices (Patall, et al., 2008).
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
82
Second, fostering relevance signifies providing the rationales and connections between learners and learning materials, which promotes behavior and cognitive engagement as well as positive feelings (Reeve, Jang, Hardre, & Omura, 2002 ; Jang, 2008 ; Oliver et al., 2008). This finding was proven by some experimental design studies. One experimental design study, for example, Jang investigated the effect of rationales in an uninteresting activity in college introductory statistics and found it can enhance learners’ motivation and engagement.
“Learning Climate questionnaire” (LCQ) is also used to survey teachers’
autonomy supportive behaviors, but teacher foster relevance is not included in this questionnaire. Therefore, fostering relevance could become not only practical implication for teacher education but also could suggest this autonomy-supportive behavior including in the future studies.
The third subscale was monitoring, which suggested that teachers observe and monitor students’ learning timely and check if they understand learning materials or need extra helps. No matter which kinds of instructions were conducted in the class, monitoring student learning could help understand the needs of students and then dynamically tune into a better classroom environment. This could help create effective classes for both learners and instructors.
Overall, teacher autonomy support could predict learning engagement. The first implication can be proposed that instructors should further activate cognitive support and deep learning under autonomy-supportive teaching. Teachers may see motivation as a kind of learning attitude rather than cognitive engagement and active participation in a classroom (Blumenfeld et al., 1991; Stefanou, et al., 2004).
Providing cognitive support is as helpful as providing autonomy-supportive instruction. For example, Tsai et al. (2008) found that both cognitive autonomy
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
83
support and autonomy-supportive climate could predict student interest experience.
Creating an instructional program that facilitates cognitive engagement may also be important (Raphael, Pressley, & Mohan, 2008). Raphael, Pressley, and Mohan found that elementary school students who were scaffolded cognitively, tended to highly engage in the classrooms.
Not only cognitive engagement but also proper challenge could add to the autonomy-supportive classrooms. Studies showed that flow experience (an optimal balance of skill and challenge) was related to intrinsic motivation, need for autonomy, and challenge of tasks (Kowal & Fortier, 1999 ; Mills & Fullagar, 2008).
Therefore, providing learners with cognitive, meaningful challenge might help their behavioral, emotional, and cognitive engagement. Autonomy-supportive schools can benefit students’ engagement and motivation. Therefore, the implication of the present finding for school administrators is that autonomy-supportive environment could be a school guideline and a philosophy of education. For example, Rathunde and Csikszentmihalyi (2005) compared traditional and Montessori high schools.
They found that Montessori students showed more flow experience and intrinsic motivation in academic activities than students in traditional high schools. Based on the results of this study and previous research, the application of autonomy-supportive principles can have a better classroom outcomes by giving proper cognitive challenging activities.
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
84
5.3 The effect of teacher involvement
Teacher support of involvement included three subcategories: affection, attunement, and dedication of resources. Teacher support of involvement could not only promote behavioral and emotional learning engagement but also fulfill need of relatedness. Teachers’ caring, respect, and understanding promoted engaged behaviors. This result extended the application of teacher involvement to higher levels of students (Skinner & Belmont, 1993 ; Ryan & Patrick, 2001; Skinner, et al., 2008). People believe that teacher-student relationship in high school context may have less intimacy and familiarity compared with the relationship in junior or senior high schools. However, this finding suggested that even in high school context, the connectedness between teachers and students was still critical.
Teacher support of involvement could foster learners’ need of relatedness and engagement. It indicated that teacher behaviors and learner behaviors are interactive and connected. Teachers provided caring, positive attitudes, and understanding could satisfy learners’ satisfaction of relatedness as well as facilitate their engagement.
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
85
5.3.1 Need of relatedness and learning engagement
Relatedness indicated the degree to which one felt a sense of belonging and connection with others, which was related to high school student motivations and learning behaviors (Skinner & Belmont, 1993 ; Furrer & Skinner, 2003 ; Legault, et al. 2006 ; Marchand & Skinner, 2007).
Collectivistic cultures emphasize the connections and relationships among individuals. Individualistic cultures tend to value more personal interests and pleasures. Self-determination theorists point out that basic needs are universal. Need of autonomy and competence are based on the relationship with significant others (Deci & Ryan, 2008).
The present study found that need of relatedness was not a predictive variable in explaining behavioral and emotional engagement. This finding is consistent with the result found among Korean high school students (Jang et al., 2009).
