There is increasing pressure for students to attain a high level of academic achievement in Malaysia. A high level of academic achievement is the basic requirement for students to enter the Sixth Form, which is the equivalent of 11th and 12th Grades in the United States educational system. The Sixth Form is the only pathway to qualify students to apply for further education in a Malaysian public university (Ministry of Education, 2004). Similar to trends in other countries, getting a degree increases the prospects of getting a higher paying job and increases the chances of obtaining a better lifestyle (Pascarella & Terenzini, 2005).
One of the goals of education is to ensure that students are equipped to fulfill their roles as citizens through personal, social, and civic responsibilities. Many teachers and parents want students to have meaningful and enduring learning experiences and develop a greater capacity to think critically and creatively. Critical and creative thinking are skills that are sought after by prospective employers and can help individuals adapt and thrive in the working environment (Driver, 2001; Scott & Bruce, 1994; Kane, Berryman, Goslin & Meltzer, 1990). The global economy of today demands that individuals should be more creative in addressing unique problems and situations (Leitch, 2006; Kanter, 1995). Therefore, schools have been encouraged to take up the challenge to prepare students to not only strive for academic achievement (e.g. getting A's) but to function effectively (i.e. solve semi-structured problems where hypotheses must be tested) in the challenging and changing global workplace (Levy & Murnane, 1996). It has been suggested that classrooms that have higher levels of autonomy supportive will meet both those functions (Niemiec & Ryan, 2009).
Although there are many factors that influence academic achievement (e.g. SES, family background, nutrition, income), school factors such as school culture, classroom size, curriculum and teachers play important roles in affecting achievement outcomes (Berliner, 2009). For example, based on a meta-analysis, Marzano (2000) reported that, on average, teachers accounted for 13% of the variance in students' academic achievement. Teachers are actively involved in creating a classroom environment that promotes learning through their instruction, curriculum design and classroom management (Cotton, 1995; Fraser, Williamson, & Tobin, 1987 and Wahlberg, 1969). Wigfield & Eccles (2000) proposed that "high-quality" (e.g., clear instructions) classroom environments can support students' persistence in a task and attainment of higher achievement. In contrast, a classroom climate that pressures students to merely achieve high scores in high-stakes testing can undermine students' motivation levels to learn (Pintrich & De Groot, 1990).
This study examines the relation between Malaysian classrooms' autonomy supportive learning environment and how it can set the stage for students' intrinsic motivation and academic performance. Although there is ample evidence for the relation between autonomy supportive classrooms and how it affects students' motivation in the United States, there is no empirical research that has examined this relationship in Malaysian schools. The theoretical framework for conceptualizing student motivation is centered on Deci and Ryan's Self-Determination Theory (Ryan & Deci, 2000), which posits that intrinsic motivation is innate and individuals can experience self-determination in a setting (e.g., classroom) that fosters it. The clarification of knowledge of autonomy support and self-determination from this study will introduce and give Malaysian teachers an alternate perspective and approach towards addressing academic performance issues in their classrooms. The knowledge will also help teachers to create effective intervention strategies to help students improve academic performance. In addition, teachers will be better equipped to prepare their students at the societal level to be more competitive in the Malaysian and global market.
There are three primary hypotheses (Hypothesis 1 to 3):
1) Students' intrinsic motivation is positively related to their perceived autonomy support.
2) Students' self-reported grades are positively related to their level of intrinsic motivation
3) Students' self-reported grades are positively related to their perceived autonomy support.
In addition to the primary hypotheses, exploratory analyses were conducted to explore the association of intrinsic motivation and students' goal orientations (Exploratory Analysis 1 to 3). It is estimated that goal orientations (mastery or performance-oriented) may influence intrinsic motivation and therefore play a role in promoting academic achievement too.
Chapter 2: Literature Review
History of Motivation Theory and Research
General scientific developments, such as Darwin's theory of evolution in biology, had strong influences on early motivation theories. These scientific ideas led to the conception of living organisms as types of machines, with motivation as the energy that fuels the machine (Weiner, 1990). The organism was thought to strive toward homeostasis, or an optimal state of satiation. Motivation for action was thought to derive from a deprivation that created a disruption of the homeostatic state. For example, deprivation of nourishment leads to motivation to seek food, and deprivation of interesting surroundings leads to motivation to seek stimulation. In these theories, the main explanation was attributed to "drive", a concept that aimed to restore homeostasis to an organism.
"Instincts," "Needs," and other related motivational concepts of the early twentieth century were similar in their emphasis on the general organismic and "energetic" character of behavior. Motivational theories of the period explained how the energy provided by the drive, instinct, or need combined with the organism's skill in a certain behavior and the relative value of the behavior's reward elicit and guide action (Hull, 1952).
The inclusion of cognitive processes, such as expectancies for success and perceptions of its value was added to the major theories of achievement motivation of the middle of the twentieth century. Arguably, the most notable among these theories was that of David McClelland, John Atkinson, and their colleagues (McClelland, Atkinson, Clark & Lowell, 1953). For these researchers, achievement motivation was based in a personality characteristic that manifested as a predisposed need to improve and perform well according to a certain standard of excellence. This achievement motive, which the researchers labeled n-Achievement, or nAch, was believed to form during the first years of life through parents' child-rearing practices: primarily, how early parents expected and rewarded, either tangibly or affectively with warmth and affection, independence in their children. McClelland and his colleagues hypothesized that these early experiences led to the propensity to experience a strong emotional arousal when cues in the environment were interpreted as an opportunity to achieve. Individuals were thought to differ from each other in the strength of this arousal and in the breadth of cues that elicited it.
