Teaching and Teacher Education 142 (2024) 104516

Contents lists available at ScienceDirect

Teaching and Teacher Education

journal homepage: www.elsevier.com/locate/tate

Research paper

Perceived autonomy support as a predictor of rural students’ academic

buoyancy and academic self-efficacy
Andrew A. Kingsford-Smith a, Dennis Alonzo a, Kim Beswick a, *, Tony Loughland a,
Philip Roberts b
a
School of Education, University of New South Wales (UNSW), NSW, 2052, Sydney, Australia
b
Faculty of Education, University of Canberra, ACT, 2601, Canberra, Australia

A B S T R A C T

Students’ academic self-beliefs are associated with their school achievement and enjoyment. However, academic self-beliefs appear to be lower in rural schools. In a
sample of students in Australian rural schools (N = 974), this study investigated whether perceived autonomy support (PAS) predicted two important self-belief
constructs: academic buoyancy and academic self-efficacy. The results revealed that PAS positively predicted academic buoyancy and academic self-efficacy.
Multigroup structural equation modeling further identified that primary school students reported more adaptive school experiences than high school students.
This research has implications for how teachers can best support students’ academic self-beliefs in rural schools.

1. Introduction setting” (Martin & Marsh, 2008b, p. 172). This represents how well
students can manage the daily challenges of school, including experi­
In many countries there are significant gaps in the educational out­ encing failure and academic difficulties. On the other hand, academic
comes between students living in rural and metropolitan areas, self-efficacy refers to students’ confidence in their ability to perform
including academic achievement in standardized tests (Organisation for well at academic tasks (Chemers et al., 2001). This represents students’
Economic Co-operation and Development [OECD], 2018) and student self-beliefs about their ability to successfully complete work and as­
expectations of completing a university degree (Echazarra & Radinger, sessments. PAS may foster both academic buoyancy and academic
2019). Students’ academic self-beliefs are implicated in these dispar­ self-efficacy because it nurtures students’ psychological needs and
ities, as students’ self-beliefs are associated with their academic self-beliefs (Ryan & Deci, 2000; 2017).
achievement (Honicke & Broadbent, 2016; Talsma et al., 2018; Valen­ Autonomy support has received considerable supportive evidence
tine et al., 2004) and positive intentions to continue their education across the literature as an adaptive teaching approach (Reeve & Shin,
(Lamb et al., 2020; Martin, 2009). Students’ academic self-beliefs also 2020; Ryan & Deci, 2017). However, fewer studies have investigated
generally decrease the further remote a school (Thomson et al., 2020; PAS in rural educational contexts (Simões & Calheiros, 2023). The
Young, 2000). uniqueness of rural education has often been neglected in the wider
Teachers’ interpersonal motivating styles may offer ways to support literature (Roberts & Fuqua, 2021), which has led to criticisms of
students’ academic self-beliefs. Autonomy support is one such moti­ research as reinforcing a metropolitan norm (Roberts & Guenther, 2021)
vating style that focuses on nurturing students’ internal motivation, and tending to not consider the contextual specificity of this research
instead of relying on external pressures and rewards to influence student population (Roberts, 2014). However, PAS may be particularly impor­
behavior (Reeve, 2016). When students perceive autonomy support tant for nurturing academic self-beliefs in rural settings (Karlberg-­
from their teachers, they also experience greater internal motivation and Granlund, 2023). Some teachers in rural schools have reported that
engagement (Jang et al., 2016) and academic achievement (Schuitema national curriculums are perceived to be less relevant to both students
et al., 2016). Perceived autonomy support (PAS) is also thought to be and communities in rural areas (Lock et al., 2012; Papatraianou et al.,
predictive of self-belief constructs such as academic buoyancy (Tarbet­ 2018). Notably, one aspect of autonomy supportive teaching is
sky et al., 2017) and academic self-efficacy (Uçar & Sungur, 2017). communicating task relevance to students to help foster intrinsic moti­
Academic buoyancy refers to students’ perceived ability to “effectively vation (Reeve, 2016), which may help rural students to identify personal
deal with setback, challenge, adversity, and pressure in the academic value in their learning (Hardre & Reeve, 2003). Thus, the extent to

* Corresponding author. University of New South Wales (UNSW), School of Education, Morven Brown Building, Sydney, NSW, 2052, Australia.
E-mail address: [email protected] (K. Beswick).

https://doi.org/10.1016/j.tate.2024.104516
Received 28 August 2023; Received in revised form 29 January 2024; Accepted 3 February 2024
Available online 16 February 2024
0742-051X/© 2024 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
A.A. Kingsford-Smith et al. Teaching and Teacher Education 142 (2024) 104516

which PAS predicts academic buoyancy and academic self-efficacy in and academic self-efficacy (i.e., cognitive/personal factors).
rural contexts is an important area of research.
There is also a need to investigate potential differences in these factor 1.3. Autonomy support
associations for students in different school-stage settings, such as pri­
mary, high, and central schools. Central schools in New South Wales Teachers’ interactions with students are thought to be instrumental
(NSW), Australia serve students of both primary and high school ages, in the development of academic self-beliefs (Burns et al., 2018). Recent
and are more prevalent in rural and regional areas, where school sizes research has identified two distinct forms of motivating styles that
can be much smaller (Halsey, 2018). Some research has suggested that teachers use in the classroom that are implicated in students’ classroom
PAS is experienced at higher levels by younger students (Gillet et al., functioning and cognitive appraisals: controlling and autonomy sup­
2012), but it is not known if this motivating style plays a stronger role portive styles (Reeve, 2016). In controlling styles, the teacher makes
for student self-beliefs when students are younger and have fewer aca­ external motives more salient to influence student behavior, such as
demic experiences. Additionally, primary school students usually spend using demanding tones and using rewards and punishments like merits
more time with one main teacher throughout a school year, which may and detentions. In contrast, autonomy supportive teaching styles focus
implicate the role that PAS plays towards students’ self-beliefs. on nurturing students’ internal drives to engage in learning. This is done
The aim of the current research was to examine the extent to which through providing rationales for classwork, listening to students’ per­
PAS predicts academic buoyancy and academic self-efficacy, using data spectives, providing meaningful choices for how classwork is done, and
from a sample of students in rural schools in NSW, Australia. It was encouraging questions (Reeve & Shin, 2020). Perceived autonomy sup­
hypothesized that PAS would have strong positive associations with port (PAS) refers to students’ perceptions that their teachers support
academic buoyancy and academic self-efficacy. Further, it was hypoth­ their autonomy and self-determination. PAS may be particularly
esized that academic buoyancy would partially mediate the association important for students in rural schools because it encompasses identi­
between PAS and academic self-efficacy. This research also examined fying learning relevance (Hardre & Reeve, 2003; Karlberg-Granlund,
potential differences in these construct associations between students in 2023). Teachers in rural contexts in Australia have expressed difficulties
different school-stage settings. The sample included students in primary in needing to adapt national curriculum topics to be relevant to their
(ages 9–14), high (ages 12–17), and central schools (ages 10–17). This students (Lock et al., 2012; Papatraianou et al., 2018). If students do not
research is significant because it explored the importance of autonomy find their work relevant, then this can affect their motivation, engage­
support across distinct rural school-stage settings. ment, and achievement (Reeve, 2012).