This result implied that in high school context, the satisfaction of relatedness among high school students did not facilitate their behavioral and emotional engagement. Students with different profiles of satisfaction in relatedness did not lead to adaptive engagements. Literature investigating social goals adoption might explain this result. At school, learners might engage in a learning context because of the social obligations, social compliance or other types of social goals. Social goals such as searching for approval or hoping for dependency from teachers might also be related to students’ need of relatedness and engagement in the classroom environments (Davis, 2006). For example, Kiefer and Ryan (2008) investigated social goals (dominance, intimacy, and popularity) from sixth to seventh graders and found that social dominance goal was related to maladaptive outcomes, intimacy goals were related with grades. Popular goals were not associated with engagement or
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
86
achievement. Another research conduced in Taiwan. Hsieh (2008) investigated Taiwanese junior high school students and found need of relatedness could positively predict social goals.
Therefore, social goal adoption may affect learners’ relatedness in the classroom.
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
87
5.4 The effect of teacher support of structure
The present study found that teacher support of structure could include teachers’ dependability, contingency, and non-control. Teacher support of structure was negatively related to need of autonomy, and could not predict learning engagement.
The first subcategory was dependability, which was originally a part of teacher involvement, included in this factor. The second subcategory was contingency, which indicated that the degree to which how teachers behave consistently. The third category was control, which was a part of teacher autonomy support in TSAC (Teacher as Social Context Questionnaire).
5.4.1 Teacher support of structure and need of autonomy
Teacher support of structure including teachers’ dependability, consistency, and non-intrusive or non-controlling behaviors had a reverse effect on need of autonomy, which indicated a desire for autonomy and independent. This might be the reason that teachers’ dependency, consistency, and nonintrusive behaviors would not fulfill learners’ need of autonomy. Students with high need of autonomy searched for independency, consistency, and more freedom, but not dependency, so a negative effect existed between two variables.
Based on this result, some profiles of learners with different basic needs profiles would perceive teacher support of structure differently. For example, a study investigated Taiwanese high school students and also found that students who felt more capable in learning tended to react more positively towards teachers (Hardre et al., 2006). Another research, for example, Tapola and Niemivirta (2008) found that students with more sense of autonomy and competence tended to response positively
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
88
toward autonomy-supportive teaching style. In summary, students with need of autonomy were less adaptive profiles, so they perceived teacher support structure more negatively.
Another dimension is teacher characteristics. Teachers with different profiles of motivation and regulatory behavior tend to produce different support and cognitive activation (Roth, et al., 2007; Klusmann et al., 2008). For example, Klusmann et al.
found that teacher with different motivation, engagement, and resilience to teach, tended to influence students differently. More specifically, teachers with better regulatory behavior produced more positive outcomes in providing support and cognitive activation.
Therefore, providing various trainings and supports for teachers could produce more effective teaching behaviors as well as effective classrooms.
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
89
5.5 Conclusion
The present study explored the proposed model containing the relationship among teacher support, basic needs and learning engagement so as to answer research questions. Self-system processes propose that teacher support effect influence personal processes which in turn shape learning actions and outcomes (Skinner & Belmont, 1993 ; Marchland & Skinner, 2007). Findings of this study partially supported this model, which shows as follows.
Significant relationships were found between teacher support and basic needs.
First, teacher involvement could positively affect relatedness to a large degree.
Second, teacher autonomy support was predictive to need of competence. Third, teacher structure was negatively related to need of autonomy.
Next, teacher autonomy support and teacher involvement could strongly explain both behavioral and emotional engagement. Lastly, between basic needs and learning engagement, need of competence was negatively predictive to learning engagement. Need of relatedness and autonomy showed no significant result in predicting learning engagement. To further explore the significant model of the hypothesized relationships, reduced form of significant model was proposed to compare the differences. It was found that reduced model showed another implication and perspective on explaining teacher support and learning engagement.
In conclusion, effective classrooms are created when teachers give choices, monitor learning as well as provide proper dependency, consistency, and caring in the classrooms. Incorporating the concepts of autonomy support, involvement, and structure could not only fulfill basic needs and but also promote learning engagement.
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
90
5.6 Limitations and suggestions for future research
The advantage of the present study was that teacher support and basic needs
The advantage of the present study was that teacher support and basic needs