Self-Determination Theory
Self-determination theory (SDT) is a broad motivational theory that addresses the issue of individuals achieving desired outcomes and personally working though setbacks. SDT assumes that humans are naturally active and have the tendency to interact with the environment in order to achieve a desired outcome or goal. From birth onward, humans, in their healthiest states, have the propensity to act on the environment, minimizing the need for external incentives. This innate motivational tendency is hypothesized to be a critical element in cognitive, social, and physical development because it is through acting on one's inherent interests that one grows in knowledge and skills. The inclinations to be curious of something new, to actively assimilate the experience, and to creatively apply new skills learned is not limited to childhood, but is a significant feature of human nature that affects performance, persistence, and well-being across the lifespan (Ryan and Deci, 2000; Ryan & LaGuardia, 2000).
Under SDT, different types of motivation-namely, intrinsic motivation, extrinsic motivation and amotivation-are clearly distinguished. Behavior can be brought about through intrinsic motivation (pleasure and interest-related motives), extrinsic motivation (instrumental motives), and amotivation (an absence of motivation; Deci & Ryan, 1985, 2002).
When individuals experience intrinsic motivation, they engage in behaviors they perceive as inherently interesting, satisfying, enjoyable, and absorbing (Deci & Ryan, 1985). Intrinsically motivated individuals will move, act, or direct their energy and effort into tasks for the enjoyment of the experience or the challenge involved. They are less moved by the external factors such as rewards or pressure. Intrinsic motivation helps individuals to seek challenges, discover new perspectives, stretch their capacities, and express their talents to become more self-determined individuals who have actualized their human potentials (Deci & Ryan, 2002).
With extrinsic motivation, individuals engage in behaviors because of the external outcomes they produce, such as praise or rewards. Unfortunately, in the long run, extrinsic motivation may cause individuals to perform and direct their energy and effort into actions based primarily on external factors and pressure that could cause resentment, resistance, and disinterest in a task (Deci, 1971; Lepper, Greene, Nisbett, 1973; Deci, Koestner, and Ryan, 1999). Originally, the intrinsic and extrinsic forms of motivation were hypothesized to lead to similar increases in the same behaviors (Atkinson, 1964). However, an extensive array of studies has shown that extrinsic rewards over time, such as deadlines (Amabile, DeJong, & Lepper, 1976) and surveillance (Plant & Ryan, 1985) tend to lower intrinsic motivation for engaging in certain behaviors (Deci, Koestner, & Ryan, 1999).
Those who experience amotivation cannot predict the consequences of their behavior, nor can they see the motive behind it. They may feel disintegrated or detached from their actions and will thus invest little effort or energy into a task. In other words, amotivation is when individuals experience a relative absence of motivation.
Basic Psychological Needs
SDT theorizes that three essential psychological needs leading to self-determination are competence, relatedness, and autonomy. In theory, meeting these three needs can improve intrinsic motivation (Deci and Ryan 1985; Ryan and Deci, 2000; Ryan, Kuhl & Deci, 1997). Intrinsic motivation that arises from meeting these basic psychological needs has been associated with various positive outcomes, such as increase in cognitive flexibility, conceptual understanding, and information processing (Grolnick & Ryan, 1987), as well as enhancing academic performance and academic self-concept (Deci, Vallerand, Pelletier, & Ryan, 1991; Reeve, Bolt, & Cai, 1999).
SDT hypothesizes that interpersonal events and structures in a social setting (e.g., rewards, communications, feedback) contribute towards meeting basic psychological needs of feeling competent. Individuals tend to seek challenges and take initiative to improve skills in order fulfill the need for competence. Intrinsic motivation is enhanced when social and contextual activities provide an environment that supports feelings of competence.
SDT also hypothesizes that secure attachments encourage feelings of belonging or the need for relatedness (Ryan & Deci, 2000). Relatedness is the need to experience oneself as connected to others and having affirmations of security to produce exploratory behavior without feelings of threats. In infancy, exploratory behavior, which indicates that the presence of intrinsic motivation (e.g., Bowlby, 1979; Frodi, Bridges & Grolnick, 1985), is more evident when the infant is securely attached.
The third basic principle of SDT is that every individual has the need for autonomy. Autonomous behavior or action is, in theory, the result of individuals' objectives. They also want to experience their "self" as the emanating source of their own actions and behaviors (deCharms, 1968; Deci & Ryan, 1985). Individuals, who experience the freedom of making personal choices and actions, fulfill their need for autonomy and demonstrate higher levels of intrinsic motivation (Frederick-Recascino, 2002).
SDT argues that individuals will desire outcomes when they are able to fulfill their basic psychological needs (i.e. autonomy; Deci, Vallerand, Pelletier, & Ryan, 1991). Furthermore, SDT posits that prospects of rewards, punishment, competition, and imposed goals generally undermine intrinsic motivation in activities that are uninteresting and reduce an individual's sense of autonomy (Deci et al., 1991). It follows, then, that classroom environments supporting students' feeling of autonomy may be related to learning and motivation (Reeve & Jang, 2006; Stefanou, Perencevich, DiCintio & Turner, 2004). Thus, self-determination theory suggests a possible mechanism underlying the potential success on students' learning.