1.1. Rural perspective 1.4. Academic buoyancy and the predictive role of perceived autonomy
support
In this research, rural contexts are understood as unique social set­
tings in which students’ academic buoyancy and academic self-efficacy Academic buoyancy refers to students’ ability to handle the day-to-
are formed and enacted. Such a focus may help better understand the day challenges they experience at school (Martin & Marsh, 2008b).
educational development of rural students, and better develop policies Academic buoyancy is related to but distinct from academic resilience,
and practices that can lift student achievement relative to their context which describes students’ resilience to more severe adversities, such as
(Roberts & Fuqua, 2021). Autonomy support may be particularly poverty (Das, 2019; Rojas Flórez, 2015). In contrast, academic buoyancy
effective in rural contexts, as it focuses on teachers nurturing students’ focuses on challenges and setbacks that most students frequently expe­
autonomous motivation to learn (Karlberg-Granlund, 2023). We rience at school, such as critical feedback, heavy school workloads, and
acknowledge that a potential limitation in the present study may be that difficult to understand concepts (Martin & Marsh, 2009). Academic
“academic study” is seen by educational policy and research as value buoyancy may be particularly relevant to students in rural settings
neutral, when instead it may be deeply rooted in context and family because these students generally experience lower academic achieve­
background. While testing such theories is beyond the scope of the ment (OECD, 2018) and thus may face academic setbacks more often.
current study, this study may gesture to some differences that are worth PAS may predict greater academic buoyancy, although this associ­
exploring from that perspective. Importantly, different levels of aca­ ation has yet to be empirically tested. Granziera et al. (2022) identified
demic buoyancy and academic self-efficacy should not be inadvertently that instrumental teacher support (i.e., the provision of instrumental
constructed as deficits when compared to a metropolitan norm, but resources and practical help) predicted greater academic buoyancy in
rather understood as potentially different and distinct in rural contexts. samples of both high school students in Singapore and primary school
students in Australia. PAS may play a similar role because of its focus on
1.2. Theoretical framework providing rationales and encouraging questions. Additionally, Tarbetsky
et al. (2017) hypothesized that PAS may nurture academic buoyancy by
Social cognitive theory (Bandura, 1986) was used as the theoretical fostering positive teacher-student relationships. PAS may also nurture
framework for the current study. Self-efficacy plays a pivotal role in academic buoyancy through helping students to focus on personal
social cognitive theory as a precursor of agentic behavior and optimal relevance in learning. When students engage in classwork for autono­
engagement, as people are more inclined to participate in activities in mous reasons, instead of external reasons, they are more likely to focus
which they feel they will experience success (Bandura, 2001). Bandura on mastering new skills (Benita et al., 2014) and may feel more resil­
defined self-efficacy as “people’s beliefs in their capabilities to produce ience because challenge is a part of the learning process (Dweck &
given attainments” (Bandura, 2006, p. 307). The origins of self-efficacy Master, 2007). Thus, in the current research, we hypothesized that PAS
beliefs are thought to reside within a triadic model of reciprocal in­ would predict greater academic buoyancy.
teractions between environmental, cognitive/personal, and behavioral A recent longitudinal study by Bostwick et al. (2022) questioned the
factors (Bandura & Jourden, 1991). In education studies, the triadic direction of the association between teacher support-related factors and
model is often used to examine how students’ self-efficacy and inner academic buoyancy. Their results demonstrated that academic buoy­
resources (i.e., cognitive/personal factors) mediate the associations ancy at Time 1 predicted teacher learning support (i.e., perceptions of
between classroom support (i.e., environmental factors) and engage­ teachers’ academic support and care) at Time 2 a year later, but this
ment and achievement outcomes (i.e., behavioral factors; Burns et al., association was not reciprocal. The authors noted that this unexpected
2018; Schunk & Mullen, 2012). In the current study, this theoretical finding may have occurred because of the dynamic relationships be­
framework was used to examine perceived autonomy support (i.e., an tween students and teachers, in which students can influence their
environmental factor) as a predictor of students’ academic buoyancy teachers as well as their own perceptions of their learning environments

2
A.A. Kingsford-Smith et al. Teaching and Teacher Education 142 (2024) 104516

(Bostwick et al., 2022; Nurmi & Kiuru, 2015). Another explanation may 1.5.1. Associations between perceived autonomy support, academic
be that students often have different teachers across school years, which buoyancy, and academic self-efficacy across different school-stage settings
may influence their perceptions of teacher support. In Bostwick et al.’s It is also worth considering whether the role that PAS plays towards
(2022) study, learning support at Time 1 was more strongly correlated students’ academic buoyancy and academic self-efficacy differs for
with academic buoyancy at Time 1 than with academic buoyancy at students in different school-stage settings, such as primary, high, and
Time 2 a year later when the students may have different teachers. central schools. In general, older students perceive their teachers to be
Although more longitudinal research is needed to further explore the less autonomy supportive (Gillet et al., 2012; Kleinkorres et al., 2023;
direction of such associations, there may be substantive benefits to Waxman & Huang, 1998). This may be because primary school students
investigating PAS as a predictor of academic buoyancy. This is because spend more time with one teacher than do high school students,
teachers can be trained to be more autonomy supportive through allowing primary school teachers to better know their students and
intervention, which can result in both short and long-term improve­ provide individualized support (Lester & Cross, 2015). There is also
ments in student outcomes (Reeve et al., 2019). often more emphasis on discipline and controlling teacher behaviors in
high school settings (Bergin & Bergin, 2009; Lewis, 2006), despite ad­
1.5. Academic self-efficacy and the predictive role of perceived autonomy olescents having greater desire for autonomy that is often not met at
support and academic buoyancy school (Mahatmya et al., 2012). This has been described as a misfit
between adolescents’ needs and their learning environments (Booth &
Academic self-efficacy refers to a student’s confidence that they will Gerard, 2014; Eccles et al., 1993).
be successful in academic tasks at school (Bandura, 2001; Valentine Although PAS is generally experienced at lower levels in high
et al., 2004). Although students can hold distinct self-efficacy beliefs schools, it is not known if it plays a larger role in developing students’
towards individual school subjects, Green et al. (2007) demonstrated academic self-beliefs in either primary or high school settings. On one
that there is significant shared variance across academic domains. Ac­ hand, PAS may be particularly important for younger students’ aca­
ademic self-efficacy is associated with many positive student outcomes demic buoyancy and academic self-efficacy because younger students’
that are pertinent to students in rural areas, including academic self-beliefs can be more malleable (Martinek & Kipman, 2016). Students
achievement (Multan et al., 1991; Valentine et al., 2004; Weiβenfels in primary school have less academic experience and interact with fewer
et al., 2023) and future study intentions (Bong, 2001; Martin, 2009). teachers, and therefore the autonomy support they receive may be
Wettersten et al. (2005) found that the academic self-efficacy of high particularly important for developing their academic self-beliefs
school students in rural schools in the US was predictive of both career (Skinner et al., 1998). On the other hand, PAS may be particularly
and academic outcome expectations and of school engagement. Simi­ salient for high school students because older students generally expe­
larly, in a sample of high school students in rural schools in the US, rience less of this type of support (Gillet et al., 2012; Kleinkorres et al.,
Hardre and Reeve (2003) identified that perceived competence (a factor 2023; Waxman & Huang, 1998), yet developmentally desire greater
related to self-efficacy) was a significant predictor of students’ in­ autonomy (Mahatmya et al., 2012). Research findings are mixed in this
tentions to persist at school. Thus, academic self-efficacy appears to be area. In a comparison between middle school (i.e., early high school)
an important outcome for students in rural areas. and primary school students’ experiences, Lei et al. (2018) found
Academic buoyancy may predict greater academic self-efficacy. In teacher support had a stronger association with negative academic
Bostwick et al.’s (2022) longitudinal study, academic buoyancy was emotions and a weaker association with positive academic emotions for
found to be a significant predictor of students’ perceived academic middle school students, which may implicate other academic self-belief
confidence (which is related to academic self-efficacy) a year later, but constructs. Due to these mixed findings, we investigated whether the
this association was not reciprocal. Other cross-sectional studies have associations between PAS, academic buoyancy, and self-efficacy differed
provided support that academic buoyancy can indirectly predict aca­ for students in different school-stage settings as an open research
demic achievement via academic self-efficacy (Weiβenfels et al., 2023) question.
and related constructs such as academic self-concept (Colmar et al., Academic buoyancy’s role towards academic self-efficacy may also
2019). Although it is likely that academic buoyancy and academic differ for students in different school-stage settings. As mentioned,
self-efficacy are reciprocal in nature (Martin et al., 2010), it is mean­ younger students’ academic self-beliefs are likely more malleable
ingful for educators to understand the predictive role that academic because they have less school experience (Martinek & Kipman, 2016).
buoyancy plays towards academic self-efficacy. This is because school Academic buoyancy may therefore be critical for younger students to
experiences involve innate challenges that could shape students’ aca­ positively evaluate their abilities in the face of challenge (Colmar et al.,
demic self-efficacy. Experiencing challenges and learning from failure is 2019). However, academic buoyancy may also be particularly important
a core part of the learning process (Tawfik et al., 2015). Academic for high school students’ academic self-efficacy because older students
buoyancy may help students to view challenges as opportunities to may experience less positive and individualized support from their
learn, instead of as indicators of failure (Dweck & Master, 2007) and teachers (Gillet et al., 2012; Kleinkorres et al., 2023; Lester & Cross,
protect students’ academic self-efficacy in the face of adversity (Martin 2015; Waxman & Huang, 1998), alongside increasing academic pressure
& Marsh, 2020; Weiβenfels et al., 2023). Therefore, in the current study, related to end of school examinations (Granziera et al., 2022). Thus, it is
we hypothesized that academic buoyancy would predict greater aca­ not known if academic buoyancy is more or less salient to academic
demic self-efficacy. self-efficacy across different school-stage settings. In the current study,
PAS has also been shown to predict greater academic self-efficacy we investigated potential differences in the role that academic buoyancy
(Gutiérrez & Tomás, 2019; Zhao & Qin, 2021), as well as related con­ plays towards academic self-efficacy in different school-stage settings as
structs like perceived competence in rural school settings (Zhou et al., an open research question.
2009). PAS may nurture academic self-efficacy because it is associated There is currently scant research on these constructs for students in
with student engagement (Cheon et al., 2016) and academic achieve­ central schools because central schools are particular to the NSW context
ment (Jang et al., 2012). Autonomy supportive teaching may help draw (Boylan, 1988). There are less than 70 central schools in NSW and most
attention away from peer comparisons (Burns et al., 2017) and instead research on these schools has focused on the experiences of principals
help students to focus on their own personal accomplishments. In the and teachers (Green, 2008; Handal et al., 2018; Pietsch & Williamson,
current research, we hypothesized that PAS would predict greater aca­ 2008). However, since central schools serve both primary and high
demic self-efficacy, both directly (Gutiérrez & Tomás, 2019; Zhao & Qin, school-aged students, exploring these school contexts may help to
2021) and indirectly via academic buoyancy (Granziera et al., 2022; disentangle whether student age and the school learning environment
Weiβenfels et al., 2023). are implicated in the role that PAS plays towards students’ self-beliefs.