Autonomy Support
Autonomy supportive environments help to facilitate intrinsic motivation. Grolnick, Deci & Ryan (1997) and Ryan, Sheldon, Kasser, & Deci (1996) state that the cluster of autonomy-supportive actions includes behaviors such as providing choice, encouraging self-initiation, minimizing the use of controls, and acknowledging the other's perspective and feelings. Opportunities to experience oneself as autonomous, and thereby increase intrinsic motivation, are facilitated by environments that are autonomy supportive (Deci & Ryan, 1987, 1991; Grolnick & Ryan, 1989; Reeve, Bolt & Cai, 1999). Autonomy supportive environments are characterized by opportunities that allow individuals the freedom of expression and action, support them in their choices, and uphold identity development (Deci and Ryan, 1987, 1991; Ryan and Solky, 1996). Importantly, Skinner & Edge (1998) state that autonomy support fosters individuals to actively discover, explore, and articulate their views, goals, and preferences in decision-making and problem-solving processes (Deci, Koestner & Ryan, 1999; Zuckerman, Porac, Lathin, Smith, & Deci, 1978). The effects of autonomy support on intrinsic motivation have focused on autonomy versus control (Deci, 1971; Lepper, Greene, & Nisbett, 1973). A meta-analysis by Deci, Koestner, & Ryan (1999) found almost every type of reward made contingent on task performance thwarts intrinsic motivation. Research also suggests that threats, deadlines, directives, and competitive pressures (such as those faced by students to do well in examinations) diminish intrinsic motivation because they leave individuals feeling controlled (e.g. manipulated), thereby diminishing their sense of autonomy (Amabile, DeJong, & Lepper, 1976; Deci & Cascio, 1972; Koestner, Ryan, Bernieri, & Holt, 1984; Reeve & Deci, 1996).
SDT posits that autonomy support in classrooms can enhance students' learning, ultimately leading to higher levels of achievement and skill development (Deci & Ryan 1985; Lepper & Green, 1975; Reeve, 2002; Ryan, 1982). Teachers who are more autonomy supportive are able to look at things from the learner's perspective, and allow opportunities for self-initiation and choice; they are able to provide a meaningful rationale if choice is constrained, refrain from the use of controlling behavior to motivate, and provide timely positive feedback (Deci et al., 1994, Reeve & Jang, 2006).
Research supports the importance of autonomy support in academic motivation and success (Deci & Ryan, 1991; Hardre & Reeve, 2003; Reeve, 2002). Reeve et al., (1999) discovered that students in autonomy supportive classrooms were more intrinsically motivated to do well in school, such as learning complex concepts better, demonstrating more conceptual and creative processes, producing higher levels of academic achievement (Benware & Deci, 1984; Grolnick & Ryan, 1987), having deeper and fuller learning experiences (i.e. understood how these facts fit together thematically vs. rote learning; Grolnick & Ryan, 1987), and were more likely to stay in school (Vallerand, Fortier & Guay, 1997).
Autonomy Support in Non-Western Contexts
Cross-cultural research (Markus & Kitayama,, 1991; Miller & Bersoff; 1992; Rudy, Sheldon, Awong & Tan, 2006) typically indicates that collectivists cultures see themselves as being more interdependent of each other, with shared interests, goals and values, and less independent (i.e. being distinct from others with divergent interests, goals and values). Children from collectivists cultures, are exposed to more interdependent norms in their respective societies (Miller, 2003). There is an implicit idea that autonomy and choice have strong associations with independence and when a society emphasizes interdependence, and then autonomy and choice are naturally less important (Markus & Kitayama, 2003).
In response to the critiques, Deci & Ryan (2000) maintain that all individuals have the need to experience autonomy. To support their claim, a full understanding of autonomy within SDT is needed. Deci & Ryan (2000) refer to autonomy within SDT as reflecting on intrapersonal and personal experience of "volition and choice" (Vansteenkiste, Zhou, Lens & Soenens, 2005:470) while the opposite of autonomy is the experience and feeling of being manipulated and controlled (Ryan & Lynch, 1989). Within the SDT, autonomous individuals can accept external guidance and support but they can also feel coerced. Therefore, autonomy does not carry the idea that individuals do not rely on others or are separate from others as suggested by Iyenger &DeVoe (2003) and Markus & Kitayama (2003).
There is emerging cross-cultural research that supports Ryan & Deci's claim that the need for autonomy is applicable to collectivists cultures. Sheldon, Elliot, Ryan, Chirkov, Kim, Wu, et al (2004) showed the subjective well-being of individuals in China, South Korea, Taiwan and the USA was positively predicted by higher autonomy. Other studies by Chirkov, Ryan, Kim & Kaplan (2003), Chirkov, Ryan & Willness (2005), Hayamizu (1997), Tanaka & Yamauchi (2000) and Yamauchi & Tanaka (1998) showed similar findings of higher levels of autonomy support among Russian, South Korean, Turkish, Brazilian, Chinese, Japanese, American, and Canadian individuals to be positive predictors of learning, intrinsic motivation, more optimal learning styles and academic achievement.
Although Deci & Ryan's (1985) self-determination theory has empirically supported the need for people to feel autonomous and being agents of their own actions, some cross-cultural theorists such as Brickman & Miller (2001), Markus & Kitayama (2003), Oishi (2000) and Oishi & Diener (2001) have argued that autonomy is not a universal need and is only useful when predicting and understanding the dynamics of motivation and achievement among Western cultures. They have argued that the need to feel autonomous is less valued in Eastern cultures where societal structures value conformity and interdependence (Vansteenkiste, Zhou, Lens & Soenens, 2005) Lyengar & Lepper (1999) conducted series of studies, which suggested that individual choice as displayed in autonomous decisions were deemed less important to individuals from collective cultures; the Asian American individual persisted longer at task they and their mothers chose for them and less when an outsider chose. The Anglo American individuals instead persisted longer only when they personally chose a task themselves. This, Lyengar & Lepper (1999) argued, demonstrated that autonomy was less relevant to Asian American individuals than Anglo American ones. Oyserman, Coon & Kemmelmeier (2002) found similar results when studying Brazilians, who emphasized collectivism, than Americans, who emphasized individualism.