3
A.A. Kingsford-Smith et al. Teaching and Teacher Education 142 (2024) 104516

1.6. Age, gender, and socioeconomic status 0.9% identified as Torres Strait Islander, 4.4% identified as both
Aboriginal and Torres Strait Islander, and 2.3% did not respond about
Three student demographic covariates were used as statistical con­ their Aboriginality. Students who spoke a language other than English at
trols in the current study: age, gender, and socioeconomic status (SES). home represented 7.3% of the sample, with missing data representing
Student age can be predictive of PAS (Gillet et al., 2012; Kleinkorres 2.2% in this variable. Most of the sample had attended preschool
et al., 2023; Waxman & Huang, 1998), academic buoyancy (Martin (88.3%), with 2.2% not responding to this item.
et al., 2010), and self-efficacy (Pajares et al., 2007; Watt, 2004; Wigfîeld
& Eccles, 1994), with younger students often experiencing more adap­ 2.2. Measures
tive school experiences. Female students have also been shown to report
greater PAS (Lietaert et al., 2015). For student self-beliefs, some research Data were collected using student self-report surveys. All measures
suggests male students report greater academic self-efficacy, though this were collected in the same survey. The students completed printed
varies across subject domains (Huang, 2013). In Australia and other versions of the survey at their school, which were then returned to the
countries, male students are more likely than female students to be researchers and converted into a digital format for analysis. The mea­
disengaged, experience high levels of anxiety, and achieve lower aca­ sures included in this study are described below.
demic outcomes (Centre for Education Statistics and Evaluation [CESE],
2017; Lietaert et al., 2015). Regarding SES, there is evidence that 2.2.1. Perceived autonomy support
teachers in lower SES areas (in comparison to higher SES areas) may Perceived autonomy support (PAS) was measured using an adapted
have lower beliefs about their students’ capacities (Archambault et al., version of the short Learning Climate Questionnaire (Williams & Deci,
2012; Beswick et al., 2019; Solomon et al., 1996), which may implicate 1996). Students responded to six items on a 7-point Likert-like scale
teachers’ motivating styles and students’ academic self-beliefs. Addi­ (“strongly disagree” to “strongly agree”). The items asked students the
tionally, students from lower SES backgrounds may experience different degree to which they agreed that their teachers use autonomy sup­
forms of disadvantage that implicate their educational progress and portive approaches and practices (example item: “My teachers try to
school experiences (Rosenthal, 1998). Thus, it is important to account understand how I see things before suggesting a new way of doing
for these three demographic covariates. things”). The short version of the Learning Climate Questionnaire has
demonstrated appropriate reliability and construct validation in previ­
1.7. Research purpose and aim ous research (Jang et al., 2012). In the current study the scale demon­
strated adequate internal reliability (α = 0.94).
The aim of the current study was to examine the extent to which
perceived autonomy support (PAS) predicts academic buoyancy and 2.2.2. Academic buoyancy
academic self-efficacy for students in rural schools. We also examined Academic buoyancy was measured using the Academic Buoyancy
whether academic buoyancy partially mediates the associations be­ Scale (Martin & Marsh, 2008b). Students responded to four items on a
tween PAS and academic self-efficacy. Further, we examined whether 7-point Likert-like scale (“strongly disagree” to “strongly agree”). The
these statistical associations varied for students in different rural school- items asked students the degree to which they agreed that they could
stage settings, including primary, high, and central schools. A unique handle the day-to-day challenges experienced at school (example item:
aspect of this study is its sole focus on non-metropolitan students and “I think I’m good at dealing with schoolwork pressures”). The Academic
schools. Buoyancy Scale has demonstrated appropriate reliability and construct
validation in previous research (Martin & Marsh, 2008a). In the current
2. Methods study the scale demonstrated adequate internal reliability (α = 0.85).

2.1. Participants 2.2.3. Academic self-efficacy

Academic self-efficacy was measured using the Academic Self-
The dataset for this study was collected as a part of a larger study Efficacy Scale (Chemers et al., 2001). Students responded to eight
across 17 schools (Rural and Regional Education Project; Beswick et al., items using a 7-point Likert-like scale (“never true” to “always true”).
2023). Ethics was approved by the ethics committees of both the uni­ The items asked students how true they believed statements to be about
versity (HC200613) and the NSW Department of Education. The 17 their abilities at academic tasks (example item: “I usually do very well in
schools were selected by convenience sampling. Consent to participate school and at academic tasks”). The Academic Self-Efficacy Scale has
in the study was received from both students and their parents. Only 16 demonstrated appropriate reliability and construct validation in previ­
of the 17 schools participated in the student survey. The sample for this ous research (Chemers et al., 2001). In the current study the scale
study included 974 students across 16 government schools in outer demonstrated adequate internal reliability (α = 0.93).
regional and remote areas in New South Wales (NSW) in Australia
(Australian Bureau of Statistics, 2021). The 16 schools had index of 2.2.4. Student covariates
community socio-educational advantage (ICSEA) scores—which is Three student covariate measures were used as statistical controls.
indicative of SES—below the national average. Seven schools were Age was computed as a continuous variable using students’ date of birth.
within 1 standard deviation (SD) below the national average, seven Gender was measured as a binary variable (0 = female, 1 = male). Due to
schools were between 1 and 2 SD below the national average, and two the small percentage of students who selected “other” (1.3%) for their
schools were between 2 and 3 SD below the national average. Seven gender, this response was treated as missing data. For SES, students
were primary schools (student sample = 263), three were high schools reported their parents’ occupations using five categories, which were
(student sample = 523), and six were central schools (student sample = then converted into a 5-point Likert-like scale. Occupation-based scales
188). have previously been used to represent SES in Australia (McMillan et al.,
The students were in Year 5 (19.9%), Year 6 (15.2%), Year 7 2009). The highest category across both parents for each student was
(18.8%), Year 8 (16.0%), Year 9 (12.7%), Year 10 (10.2%), Year 11 used as students’ SES value. The categories were: 1 = Have not been in
(4.3%), and Year 12 (2.8%). Only 0.6% of students did not provide their paid work in the last 12 months (6.3% of the sample); 2 = Machine op­
year group. The mean age was 13.4 years old. Male students represented erators, hospitality staff, assistants, labourers, or related workers (29.1%); 3
46.5% of the sample, female students represented 50.2%, 1.4% identi­ = Tradespeople, clerks and skilled office, sales, or service staff (26.5%); 4 =
fied as “other” as their gender, and 2.0% did not disclose their gender. Other business managers, arts/media/sportspersons or associate pro­
Students who identified as Aboriginal represented 27.9% of the sample, fessionals (15.9%); 5 = Senior management in large business organisation,

4
A.A. Kingsford-Smith et al. Teaching and Teacher Education 142 (2024) 104516

government administration and defence, or qualified professionals (13.7%). 3. Results

2.3. Statistical analyses 3.1. Measurement model

Analyses were conducted using Mplus, Version 8 (Muthén & Muthén, The descriptive statistics are reported in Table 1. All measures
2017). Robust maximum likelihood (MLR) was used as the estimator for demonstrated approximate normal distribution. Next, CFA was con­
the analyses. Full information maximum likelihood was used to handle ducted to test the measurement model. The model fit indices indicated
missing data. A measurement model with the three substantive latent acceptable model fit: df = 177, χ2 = 461.84 (p < .001); RMSEA = 0.04;
factors and three covariates was first tested through confirmatory factor CFI = 0.97; TLI = 0.96; SRMR = 0.04. The standardized factor loadings
analysis (CFA) to check for appropriate model fit of the constructs. can be seen in Table 1. These loadings were high and statistically sig­
Covariates were measured as single-item manifest factors, with item nificant for all factors, indicating that the measurement model provided
factor loadings fixed to 1 and item residual variances fixed to 0. Model fit a sound base for further analyses. The correlations between the latent
was assessed using the following criteria recommended by Keith (2015): variables and student covariates can be seen in Table 2. The correlations
root mean square error of approximation (RMSEA) scores ≤0.05 (≤0.08 between the substantive factors had expected relationships, such that
for adequate fit), Tucker-Lewis index (TLI) scores ≥0.95 (≥0.90 for perceived autonomy support (PAS), academic buoyancy, and academic
adequate fit), comparative fit index (CFI) scores ≥0.95 (≥0.90 for self-efficacy had strong and positive inter-correlations.
adequate fit), and standardized root mean squared residual (SRMR)
scores ≤0.06 (≤0.08 for adequate fit). 3.2. Whole sample SEM results
Following the CFA, a structural equation model (SEM) was tested for
all students in the sample. In this model, academic buoyancy and aca­ Following the measurement model, a whole sample SEM model was
demic self-efficacy were regressed onto PAS, and academic self-efficacy tested for all students in the dataset. The results can be seen in Table 3.
was regressed onto academic buoyancy. All three substantive factors The fit indices indicated acceptable model fit: df = 177, χ2 = 461.84 (p
were also regressed onto the three covariate measures (i.e., gender, age, < .001); RMSEA = 0.04; CFI = 0.97; TLI = 0.96; SRMR = 0.04. The SEM
and SES). A secondary analysis was conducted to inspect the indirect results can be seen in Table 3. The results demonstrated that PAS pre­
associations between PAS and academic self-efficacy via academic dicted academic buoyancy (β = 0.50, p < .001) and academic self-
buoyancy. Confidence intervals (95%) were inspected with non- efficacy (β = 0.26, p < .001). Academic buoyancy predicted academic
parametric bootstrapping for the indirect associations, using 1000 self-efficacy (β = 0.50, p < .001). The indirect effects from the bootstrap
draws with maximum likelihood estimation. SEM can be seen in Table 4. PAS had a significant indirect association
Next, multigroup analyses were conducted to test for differences with academic self-efficacy via academic buoyancy (β = .25; 95% CI =
between the school-stage setting groups (i.e., primary, high, and central [0.20 0.31]), with a total association (direct and indirect) of β = 0.52;
schools). The means of the covariate and substantive variables were 95% CI = [0.45 0.58]).
compared across the school-stage setting groups. Since Mplus computes Some significant associations were found between the student
latent variable means to be zero, mean-based composite variables were covariates and substantive factors. Older student age predicted lower
used to compare the means of PAS, academic buoyancy, and academic PAS (β = − 0.28, p < .001) and academic buoyancy (β = − 0.08, p < .05).
self-efficacy. T-tests were conducted using SPSS (version 26; IBM, 2022) Female students reported significantly greater PAS (β = − 0.09, p < .01)
to test for significant differences. and academic self-efficacy (β = − 0.16, p < .001). Higher SES predicted
Models with increasing parameter restraints were then evaluated to greater academic self-efficacy (β = 0.07, p < .05). In total, the model
test the invariance of the measurement model across the groups. This explained 28.5% of the variance of academic buoyancy and 51.0% of the
included a configural model where factor loadings and intercepts were variance of academic self-efficacy.
freely estimated between the groups; a metric model, where factor
loadings were constrained to be equal between the groups; and a scalar 3.3. Multigroup invariance tests
model, where factor loadings and intercepts were constrained to be
equal between the groups. Minimal changes in model fit (ΔRMSEA Next, multigroup analyses were conducted. Construct invariance
≤0.015; Δ CFI ≤0.01) would indicate measurement invariance across between the school-stage setting groups was tested through comparing
the groups (Chen, 2007; Cheung & Rensvold, 2002). Since the aim of this the model fit of configural, metric, and scalar models. The model fit
multigroup analysis was to test for meaningful differences between the indices are shown in Table 5. Model fit remained similar across the three
school-stage setting groups, scalar invariance was the desired level of models (Chen, 2007; Cheung & Rensvold, 2002), indicating that the
invariance (Collie et al., 2020; van de Schoot et al., 2012). measurement model structure was invariant across the school-stage
Multigroup SEM was then conducted to test whether there were setting groups.
differences in the associations between the substantive factors for stu­
dents in primary, high, and central schools. In this multigroup analysis, 3.4. Multigroup mean-level differences
the scalar invariance measurement model was used, with the same
regression paths that were included as the initial SEM. However, The means of the substantive factors and covariates were compared
regression betas were allowed to vary between the three groups. Indirect across the school-stage setting groups (see Table 6). Both primary and
associations were again inspected using non-parametric bootstrapping central school students reported significantly greater PAS than did high
(1000 draws) and 95% confidence intervals. school students (primary vs. high: t(769) = 9.10, p < .001; high vs.
Two approaches were used to explore differences and similarities central: t(691) = − 7.05, p < .001). There was no significant difference
between the school-stage setting groups in the multigroup SEM. First, between primary and central school students’ PAS (primary vs. central: t
overall model invariance was tested by investigating change in model fit (442) = 1.22, p = .23). Primary school students reported significantly
when the beta values between the substantive factors were constrained greater academic buoyancy than did central and high school students
to be equal between groups. Minimal changes in model fit (ΔRMSEA (primary vs. high: t(751) = 7.00, p < .001; primary vs. Central: t(436) =
≤0.015; Δ CFI ≤0.01) would indicate broad invariance across the 3.80, p < .001). Central school students’ academic buoyancy was greater
groups (Chen, 2007; Cheung & Rensvold, 2002). Next, path-level than that of high school students, but the difference was only borderline
invariance was tested by conducting Wald difference tests (p < .05) on significant (high vs. central: t(669) = − 1.95, p = .05). Primary school
beta values between groups. Each beta value for the substantive factor students reported significantly greater academic self-efficacy than did
associations were tested between the groups. central and high school students (primary vs. high: t(733) = 6.08, p <

5
A.A. Kingsford-Smith et al. Teaching and Teacher Education 142 (2024) 104516

Table 1
Descriptive statistics and factor loadings for whole sample.
Min Max Missing Mean/% SD Skewness Kurtosis Factor loading Alpha
%

Age 9.66 17.98 7.3% 13.38 1.93 0.22 − 0.91

Gender (male) 0 1 3.3% 46.5% 0.50 0.08 − 2.00
SES 1 5 8.6% 3.02 1.17 0.27 − 0.88
Perceived autonomy support .94
Item 1 1 7 2.7% 4.55 1.71 − 0.46 − 0.54 .83
Item 2 1 7 2.7% 4.93 1.59 − 0.71 − 0.02 .89
Item 3 1 7 3.4% 4.69 1.69 − 0.52 − 0.48 .89
Item 4 1 7 2.5% 5.07 1.55 − 0.70 − 0.03 .89
Item 5 1 7 2.9% 5.01 1.64 − 0.72 − 0.11 .81
Item 6 1 7 2.9% 4.76 1.69 − 0.57 − 0.37 .83
Academic buoyancy .85
Item 1 1 7 5.0% 4.56 1.62 − 0.47 − 0.44 .72
Item 2 1 7 5.4% 4.28 1.88 − 0.32 − 0.97 .78
Item 3 1 7 5.2% 4.21 1.79 − 0.27 − 0.92 .85
Item 4 1 7 5.4% 4.73 1.79 − 0.58 − 0.64 .70
Academic self-efficacy .91
Item 1 1 6 7.2% 4.39 1.69 − 0.34 − 0.70 .80
Item 2 1 6 6.9% 4.91 1.71 − 0.61 − 0.49 .76
Item 3 1 6 7.4% 4.49 1.80 − 0.33 − 0.85 .81
Item 4 1 6 7.5% 4.60 1.77 − 0.46 − 0.70 .81
Item 5 1 6 7.9% 4.89 1.80 − 0.59 − 0.59 .73
Item 6 1 6 8.8% 4.81 1.67 − 0.57 − 0.39 .87
Item 7 1 6 8.1% 4.32 1.74 − 0.36 − 0.70 .77
Item 8 1 6 7.2% 4.99 1.78 − 0.70 − 0.42 .79

Note. Factor loadings are standardized coefficients. For gender, 0 = female, 1 = male. SES = socioeconomic status.