Due to the opposing thoughts on the need for collectivists cultures for autonomy support, it is suggested that a look into how relevant autonomy support is in the Malaysian classroom context that is very much collectivist in nature. This study will help teachers to evaluate to what extent students are influenced or not influenced by autonomy support.
Further, to assist the Malaysian government's task of preparing individuals to be more competitive in the global economy (i.e. well-rounded individuals who are able to achieve high academic standards and be equally adaptive and thrive in the new working environment), the need to investigate whether SDT's claim that autonomy support can foster intrinsic motivation and improve academic achievement in the Malaysian context, is worth investigating.
Autonomy Support in the Malaysian Education Context
In my observations as a teacher in the Malaysian education context for twelve years, I believe the SDT can be helpful to teachers, in the Malaysian education context, who place importance on trying to keep students focused on achieving high scores in standardized examinations. A move towards increasing students' intrinsic motivation through the means of autonomy may widen teachers' options. As observed, the traditional model of classroom typically used in Malaysia emphasizes lecture-style instruction that involves mostly one-way transfer of information, with homework being the usual vessel of getting students motivated in their learning (Lau & Ellias, 2011). The Malaysian educational context has a more controlling atmosphere where desired behaviors or outcomes are brought about by various methods of explicit rewards (e.g. demerit points, seeing the discipline teacher), which I suggest, can undermine intrinsic motivation. For example, the students remain as passive and obedient recipients of knowledge and information and play little part in the learning process (Neo & Neo, 2003).
In the Malaysian classroom context, there is sense that teachers need to be in control of the learning environment, activities, and behaviors of the students. The teachers also face demands and pressures from administrators and parents to ensure students are performing well (e.g. achieving A's) in high-stakes standardized examinations and school ranking, which can lead teachers toward feeling a need to be more explicitly in control of the classroom environment. This idea of providing autonomy support for students is new for Malaysian teachers. Research in other countries has shown that autonomy supportive classrooms can be helpful in maintaining or enhancing intrinsic motivation in learners. However, the effects of autonomy supportive classrooms on intrinsic motivation and academic achievement have not been studied in Malaysia. Therefore, I argue that Malaysian teachers in general could benefit from seeing how support for students' autonomy may hold promise for enhancing students' achievement.
Method
This study describes autonomy support in Malaysian classrooms and students' perception of autonomy support. This study evaluated whether there was an association between autonomy support and both intrinsic motivation and student academic achievement. It was hypothesized that students who perceive greater autonomy support in their classrooms would have higher levels of intrinsic motivation and higher levels of academic achievement. Questionnaires were administered to students.
Participants
Participants in this study consist of sixteen-year-old students from two high schools (Convent Bukit Nanas High School (CBN) and Bandar Sri Permaisuri High School (SMP)) in Kuala Lumpur, Malaysia. There were 46 and 98 students from each school respectively. The former is an all-girls school in the city with students who are generally more academically inclined. The latter is a co-educational school in the suburbs with generally less academically inclined students. Malaysia has a strong English language background as it was previously colonized by the British. This convenience sampling included 144 students (CBN, n = 146, SMP, n = 98). The students were asked to complete questionnaires. All participants were included in the sample. All students are proficient in the English language as English is learned as a second language alongside with the Malay language from pre-school to university.
Procedures
The questionnaires (Appendix A and B) were distributed to both schools respectively. Students were informed of their rights as research participants and that their participation would not affect their grades in school. Students were also informed that their responses would be held completely confidential and their teachers would not have access to students' responses. Students' assent of participation was obtained. Students completed questionnaires and placed completed questionnaires into an envelope. The last student to complete the questionnaire sealed the envelope.
Instruments
Learning Climate Questionnaire (LCQ). Perceived autonomy support was measured using the LCQ that was adapted by Williams and Deci (1996) from the Health-Care Climate Questionnaire (Williams, Grow, Freedman, Ryan, & Deci, 1996). There were 15 items in the measure that are answered on a seven-point Likert scale (ranging from 1 = 'strongly disagree' to 7 = 'strongly agree'), which indicated the degree to which students perceived autonomy support. A sample item was, 'I feel that my teachers provide me with choices and options.' The LCQ has a single underlying factor and in previous studies has had a high internal consistency with an alpha of 0.93 (Black & Deci, 1999). Williams & Deci (1996) has also validated the 15 items on the LCQ in a study of medical students who experienced more autonomy supportive environment experienced an increase interest and perceived competence. They found that across domains, the alpha coefficient of internal consistency is virtually always above 0.90. Scores on the questionnaire were calculated by averaging the individual item scores. Higher average scores represented a higher level of perceived autonomy support.
Intrinsic Motivation Inventory (IMI). The IMI (Ryan, 1982) assessed students' subject experience of IM and characteristics in a setting that promote IM for a target activity in the classroom. It has been used in several experiments related to intrinsic motivation and self-regulation (e.g., Ryan, Koestner & Deci, 1991) and in published research it has an internal reliability of an alpha 0.85 (McAuley, Duncan & Tammen,1989). The IMI consisted of seven domains related to IM; one of these domains was considered a direct measure of IM while the remaining assess constructs related to it. For this study, three domain scales were used: IM, perceived choice (in activity), and perceived anxiety. (Perceived choice is closely related to Ryan and Deci's construct, autonomy support, described above).
The IMI consists of varied numbers of items from these subscales, all of which had been shown to be factor analytically coherent and stable across a variety of tasks, conditions, and settings. The general criteria for inclusion of items on subscales have been a factor loading of at least 0.6 on the appropriate subscale, and no cross loadings above 0.4. Typically, loadings substantially exceed these criteria. The incremental R for every item above 4 for any given factor is quite small. Still, it is very important to recognize that multiple item subscales consistently outperform single items for obvious reasons, and they have better external validity.