Table 2 Table 4
Correlations for whole sample. Indirect results from bootstrap SEM for whole sample.
1. Perceived 2. Academic 3. Academic self- Indirect Total (indirect +
autonomy support buoyancy efficacy Beta (β) [95% CI direct]
Values] Beta (β) [95% CI
1. Perceived
Values]
autonomy
support Perceived autonomy support - > .25 [.20 .31] .52 [.45 .58]
2. Academic .52*** Academic buoyancy ->
buoyancy Academic self-efficacy
3. Academic self- .55*** .65***
efficacy Note. All β values are standardized. CI = confidence interval.
Age ¡.28*** ¡.22*** ¡.19***
Gender ¡.08* − .01 ¡.19***
SES .02 .06 .10** Table 5
Multigroup CFA and SEM model fit.
Note. SES = socioeconomic status.
*p < .05; **p < .01; ***p < .01. Chi- df RMSEA CFI TLI SRMR
Square

Preliminary CFA Invariance Tests

Table 3 Configural 655.36 396 .045 .97 .96 .05
SEM results for whole sample. Metric 690.32 426 .044 .97 .96 .06
Scalar 744.19 456 .045 .96 .96 .06
Standardized Multigroup SEM 982.04 591 .045 .96 .95 .05
Beta (β) Multigroup SEM with 996.73 597 .046 .96 .95 .06
Perceived autonomy support Constrained Betas Values
Age - > Perceived autonomy support ¡.28***
Note. df = degrees of freedom; RMSEA = root mean-square error of approxi­
Gender (male) - > Perceived autonomy support ¡.09**
SES - > Perceived autonomy support .03
mation; CFI = comparative fit index; TLI = Tucker-Lewis index; SRMR = stan­
Perceived autonomy support - > Academic buoyancy .50*** dardized root mean square residual.
Age - > Academic buoyancy ¡.08*
Gender (male) - > Academic buoyancy .03 The means of the covariate factors were also compared across the
SES - > Academic buoyancy .05
Perceived autonomy support - > Academic self-efficacy .26***
school-stage setting groups. As expected, student age varied across the
Academic buoyancy - > Academic self-efficacy .50*** groups (primary vs. high: t(719) = − 32.87, p < .001; primary vs. central:
Age - > Academic self-efficacy − .01 t(416) = − 14.11, p < .001; high vs. central: t(665) = 11.12, p < .001).
Gender (male) - > Academic self-efficacy ¡.16*** All three school-stage setting groups were similar in regard to gender
SES - > Academic self-efficacy .07*
proportions (primary vs. high: t(758) = − 0.05, p = .96; primary vs.
Note. SES = socioeconomic status. All β values are standardized. central: t(436) = 0.47, p = .64; high vs. central: t(684) = 0.57, p = .57).
*p < .05; **p < .01; ***p < .001. Similarly, all three school-stage setting groups were similar regarding
SES (primary vs. high: t(706) = − 0.12, p = .91; primary vs. central: t
.001; primary vs. central: t(434) = 3.90, p < .001). There was no sig­ (421) = − 0.73, p = .46; high vs. central: t(647) = − 0.67, p = .51).
nificant difference between central and high school students’ academic
self-efficacy (high vs. central: t(653) = − 1.13, p = .26).

6
A.A. Kingsford-Smith et al. Teaching and Teacher Education 142 (2024) 104516

Table 6 [0.44 0.69]); high school group (indirect association: β = 0.18, 95% CI
Comparisons of group means. = [0.12 0.25]; total association: β = 0.47, 95% CI = [0.38 0.56]); central
Primary School High School (2) Central School school group (indirect association: β = 0.21, 95% CI = [0.10 0.33]; total
(1) (3) association: β = 0.51, 95% CI = [0.38 0.64]).
Age M = 11.28, SD = M = 14.52, SD = M = 13.05, SD = Overall, the model explained the most variance for the primary
0.70 (2, 3) 1.43 (1, 3) 1.75 (1, 2) school group, explaining 47.7% of the variance of academic buoyancy
Gender (female = 0; M = 0.48, SD = M = 0.49, SD = M = 0.46, SD = and 69.7% of the variance of academic self-efficacy. For the high school
male = 1) 0.50 0.50 0.50 group, the model explained 18.4% of the variance of academic buoyancy
SES M = 3.00, SD = M = 3.01, SD = M = 3.08, SD =
1.06 1.21 1.21
and 39.4% of the variance of academic self-efficacy. For the central
Perceived M = 5.39, SD = M = 4.40, SD = M = 5.23, SD = school group, the model explained 25.2% of the variance of academic
autonomy 1.42 (2) 1.42 (1, 3) 1.19 (2) buoyancy and 54.5% of the variance of academic self-efficacy.
support
Academic buoyancy M = 4.95, SD = M = 4.18, SD = M = 4.43, SD =
4. Discussion
1.39 (2, 3) 1.45 (1, 3^) 1.44 (1, 2^)
Academic self- M = 5.12, SD = M = 4.45, SD = M = 4.59, SD =
efficacy 1.42 (2, 3) 1.42 (1) 1.35 (1) This study investigated the associations between perceived auton­
omy support (PAS), academic buoyancy, and academic self-efficacy for
Note. Numbers in brackets identify groups that have significantly different
means (p < .05). M = mean; SD = standard deviation; ^ = borderline significance
students in rural schools in Australia. The results revealed that students’
(p = .05); SES = socioeconomic status. PAS was a significant predictor of their academic buoyancy and aca­
demic self-efficacy. Further, academic buoyancy was found to partially
mediate the association between PAS and academic self-efficacy in high
3.5. Multigroup SEM results
and central school contexts, and to fully mediate this association in
primary school contexts. These results provide support for autonomy
The multigroup SEM fit indices indicated appropriate fit (see
support as an effective motivating style across different school-stage
Table 5). Model fit remained similar when the substantive factor beta
settings in rural areas for nurturing students’ ability to handle chal­
values were constrained to be equal across the groups, indicating broad
lenges and feel efficacious in their school studies.
model-level invariance (see Table 5). The multigroup SEM results for the
This study offers several important contributions to the literature.
unconstrained beta model can be seen in Fig. 1 and Table 7.
First, rural education has received much less academic attention than
Next, path-level similarities and differences in the substantive factors
metropolitan school settings (Roberts & Fuqua, 2021; Stockard, 2011).
between the school-stage setting groups are reported. PAS predicted
The current results are consistent with other rural-focused studies in
greater academic buoyancy for all three groups (primary: β = .66, p <
showing that PAS was associated with adaptive student self-beliefs in
.001; high: β = 0.42, p < .001; central: β = 0.44, p < .001). Wald dif­
rural schools (Hardre & Reeve, 2003; Zhou et al., 2009). Together, the
ference tests found the betas were significantly different between the
findings suggest that autonomy support may help to address educational
primary and high school groups (Wald = 4.33, p < .05), but were not
outcome gaps in rural schools (Echazarra & Radinger, 2019; OECD,
significantly different between the primary and central school groups
2018). Second, the finding that PAS predicted greater academic buoy­
(Wald = 0.50, p = .48), nor between the high and central school groups
ancy is a novel contribution of this study. This finding is consistent with
(Wald = 0.70, p = .40).
previous research that has shown positive associations between teacher
PAS predicted greater academic self-efficacy for the high school (β =
support and academic buoyancy (Granziera et al., 2022; Tarbetsky et al.,
0.29, p < .001) and central school (β = 0.30, p < .001) groups; however,
2017). This suggests that autonomy support may help students to better
this association was not significant for the primary school group (β =
handle the day-to-day challenges of school life. Third, the current
0.10, p = .28). Wald difference tests did not find significant differences
study’s multigroup model demonstrated important similarities and dif­
in the beta values between the primary and high school groups (Wald =
ferences in the levels of and associations between the substantive factors
2.57, p = .11), nor between the primary and central school groups
in primary, high, and central school settings. Scant research has inves­
(Wald = 2.86, p = .09), nor between the high and central school groups
tigated such differences in rural contexts. The current results suggest
(Wald = 0.20, p = .65).
that primary school students may have the most adaptive school expe­
Academic buoyancy predicted greater academic self-efficacy for all
riences. These findings are discussed below.
groups (primary: β = 0.70, p < .001; high: β = 0.44, p < .001; central: β
= 0.48, p < .001). Wald differences tests found the betas were signifi­
cantly different between the primary and high school groups (Wald = 4.1. School-stage setting similarities
7.44, p < .01), and between the primary and central school groups
(Wald = 4.89, p < .05), but were not significantly different between the The current study identified several similarities across the three
high and central school groups (Wald = 0.10, p = .75). school-stage settings in the associations between PAS, academic buoy­
Turning to the covariate associations, age predicted lower PAS for ancy, and academic self-efficacy. First, PAS predicted greater academic
the central school group (β = − 0.40, p < .001), but this association was buoyancy in all three school-stage settings. This is a novel finding in the
not significant for the primary and high school groups. Gender (male) current study. From this, it appears that supporting students’ internal
predicted lower PAS for the primary school group (β = − 0.22, p < .001), motivational resources is associated with students better coping with
but this association was not significant for the central and high school academic setbacks. This may be because students who receive autonomy
groups. Gender (male) predicted lower predicted academic self-efficacy supportive teaching may value and enjoy their schooling (Reeve, 2016)
for all three school-stage setting groups (primary: β = − 0.19, p < .001; and thus may see academic setbacks as part of this valued learning
high: β = − 0.14, p < .001; central: β = − 0.18, p < .01). SES predicted process (Dweck & Master, 2007). Academic buoyancy is a relatively new
greater academic self-efficacy for the central school group (β = 0.17, p < construct in the literature, yet it has been linked to several important
.01), but this association was not significant for the primary and high educational outcomes including school engagement (af Ursin et al.,
school groups. 2021), valuing and enjoyment of school (Martin et al., 2017), and future
PAS had a significant indirect association with academic self-efficacy study intentions (Martin et al., 2013). It is thus beneficial to better un­
via academic buoyancy for all three school-stage setting groups (see derstand how academic buoyancy can be supported in educational set­
Table 8): primary school group (indirect association: β = 0.46, 95% CI = tings. Although the current results are only cross-sectional, previous
[0.30 0.63]; total association [direct and indirect]: β = 0.56, 95% CI = longitudinal research has demonstrated that students’ prior levels of
autonomy satisfaction in learning contexts are predictive of future