There were 19 items in the measure that were answered on a seven-point Likert scale (ranging from 1 = 'not at all true' to 7 = 'very true'), which indicated the degree to which students agree to a statement. A sample statement was, 'I enjoy doing the classroom activities very much.' Items for which and (R) is shown were reversed scored. Then, scores on the each subscale were calculated by averaging across all of the items on the subscales. A higher score indicates more of the concept described in the subscale name.
Academic Achievement. Students' academic achievement was based on their voluntary responses to what grades generally describe them (Grades). The sample statement was, 'Which of the following best describes your grades in school?'
Patterns of Adaptive Learning Scales (PALS). As much as autonomy support and intrinsic motivation have been argued to positively affect academic achievement, the possibility of students' goal orientation playing a role in students' intrinsic motivation and academic achievement could also be important (Guskey & Gates, 1986; Ritchie & Thorkildsen, 1994 and Whiting & Render, 1984). Goal orientation represents personal disposition to pursue either learning (i.e. Goal Mastery Learning) or performance goal orientations (Performance Approach and Performance Avoidance) in achievement situations (Dweck, 1999). A learning goal orientation is associated with the belief that ability can be developed; in theory students are intrinsically motivated to increase their mastery and competence over challenging situations to pursue a set goal (Wentzel 1991, 1993). The focus of attention is on the intrinsic value of learning (Nicholls, 1984) as they master and understand the content and find enjoyment and interest in the actual act of learning (Brophy, 1983).
In contrast, Ames (1992) states that performance goal orientation is associated with the belief that ability is fixed, and difficult to develop and individuals are motivated to establish the adequacy of their ability in the eyes of others and to avoid situations where they may appear inadequate (i.e. performance approach and performance avoidance). Individuals who adopt performance approach goal orientation work towards a goal with the intention of looking good to others while those who adopt performance avoidance goal orientation tend to be more concerned with drawing as little attention as possible from others (Ames, 1992).
Both mastery and performance goal orientation allow students to reach academic achievement goals but like intrinsic motivation, research suggests that mastery goal is associated with motivation-related variables that promote positive achievement activity. Ames & Archer (1988) linked mastery goals to personal beliefs that effort lead to success. There is pride and satisfaction from the effort put into a task.
However, research also indicates that both types of goal orientations have their place in students' learning and academic achievement (Wentzel 1991, 1993; Wolters, Yu & Pintrich, 1996). Harackiewicz and Sansone (1991) indicate that in specific situations performance goals can also promote the development of competences; and acknowledge the positive effects of performance goals. It has also been pointed out that the different goal orientations do not necessarily need to be treated as opposites. For example, Meece and Holt (1993) found that students could be high in mastery motivation and also high in performance orientation, while others could be low in both dimensions. From this viewpoint, achievement goals are seen as complementary and it is acknowledged that students can pursue a mastery, performance, or work-avoidance orientation simultaneously (e.g., Valle et al., 2003).
Looking into students' achievement goals can add valuable insights into differing ways students engage in, evaluate, and perform in academic learning, as well as consider the additional influences of issues affecting academic achievement.
PALS examines the relation between the learning environment and students' motivation, affect and behavior. The student scale used in this research is focused on students' personal achievement goal orientation. There are 14 items in the measure that are answered on a seven-point Likert scale (ranging from 1= 'strongly disagree' to 7= 'strongly agree'). The scales are based on research showing that a differential emphasis on "mastery" and "performance" goals is associated with adaptive or maladaptive patterns of learning (e.g., Ames, 1992; Dweck, 1986). In addition, recent evidence (Elliot & Harackiewicz, 1996; Middleton & Midgley, 1996; Skaalvik, 1997) suggests that a performance goal orientation can be conceptualized in terms of both approach and avoidance components. Therefore, the goal scales for personal goal orientations include not only mastery and performance goals but also to differentiate between performance-approach and performance-avoid dimensions.
To validate the use of the revised personal goal scales, a confirmatory factor analysis on the 14 personal goal orientation items to examine the factor structure of the three sets of items (mastery, performance-approach, and performance-avoid) were conducted (Midgley, Maehr, Hruda, Anderman, Anderman, Freeman et.al.). LISREL VIII (Joreskog & Sorbom, 1993) confirmed the expected model. Goodness of fit indices suggested that the model fits the data well (GFI = 0.97, AGFI = 0.95). Specifically, personal mastery, performance-approach, and performance-avoid goals all loaded on different latent factors. To validate the use of the classroom goal structure scales, confirmatory factor analysis on the mastery goal structure, performance-approach goal structure, and performance-avoid goal structure items to examine the factor structure, was conducted. LISREL VIII (Joreskog & Sorbom, 1993) confirmed that the items loaded on different latent factors (GFI = 0.96, AGFI = 0.94).
Plan of Analysis
Regression analyses were carried out for the purpose of this study. The first step studied the associations between the independent variable (IV) perceived autonomy support (AS) in the dependent variable (DV) intrinsic motivation(IM), controlling for gender and type of school. The IV and control variables were entered in two blocks in a hierarchical regression to separate out the effects of the control variables. Step two involved a simple regression analysis of perceived autonomy support (IV) on the DV, academic achievement (self-reported grades), still controlling for gender and types of school. The third step used a simple regression analysis to assess the association between IM and self-reported grades (Grades). The final step was a multiple regression analysis with autonomy support and intrinsic motivation predicting academic achievement. Intrinsic motivation and perceived autonomy support could either be an independent or dependent variable. IM and AS were treated as IV's when making associations between them and grades. To determine associations between IM and goal orientations (goal mastery, performance approach and performance avoidance), IM was again treated as an IV.