7
A.A. Kingsford-Smith et al. Teaching and Teacher Education 142 (2024) 104516

Fig. 1. Multigroup SEM Results

Note. All paths show the standardized beta values. Only significant (p < .05) paths are shown. SES = socioeconomic status. For gender, 0 = female, 1 = male. Bold
paths are significantly different from the primary school group (p < .05). *p < .05; **p < .01; ***p < .001.

academic buoyancy (Aydın & Michou, 2020). Future studies may wish self-efficacy across different school-stage settings through longitudinal
to explore this area further through examining whether autonomy research design (Bostwick et al., 2022; Martin et al., 2010).
support interventions in schools (Reeve et al., 2019) help to increase PAS was found to have positive indirect associations with academic
academic buoyancy over time in rural schools. self-efficacy via academic buoyancy in all three school-stage settings.
Academic buoyancy also predicted greater academic self-efficacy in Although the direct associations between PAS and academic self-efficacy
all three school-stage settings. This is consistent with previous findings differed across the school-stage settings (as is discussed later), the in­
that academic buoyancy was associated with academic self-concept in direct associations indicated that higher levels of PAS were linked to
primary school settings (Colmar et al., 2019) and self-efficacy in high higher levels of academic self-efficacy, which is consisted with previous
school settings (Martin et al., 2010). From this, it appears that students research (Gutiérrez & Tomás, 2019; Hardre & Reeve, 2003; Zhao & Qin,
with lower academic buoyancy also feel less efficacious. For these stu­ 2021; Zhou et al., 2009). This could be because autonomy supportive
dents, a setback such as a lower test result may damage their sense of teaching fosters student engagement (Cheon et al., 2016) and supports
self-efficacy and begin a downward spiral of low levels of buoyancy and academic achievement (Jang et al., 2012), allowing students to feel
self-efficacy (Martin et al., 2010). Although this association has previous buoyant and successful in the classroom. Autonomy support may also
evidence (Martin et al., 2017), the current study provides empirical help students to focus on their own successes, instead of on peer com­
support for this association in primary, high, and central schools in rural parisons (Burns et al., 2017). Thus, the results suggest that autonomy
settings. Future studies may wish to further explore the potential support may be an effective pedagogical approach to support both ac­
reciprocal relations between academic buoyancy and academic ademic buoyancy and academic self-efficacy in rural schools across

8
A.A. Kingsford-Smith et al. Teaching and Teacher Education 142 (2024) 104516

Table 7 with one teacher over a school year, which may help stronger
Multigroup SEM results. student-teacher relationships develop (Roorda et al., 2011) and thus
Primary High Central may support academic buoyancy (Tarbetsky et al., 2017) and academic
School School School self-efficacy (Hughes & Chen, 2011).
Perceived autonomy support On the other hand, high school students’ lower levels of PAS may be
Age - > Perceived autonomy support .04 .02 ¡.40*** connected to their learning environments. Researchers have suggested
Gender (male) - > Perceived autonomy ¡.22*** − .03 − .08 that high school students can desire greater autonomy than current
support school systems provide (Mahatmya et al., 2012), which may limit their
SES - > Perceived autonomy support .04 .00 .05
Perceived autonomy support - > Academic .66*** .42*** .44***
perceptions of autonomy support. High school students’ lower levels of
buoyancy academic buoyancy and academic self-efficacy may also be linked to
Age - > Academic buoyancy .04 − .06 − .10 other factors, such as increasing academic pressures in high school and
Gender (male) - > Academic buoyancy − .06 .08 .08 peer comparisons (Berry & West, 1993; Granziera et al., 2022; Smith &
SES - > Academic buoyancy .09 .01 .08
Sinclair, 2000). Such extrinsic drives may limit the relevance that stu­
Perceived autonomy support - > Academic .10 .29*** .30***
self-efficacy dents see in their education, especially if tertiary education or high ex­
Academic buoyancy - > Academic self-efficacy .70*** .44*** .48*** amination marks are not valued by the student. This is particularly
Age - > Academic self-efficacy − .03 .04 − .02 important in rural schools where examples drawn from students’ expe­
Gender (male) - > Academic self-efficacy ¡.19*** ¡.14*** ¡.18** riences are often seen as an important engagement hook (Roberts,
SES - > Academic self-efficacy .04 .05 .17**
2013). Indeed, the growth of place-based education in rural schools
Note. SES = socioeconomic status. All β values are standardized. reflects the increasing recognition of needing to engage students using
*p < .05; **p < .01; ***p < .001. their own social and cultural experiences (Corbett, 2020). Thus, the
current findings suggest that high school students in rural schools may
benefit from greater autonomy support at school to help them find
Table 8
relevance in their learning, and to build greater academic buoyancy and
Indirect results from bootstrap multigroup SEM.
academic self-efficacy.
Indirect Total (indirect + The substantive factor scores for the central school group sat in be­
Beta (β) [95% CI direct]
Values] Beta (β) [95% CI
tween the primary and high school groups. However, the differences
Values] between groups were not consistently significant. There was no signif­
icant difference in PAS between primary and central school students.
Primary school students
Perceived autonomy support - > .46 [.30 .63] .56 [.44 .69] This may be due to the small student and teacher numbers in central
Academic buoyancy -> schools, which may allow closer student-teacher relationships to
Academic self-efficacy develop (Crosnoe et al., 2004) and thus may aid greater autonomy
High school students support. However, younger students in central schools did report higher
Perceived autonomy support - > .18 [.12 .25] .47 [.38 .56]
Academic buoyancy ->
levels of PAS than did older students, which suggests there may still be
Academic self-efficacy primary and high school-level differences within central schools. The
Central school students academic self-efficacy of the central school group was not significantly
Perceived autonomy support - > .21 [.10 .33] .51 [.38 .64] different from the high school group. Central school settings can vary
Academic buoyancy ->
quite considerably from each other because of school size, teachers
Academic self-efficacy
available, and proportion of students in primary and high school years
Note. All β values are standardized. CI = confidence interval. (What Works National Office National Curriculum Services, 2012). This
makes it difficult to disentangle why self-efficacy was lower in central
different school-stage settings. schools in comparison to primary schools. Further, SES predicted aca­
Female students reported greater academic self-efficacy than male demic self-efficacy in central schools, but not in the other two groups,
students across all three school-stage settings. This finding is consistent which demonstrates further complexities in the nature of academic
with the general trend of male students being less engaged and self-efficacy in central schools. More research is needed to better un­
achieving lower academic outcomes (CESE, 2017; Lietaert et al., 2015). derstand the unique attributes of central schools and how such contexts
Male students in rural schools have also been found to hold lower are associated with students’ PAS, academic buoyancy, and academic
educational aspirations than their female peers (CESE, 2015). The cur­ self-efficacy.
rent results indicate that teachers may need to be particularly aware of The current study also identified significant differences in the
supporting the academic self-efficacy of their male students in rural strength of associations between PAS, academic buoyancy, and self-
contexts. Next, differences between the school-stage settings are efficacy across the school-stage settings. Most notably, in the primary
discussed. school group academic buoyancy completely mediated the association
between PAS and academic self-efficacy, whereas this was only a partial
4.2. Differences between the school-stage settings mediation in the high and central school groups. Further, academic
buoyancy had a significantly stronger association with academic self-
Several differences were identified in the factor associations across efficacy for the primary school group than the other two groups.
the three school-stage settings. First, primary school students reported These findings suggest that academic buoyancy may be particularly
the most adaptive responses to their school experiences. These students important in primary school settings in rural contexts. This may be
reported greater PAS than did the high school group, and greater aca­ because primary school students’ academic self-beliefs are likely more
demic buoyancy and academic self-efficacy than did both the high and malleable than high school students (Martinek & Kipman, 2016). The
central school groups. Considering that PAS was found to predict greater results suggest that primary school students’ academic buoyancy and
academic buoyancy and academic self-efficacy, it seems likely that PAS academic self-efficacy are strongly linked. Future research may wish to
may be implicated in the differences in academic buoyancy and aca­ further investigate this area through longitudinal research design to
demic self-efficacy between the primary and high school groups. This better understand how academic buoyancy functions across different
finding is consistent with previous research that identified younger school-stage settings over time.
students perceive receiving greater autonomy support than older stu­ The association between PAS and academic buoyancy was also
dents (Gillet et al., 2012). Primary school students also spend more time stronger in the primary school group than the high school group. This