Results
Preliminary Analyses
Preliminary analysis of the survey items, including reliability of composite measures and descriptive statistics, was conducted. There were no missing data. Reliability coefficients and descriptive statistics with the adjusted scales for 144 participants are displayed in Table 1.
Reliability Analysis
The criteria for the composite scale reliability for each construct had to meet the recommended minimum cutoff of 0.70 (Nunnally, 1978). With the exception of mastery goals (MG), perceived autonomy support (AS), and intrinsic motivation (IM), this criterion led to the adjustment of the remaining composite scales. The reliability of the performance approach (PerfAppr) Goals scale with all five of the original items was a = .67. Reliability analysis indicated that inclusion of item-4, One of my goals is to looks smart in comparison to the other students in my class, decreased the internal consistency of the scale, thus it was dropped leading to a final reliability with four items was a = .71. The reliability of the performance avoidance (PerfAvoid) Goals scale with all four the original items was a = .63. Analysis indicated that item-7, One of the goals is to keep others from thinking I m not smart in class, substantially decreased the reliability of the scale. With item-7 removed, the internal consistency for the performance avoidance goals scale with three items was a = .74.
For perceived choice (PC), the two positively worded items (Item 1, "I believe I have some choice about doing the activities" and Item-6, "I do the activities because I have no choice") were dropped and only the five reversed items were retained to increase the reliability from .60 to .68. Even with this consideration, the minimum criterion for composite scale reliability was insufficient and the measure was dropped. Furthermore, this measure was not necessary for the analysis with the availability and reliability of the perceived autonomy support measure. Item-GPA7, " One of my goals is to keep others from thinking I'm not smart in class" in the performance avoidance goals measure was dropped as the double negative of the item structure reduced the reliability to 0.64 instead of 0.74. The perceived anxiety measure that had two positively worded and two negatively worded items was dropped as the overall reliability was too low (a = .555) and creating two-item scales one for positively and one for negatively worded items was deemed conceptually and statistically inappropriate.
Descriptive Statistics
The mean score for self-reported grades, (M = 4.26) showed students had average grades of "B's" and "C's" and "C's". Mean scores for performance approach goals, performance avoidance goals and perceived autonomy support were located close the mid-point of the 1-7 response scale (e.g., M = 4.13). mastery goal and intrinsic motivation, however, showed higher mean scores (M = 6.20 and M = 5.19, respectively).
The correlations among the study variables were mainly consistent with hypothesized relations in the literature. Seven of the 15 correlations were statistically significant with at a < .05. Four of the seven significant correlations were in the expected positive direction. For example the correlation between intrinsic motivation and mastery goals was r = .383, indicating that higher self-reported IM was associated with higher self-reported mastery goal, and vice versa. Three of the correlations were in a direction opposite of what we expected. These were between grades, and mastery goal, IM and perceived choice, indicating that higher self-reported grades were associated with lower MG, IM and PC. For example the correlation between grades and IM was r = -.321. The remaining 8 correlations which were in expected association directions were not statistically significant, consistent with literature.
Main Analyses
Gender was evaluated for inclusion in the regression models. Regression analyses were conducted with and without gender to determine if it has an impact on the pattern of findings. These analyses indicated it did not alter the pattern or significance of results and was thus dropped. Given that one of the two schools in the sample was an all-girls school, it was not surprising that adding what amounts to an additional measure for gender failed to influence the results. The type of school (0 = CBN, 1 = SMP) was included in all models. Results will be presented by hypothesis.
Hypothesis 1. Students' intrinsic motivation will be positively related to their perceived autonomy support. For Hypothesis 1, regression analyses in Model A (Table 2) showed that the type of school was associated with students' intrinsic motivation, F (1,142) = 18.570, p < .001. Type of school explained 11.6% of the variance in students' intrinsic motivation and, based on the regression coefficient for school type (CBN = 0, SMP = 1), students at SMP reported higher intrinsic motivation, β =.711, bootstrap CI = .358 - 1.043. Model B added students' perceived autonomy support in the classroom to Model A. Model B explained a significant and additional amount of variance in students intrinsic motivation, 27.1%, F (2,141) = 61.88, p < .001. The regression coefficient indicated that, controlling for type of school, higher perceived autonomy support in the classroom was associated with higher intrinsic motivation, β = .565, bootstrap CI =.430 - .701, and vice versa. The type of school coefficient remained significant and in same direction, β = .864, bootstrap CI = .580 - 1.148. The final model, Model B, explained a total of 37.8% (adjusted R2) of the variance in students' intrinsic motivation.
Hypothesis 2. Students' self-reported grade will be positively related to their level of intrinsic motivation. For Hypothesis 2, regression analyses in Model A (Table 3) indicated that the type of schools was associated with students' self-reported grades, F (1,142) = 741.437, p < .001. It explained 83.9% of the variance in students' self-reported grades and, based on the regression coefficient for school type (CBN = 0, SMP = 1), students at CBN self-reported higher grades, β = -3.861, bootstrap CI = -4.146 -3.571, than students at SMP. Model B failed to explain a significant amount of variance in students self-reported grades. That is, intrinsic motivation was not significantly associated with self-reported grades, controlling for school type. The final model then explained a total of 83.7% (adjusted R2) of the variance in students' grades.
Hypothesis 3. Students' self-reported grade will be positively related to their perceived autonomy support. For Hypothesis 3, regression analyses in Model A (Table 4) again indicated that the type of schools was associated with students' self-reported grades, F (1,142) = 741.437, p < .001. It explained 83.9% of the variance in students' self-reported grades and, based on the regression coefficient for school type (CBN = 0, SMP = 1), again students at CBN self-reported higher grades, β = -3.861, bootstrap CI = -4.146 -3.571. Model B failed to explain a significant amount of variance in students self-reported grades. Intrinsic motivation was not significantly associated with self-reported grades (0.2%), controlling for school type. The final model then explained a total of 83.7% (adjusted R2) of the variance in students' grades.
Exploratory Analysis 1. Students' intrinsic motivation will be associated with mastery goal orientation. For Exploratory Analysis 1, regression analyses in Model A (Table 5) showed that the type of school was not associated with students' intrinsic motivation, F (1,142) = 10.145, p = .002. Type of school explained 6.7% of the variance in students' intrinsic motivation and, based on the regression coefficient for school type (CBN = 0, SMP = 1), students at SMP reported higher mastery goal orientation, β = .419, bootstrap CI = .181 - .662. Model B explained a 9.8% amount of variance, F(2, 141) = 10.145, p = .002. The regression coefficient for indicated that, controlling for type of school, higher intrinsic motivation was associated with higher mastery goal orientation among students, β = .259, bootstrap CI = .134 - .395, and vice versa. The type of school coefficient remained non-significant, β =.235, bootstrap CI = -.033 - .486. The final model, Model B, explained a total of 15.3% (adjusted R2) of the variance in students' mastery goal orientation.
Exploratory Analysis 2. Students' intrinsic motivation will be associated with performance approach goal orientation. Regression analyses for Exploratory Analysis 2 (Table 6), in Model A and Model B, showed no significant association between type of school and intrinsic motivation on performance approach goal orientation, explaining only 1.7% and 2% of the variance respectively, F(1,142) = 2.390, p = .124 and F(2, 141) = .555, p = 1.469 respectively. Students' performance approach goal orientation, based on the regression coefficient for school type (CBN = 0, SMP =1), students at CBN had higher performance approach goal orientations, β = -.093, bootstrap CI = -.396 - .201. The final model explained a total of 7% (adjusted R2) of the variance in students' performance approach goal orientation.
Exploratory Analysis 3. Students' intrinsic motivation will be associated with performance avoidance goal orientation. For this analysis, regression analyses in Model A (Table 7) indicated that the type of schools was not associated with students' performance avoidance goal orientation, F (1,142) = .009, p = .918. It explained none of the variance in students' performance avoidance goal orientation and, based on the regression coefficient for school type (CBN = 0, SMP = 1), students at CBN had higher performance avoidance goal orientations, β = -.093, bootstrap CI = -.438 - .220. The final model explained a total of 11% (adjusted R2) of the variance in students' performance avoidance goal orientation.
Separate follow-up regression analyses were carried out to determine associations between individual schools on the variables. The results indicated no change. Regression analyses with both IM and AS included together in one block in Model B on the various variables also indicated no significant associations.
Discussion
The purpose of this study was to evaluate the associations between students' perceived autonomy support (AS), their level of intrinsic motivation (IM), and academic achievement (grades) in Malaysian classrooms. This study was based on Deci &Ryan's (1985) self-determination theory and posits that perceived autonomy support may influence students' intrinsic motivation in the Malaysian education context and indirectly their academic achievement. In general, some findings were consistent with theory and the hypothesized relations while others were not.
Intrinsic Motivation, Autonomy Support, & Academic Achievement
As hypothesized (Hypothesis 1), there was a strong, positive correlation between perceived autonomy support (AS) and intrinsic motivation (IM), which was consistent with literature findings (Grolnick, Deci & Ryan (1997) and Ryan, Sheldon, Kasser, & Deci (1996)). Students' who perceived greater AS were more likely to also report higher levels of IM. However, there were no significant correlations for Hypothesis 2 and 3, between IM and grades, and between AS and grades. These findings were unexpected as the research literature has generally found that both IM and AS have positive associations with academic indicators (Grolnick & Ryan, 1987, Deci & Ryan, 1991; Hardre & Reeve, 2003; Reeve, 2002). There could be several reasons for a lack of a significant association. First, the students in this sample, like most students in the Malaysian education system, have been educated in traditional classrooms that are more teacher-directed (e.g. teachers instructing and directing students' learning activities) instead of providing more individual autonomy. Second, the educational emphasis in Malaysian schools and societal expectations is on exam performance, which theory suggests would limit both IM and AS (Koestner, Ryan, Bernieri, & Holt, 1984; Reeve & Deci, 1996).
The positive association between IM and AS suggests that the Malaysian educational context may not presently support the linkage between IM/AS and academic performance but that the IM-AS relation could have relevance in other settings of student's lives. The literature that supports the idea of AS supporting the development of IM in individuals is important in the Malaysian educational context. Malaysian students at the school level can be groomed to develop competitive thinking skills in order for them to further develop their overall ability to think competitively and globally. The challenges they face in the new global working environment can only be supported by new skills that are supported by an autonomy supportive environment (Scott & Bruce, 1994; Kane, Berryman, Goslin & Meltzer, 1990). The high school classroom is an ideal environment in which Malaysian students can benefit from IM-AS relations. Therefore the significance of AS and its relation to IM can help Malaysian students discover, explore and invent new ideas, as well as sustain effort to prepare themselves to keep up with the continually changing work environment.