9
A.A. Kingsford-Smith et al. Teaching and Teacher Education 142 (2024) 104516

indicates that PAS may be more closely linked to academic buoyancy in 4.4. Limitations
primary schools than in high schools. Since primary school students
spend more time with one teacher over a school year, the autonomy Although the findings of this study have significant implications, it is
support they receive from their teacher may be more closely linked to important to note some limitations of the study. First, this study used
their daily academic resilience. cross-sectional data and thus causality between constructs cannot be
The current results indicate that autonomy support may be particu­ claimed. There are, however, some longitudinal studies that provide
larly salient in primary school settings in rural schools. The results also preliminary evidence for similar factor associations occurring between
indicate that high and central school students may benefit from auton­ different timepoints (Aydın & Michou, 2020; Martin et al., 2010). Future
omy support to nurture academic buoyancy and academic self-efficacy. studies may wish to use longitudinal analyses to build upon the current
Older students often report lower PAS, especially in middle high school findings. Second, the schools involved in this study were chosen using
years (Gillet et al., 2012) as their teachers may put more emphasis on convenience sampling. School-level convenience sampling was deemed
disciplinary action and classroom control (Bergin & Bergin, 2009). This appropriate as schools can be representative of wider student pop­
is counter to the developmental needs of adolescents for greater au­ ulations (Gravetter & Forzano, 2012). Despite this, more sophisticated
tonomy (Mahatmya et al., 2012). Indeed, disengaged students can often forms of sampling may be used in future studies, such as random sam­
receive the most controlling teacher behaviors (Sarrazin et al., 2006), pling to select schools, as used in OECD studies like the Programme for
despite the benefits that autonomy support can provide to students International Student Assessment (PISA). Third, the current study used
(Reeve, 2009). Greater autonomy support in high and central school self-report measures from students. Although academic buoyancy and
settings may help students to feel more buoyant and efficacious. Hence, academic self-efficacy are arguably best measured using students’
autonomy support may be beneficial in all school-stage settings in rural self-reports because they are internal unobservable factors, future
contexts. studies may wish to measure autonomy support through multiple
datasets (e.g., external observations, teacher reports). Fourth, we
4.3. Implications for rural schools examined student gender as a covariate, however, future research may
wish to examine whether the factor associations differ for male and fe­
The current results support the importance of autonomy support in male students because of potential differences in school experience
rural schools. This has implications for the pedagogies used in these (Lietaert et al., 2015). Finally, although we were unable to test the in­
schools. Some reports have indicated that teachers in more remote areas fluence of students’ sense of belonging in school and in the curriculum,
in Australia may on average have less experience than their provincial and the value of education in general, the results in the context of the
and metropolitan colleagues (McKenzie et al., 2014). Further, teachers broader research project (Beswick et al., 2023) from which they were
in rural areas may experience barriers to access professional develop­ drawn suggest this may well be an important avenue for further
ment due to long distances from urban centers and other schools, as well examination.
as shortages in casual staff to relieve teaching duties (Jenkins et al.,
2015). These factors may impact teachers’ ability and confidence to use 5. Conclusion
autonomy supportive practices. Schools in rural areas may benefit from
prioritizing professional learning that focuses on autonomy supportive The current study examined the associations between students’
teaching to increase teachers’ exposure to this motivating style (Reeve perceived autonomy support (PAS), academic buoyancy, and academic
et al., 2019). Studies have also shown that teachers are more likely to self-efficacy in rural schools in Australia. The results demonstrated that
adopt more autonomy-supportive approaches and feel more efficacious PAS is predictive of students’ academic buoyancy and academic self-
when their own autonomy and basic psychological needs are supported efficacy. The results also demonstrated primary school students on
at school (Moè et al., 2022; Orakci & Durnali, 2023; Pelletier et al., average report more adaptive school experiences than high and central
2002). Pre-service teachers may also benefit from greater training on school students. In these younger-year school settings, PAS was found to
autonomy support and also from experiencing autonomy support in principally predict students’ academic buoyancy, which in turn pre­
their teacher education (Chan et al., 2021; Khotimah et al., 2023). dicted academic self-efficacy. However, PAS was also found to be
The current results also have implications for national curriculums. important in high and central school contexts. High school students
Greater autonomy support may be realized through encouraging reported experiencing less of this type of support, which may lead to
teachers to develop confidence in linking their curriculum enactment lower levels of academic buoyancy and academic self-efficacy. Older
with local examples, and not being constrained by available texts that students in central schools also reported lower levels of PAS. As school
are not from a rural perspective (Roberts, 2013). However, it is also systems around the world struggle to reduce the gap between the
important that national curriculum topics are designed to be relevant to educational outcomes of rural and metropolitan schools, the current
rural students. Teachers in rural areas in Australia have reported results illustrate that autonomy support may help students in rural
needing to adapt national curriculum topics to be relevant to their stu­ schools to feel more resilient and efficacious.
dents (Lock et al., 2012; Papatraianou et al., 2018). Some researchers
have argued that the Australian national curriculum prioritizes metro­ CRediT authorship contribution statement
politan experiences and ways of being, while local community needs in
rural locations are not adequately addressed (Roberts, 2014). Indeed, Andrew A. Kingsford-Smith: Writing – review & editing, Writing –
some national standardized tests may afford cultural capital advantages original draft, Formal analysis, Conceptualization. Dennis Alonzo:
to students in metropolitan schools (Frawley & McLean Davies, 2015). If Writing – original draft, Supervision, Methodology, Formal analysis,
the curriculum is perceived to be irrelevant by students in rural areas, Conceptualization. Kim Beswick: Writing – review & editing, Supervi­
then teachers can be caught in a bind between student disengagement sion, Project administration, Methodology, Funding acquisition. Tony
linked to perceived irrelevance and the impulse of the teacher to assert Loughland: Writing – review & editing, Supervision, Investigation.
control over the classroom. Ensuring that school learning is relevant to Philip Roberts: Writing – review & editing, Project administration,
all students, no matter where they live, may help students develop Methodology, Funding acquisition, Conceptualization.
greater agency in their education and help reduce gaps in academic and
school outcomes (Australian Curriculum, Assessment and Reporting
Authority [ACARA], 2019). Declaration of competing interest

The authors have no conflict of interest to declare.