The lack of association for IM-Grades and AS-Grades in the Malaysian education context could be explained with better academic indicators such as the national examination scores (Sijil Pelajaran Malaysia). However, it is proposed that a Confirmatory Factor Analysis (CFA) be performed on the constructs of IM and AS to better understand the nature of the latent constructs in the Malaysian context. CFA of a similar but larger sample, and even between western and non-western samples could test whether the relationship between AS and IM and their underlying latent constructs exists. Performing a CFA that included western and non-western samples could provide a better understanding of whether cultural perspectives influenced the data gathered from the constructs. Any evidence of degrees of misfit in original western-based samples could also be identified and explained. However, in the present study, a CFA was not conducted as it had been assumed that the measures used were valid measurements that measured the constructs intended to be measured.
The lack of a relation between IM/AS and grades could also be attributed to issues of measurement. An autonomy support measurement that looked into teachers' perspective could also be included to explore associations between students' perceived AS and teachers' perceived provision of AS. Further, the use of self-reported grades, for example, may not provide an adequate representation of students' academic achievement or performance. Thus, future research would need to include additional measures of students' academic achievement and performance. Because of the value placed on exam performance in Malaysia, it would be important to have indicators that do not solely rely on self-report in order to rule out associations due to shared method variance. One possible measure could be scores from standardized examinations.
However, the importance of AS-IM in this study, go beyond academic performance but also to help teachers and students understand the importance of AS and IM. AS and IM can be important aspects in helping students to develop better learning strategies, which in turn can help them be more active participants of the education process.
Goal Orientations
As the study relied solely on self-reported grades, the researcher considered the possibility of an additional indicator, in the form of students' goal and performance orientations that was easily obtained from students. Therefore an exploration of the relation between AS and IM with students' goal orientations was included. Consistent with prior literature, the results indicate higher IM being associated with higher mastery goal orientation (MG) (Roeser et al., 1996). A possible explanation would be that intrinsically motivated students tend to develop self-competence and the inclination to contrast their present stage of success with previous success of their own (Pintrich, 2000a). Also, it is possible that students who adopt mastery goals cultivated a self-regulated type of learning, which could be intrinsically motivated. They seek information in order to acquire, develop, and refine their knowledge and skill (Butler, 2000). This mastery goal orientation as a possible result of being intrinsically motivated may help them to continue improving themselves.
Where association was noted between IM and mastery goal orientation, it is interesting to note that students in SMP (less academically inclined students) and not CBN (more academically inclined students) had higher levels of intrinsic motivation and more were more inclined to have mastery goal orientations. The background and socio-economic status of the SMP students suggest an internal drive to want to perform well in school in order to do well academically and break the cycle of poverty. In spite of SMP students' lower levels of achievement, the association between IM and MG seems to indicate that that these students value learning and are determined to learn and improve their competencies (Suarez, Cabanach, and Valle, 2001).
The findings between IM and performance goal orientations (PG) showed no significant associations. Students, especially those from CBN, who had higher self-reported grades, were more performance approach and performance avoidance goal oriented. When considering how critical a role high-stakes testing has in high-performing schools, this was not a surprise. The lack of association between IM and PG was opposite to prior research findings that posited that students who have high IM tend to be more mastery oriented and less performance oriented, as these students are more concerned about the actual learning process and personal capabilities rather than academic scores (Valle, Cabanach, Rodríguez, Nunez & Solano, 2007).
The lack of association between IM and Performance orientation could be explained in the Malaysian context seeing how exam-oriented students, teachers, parents and society's associations may cause students to be more performance oriented.
Research on the role of goals at the academic level coincide in noting that learning goals are beneficial for most learning-related results, including results at a motivational level, such as self-efficacy, interest, and value (Harackiewicz, Barron, Tauer, Carter, and Elliot, 2000) but it cannot be concluded that PG do not hold any advantages as empirical results about the benefits or disadvantages of performance goals are fairly controversial, some authors consider that their effects on motivation and learning are more complex and require more theoretical and empirical attention (Utman, 1997). Furthermore, the distinction between performance approach goals and performance-avoidance goals has led, among other things, to the reconsideration of the effects of performance-approach goals. Pintrich (2000b) noted that there may be situations in which performance goals are not disadvantageous to motivational and affective terms and in terms of the use of strategies and achievement. In fact, in some works, performance-approach goals seem more closely linked to achieving goals, whereas learning goals are more related to intrinsic interest in the tasks (Harackiewicz et al., 2000).
Therefore, it is suggested that a more detailed study into how the two goal orientations can help Malaysian students should be carried out. Instead of adopting one exclusive goal, Malaysian students may choose various goals to become learn better. Thus, when facing situations in which the learning activity is not very stimulating or interesting, it might be useful to find reasons other than intrinsic interest in the task to motivate their actions. In these cases, the opportunity to choose various motives such as getting others' approval (performance approach) to be a better incentive to maintain academic engagement ((Pintrich, 2000b).
Conclusion
Although there may be cultural and institutional variations on how the needs for autonomy, IM and performance are supported, satisfied, and expressed, the literature supporting the importance of AS, IM and even goal orientations in Western samples continues to indicate that non-western samples, specifically, Malaysians could benefit from them.
Intrinsically motivated students who rely on self-reward and mastery goals can be crucial elements to the learning process and to the experience of well-being. Autonomy supportive classroom environments will support Malaysian students to take a more active role in their learning. It will also allow students to be more responsible for their own learning. It is hoped that when Malaysian classrooms become more autonomy supportive, the learning environment will be more conducive and students will have more meaningful learning experienced. There will be more opportunity to think and work independently and come up with creative problem-solving solutions. It is hope too that teachers and students are more able to fully attend to and grasp the importance of AS-IM for their teaching and learning.
An optimal Malaysian system of education would support both students' autonomy, and one which its evaluative system is able to equally support this move, thereby aiding students become more intrinsically motivated to learn.