10
A.A. Kingsford-Smith et al. Teaching and Teacher Education 142 (2024) 104516

Data availability Centre for Education Statistics and Evaluation [CESE]. (2015). Student engagement and
wellbeing in NSW. https://education.nsw.gov.au/about-us/education-data-and-resea
rch/cese/publications/research-reports/student-engagement-and-wellbeing-in-nsw.
The data that has been used is confidential. Centre for Education Statistics and Evaluation [CESE]. (2017). Tell them from me: Gender
and engagement. https://education.nsw.gov.au/about-us/education-data-and-rese
arch/cese/publications/research-reports/tell-them-from-me-gender-and-engageme
Acknowledgments nt.
Chan, S., Maneewan, S., & Koul, R. (2021). Teacher educators’ teaching styles: Relation
This research was undertaken as part of the Rural and Regional Ed­ with learning motivation and academic engagement in pre-service teachers. Teaching
in Higher Education, 1–22. https://doi.org/10.1080/13562517.2021.1947226
ucation project, which was commissioned by the NSW Department of
Chemers, M. M., Hu, L. T., & Garcia, B. F. (2001). Academic self-efficacy and first-year
Education, and completed by the Gonski Institute for Education, the college student performance and adjustment. Journal of Educational Psychology, 93
School of Education at UNSW Sydney, Social Ventures Australia, and the (1), 55–64. https://doi.org/10.1037/0022-0663.93.1.55
Chen, F. F. (2007). Sensitivity of goodness of fit indexes to lack of measurement
University of Canberra.
invariance. Structural Equation Modeling, 14(3), 464–504. https://doi.org/10.1080/
10705510701301834
References Cheon, S. H., Reeve, J., & Song, Y. (2016). A teacher-focused intervention to decrease PE
students’ amotivation by increasing need satisfaction and decreasing need
frustration. Journal of Sport & Exercise Psychology, 38(3), 217–235. https://doi.org/
af Ursin, P., Järvinen, T., & Pihlaja, P. (2021). The role of academic buoyancy and social
10.1123/jsep.2015-0236
support in mediating associations between academic stress and school engagement
Cheung, G. W., & Rensvold, R. B. (2002). Evaluating goodness-of-fit indexes for testing
in Finnish primary school children. Scandinavian Journal of Educational Research, 65
measurement invariance. Structural Equation Modeling, 9(2), 233–255. https://doi.
(4), 661–675. https://doi.org/10.1080/00313831.2020.1739135
org/10.1207/S15328007SEM0902_5
Archambault, I., Janosz, M., & Chouinard, R. (2012). Teacher beliefs as predictors of
Collie, R. J., Bostwick, K. C. P., & Martin, A. J. (2020). Perceived autonomy support,
adolescents’ cognitive engagement and achievement in mathematics. Journal of
relatedness with students, and workplace outcomes: An investigation of differences
Educational Research, 105(5), 319–328. https://doi.org/10.1080/
by teacher gender. Educational Psychology, 40(3), 253–272. https://doi.org/
00220671.2011.629694
10.1080/01443410.2019.1663791
Australian Bureau of Statistics. (2021). Remoteness structure: Australian statistical
Colmar, S., Liem, G. A. D., Connor, J., & Martin, A. J. (2019). Exploring the relationships
geography standard (ASGS). Australian Bureau of Statistics. https://www.abs.gov.
between academic buoyancy, academic self-concept, and academic performance: A
au/statistics/standards/australian-statistical-geography-standard-asgs-edition-3/
study of mathematics and reading among primary school students. Educational
jul2021-jun2026/remoteness-structure.
Psychology, 39(8), 1068–1089. https://doi.org/10.1080/01443410.2019.1617409
Australian Curriculum, Assessment and Reporting Authority [ACARA]. (2019).
Corbett, M. (2020). Place-based education: A critical appraisal from a rural perspective.
Achievement in reading, writing, language conventions and numeracy: National report for
In M. Corbett (Ed.), Rural teacher education. Springer.
2019. Australian curriculum, assessment and reporting authority. https://www.nap.
Crosnoe, R., Johnson, M. K., & Elder, G. H. (2004). School size and the interpersonal side
edu.au/results-and-reports/national-reports.
of education: An examination of race/ethnicity and organizational context. Social
Aydın, G., & Michou, A. (2020). Self-determined motivation and academic buoyancy as
Science Quarterly, 85(5 SPEC. ISS.), 1259–1274. https://doi.org/10.1111/j.0038-
predictors of achievement in normative settings. British Journal of Educational
4941.2004.00275.x
Psychology, 90(4), 964–980. https://doi.org/10.1111/bjep.12338
Das, D. (2019). Academic resilience among children from disadvantaged social groups in
Bandura, A. (1986). Social foundations of thought and action: A social cognitive theory.
India. Social Indicators Research, 145(2), 719–739. https://doi.org/10.1007/s11205-
Englewood Cliffs, NJ: Prentice-Hall.
018-1899-y
Bandura, A. (2001). Social cognitive theory: An agentic perspective. Annual Review of
Dweck, C. S., & Master, A. (2007). Self-theories and motivation: Students’ beliefs about
Psychology, 52(1), 1–26. https://doi.org/10.1146/annurev.psych.52.1.1
intelligence. In Handbook of motivation at school (pp. 123–140). New York, NY:
Bandura, A. (2006). Guide for constructing self-efficacy scales. In F. Pajares, &
Routledge.
T. C. Urdan (Eds.), Self-efficacy beliefs of adolescents (pp. 307–337). Greenwich, CT:
Eccles, J. S., Midgley, C., Wigfield, L., Buchanan, C. M., Reuman, D., Flanagan, C., & Mac
Information Age.
Iver, D. (1993). Development during adolescence: The impact of stage-environment
Bandura, A., & Jourden, F. J. (1991). Self-regulatory mechanisms governing the impact
fit on young adolescents’ experiences in schools and in families. American
of social comparison on complex decision making. Journal of Personality and Social
Psychologist, 48(2), 90–101. https://doi.org/10.1037/0003-066x.48.2.90
Psychology, 60(6), 941–951. https://doi.org/10.1037//0022-3514.60.6.941
Echazarra, A., & Radinger, T. (2019). Learning in rural schools: Insights from PISA, TALIS
Benita, M., Roth, G., & Deci, E. L. (2014). When are mastery goals more adaptive? It
and the literature. https://doi.org/10.1787/8b1a5cb9-en
depends on experiences of autonomy support and autonomy. Journal of Educational
Frawley, E., & McLean Davies, L. (2015). Assessing the field: Students and teachers of
Psychology, 106(1), 258–267. https://doi.org/10.1037/a0034007
writing in high-stakes literacy testing in Australia. English Teaching: Practice and
Bergin, C., & Bergin, D. (2009). Attachment in the classroom. Educational Psychology
Critique, 14(2), 83–99. https://doi.org/10.1108/etpc-01-2015-0001
Review, 21(2), 141–170. https://doi.org/10.1007/s10648-009-9104-0
Gillet, N., Vallerand, R. J., & Lafrenière, M. A. K. (2012). Intrinsic and extrinsic school
Berry, J. M., & West, R. L. (1993). Cognitive self-efficacy in relation to personal mastery
motivation as a function of age: The mediating role of autonomy support. Social
and goal setting across the life span. International Journal of Behavioral Development,
Psychology of Education, 15, 77–95. https://doi.org/10.1007/s11218-011-9170-2
16(2), 351–379. https://doi.org/10.1177/016502549301600213
Granziera, H., Liem, G. A. D., Chong, W. H., Martin, A. J., Collie, R. J., Bishop, M., &
Beswick, K., Roberts, P., Eacott, S., Holden, R., Alonzo, D., Downes, N., Mularcyzk, L.,
Tynan, L. (2022). The role of teachers’ instrumental and emotional support in
Loughland, T., Cridge, S., Corbyn, E., & Bedford, M. (2023). Rural and regional
students’ academic buoyancy, engagement, and academic skills: A study of high
education project: Final report. https://www.gie.unsw.edu.au/rural-and-regional-e
school and elementary school students in different national contexts. Learning and
ducation-project-final-report.
Instruction, 80, 101619. https://doi.org/10.1016/j.learninstruc.2022.101619.
Beswick, K., Wright, S., Watson, J., Hay, I., Allen, J., & Cranston, N. (2019). Teachers’
Gravetter, F. J., & Forzano, L.-A. B. (2012). Research methods for the behavioral sciences
beliefs related to secondary school completion: Associations with socio-educational
(4th ed.). Belmont, CA, USA: Wadsworth Cengenge Learning.
advantage and school level. Australian Educational Researcher, 46(5), 751–774.
Green, B. (2008). The NSW rural (teacher) education project: Introduction and overview.
https://doi.org/10.1007/s13384-019-00317-3
In B. Green (Ed.), Spaces and places: The NSW rural (teacher) education project. Centre
Bong, M. (2001). Role of self-efficacy and task-value in predicting college students’
for Information Studies Press.
course performance and future enrollment intentions. Contemporary Educational
Green, J., Martin, A. J., & Marsh, H. W. (2007). Motivation and engagement in English,
Psychology, 26(4), 553–570. https://doi.org/10.1006/ceps.2000.1048
mathematics and science high school subjects: Towards an understanding of
Booth, M. Z., & Gerard, J. M. (2014). Adolescents’ stage-environment fit in middle and
multidimensional domain specificity. Learning and Individual Differences, 17(3),
high school: The relationship between students’ perceptions of their schools and
269–279. https://doi.org/10.1016/j.lindif.2006.12.003
themselves. Youth & Society, 46(6), 735–755. https://doi.org/10.1177/
Gutiérrez, M., & Tomás, J. M. (2019). The role of perceived autonomy support in
0044118X12451276
predicting university students’ academic success mediated by academic self-efficacy
Bostwick, K. C. P., Martin, A. J., Collie, R. J., Burns, E. C., Hare, N., Cox, S., Flesken, A., &
and school engagement. Educational Psychology, 39(6), 729–748. https://doi.org/1
McCarthy, I. (2022). Academic buoyancy in high school: A cross-lagged multilevel
0.1080/01443410.2019.1566519.
modeling approach exploring reciprocal effects with perceived school support,
Halsey, J. R. (2018). Independent review into regional, rural and remote education: Literature
motivation, and engagement. Journal of Educational Psychology, 114(8), 1931–1949.
review. Commonwealth of Australia. https://www.education.gov.au/quality-schools
https://doi.org/10.1037/edu0000753
-package/resources/independent-review-regional-rural-and-remote-education-liter
Boylan, C. (1988). Research on the provision of education in the New South Wales
ature-review.
central school. Research in Rural Education, 5(2), 43–48. https://jrre.psu.edu/site
Handal, B., Watson, K., Petocz, P., & Maher, M. (2018). Choosing to teach in rural and
s/default/files/2019-07/5-2_7.pdf.
remote schools: The zone of free movement. Education Research & Perspectives, 45,
Burns, E. C., Martin, A. J., & Collie, R. J. (2017). Understanding the role of adaptability
1–32. https://www.erpjournal.net/wp-content/uploads/2020/02/01_ERPV45_H
and personal best (PB) goals in students ’ academic outcomes: A social cognitive
andal.pdf.
perspective. BJEP Monograph Series II, 12, 111–143. https://doi.org/10.53841/
Hardre, P. L., & Reeve, J. (2003). A motivational model of rural students’ intentions to
bpsmono.2017.cat2023.8
persist in, versus drop out of, high school. Journal of Educational Psychology, 95(2),
Burns, E. C., Martin, A. J., & Collie, R. J. (2018). Adaptability, personal best (PB) goals
347–356. https://doi.org/10.1037/0022-0663.95.2.347
setting, and gains in students’ academic outcomes: A longitudinal examination from
a social cognitive perspective. Contemporary Educational Psychology, 53, 57–72.
https://doi.org/10.1016/j.cedpsych.2018.02.001

11
A.A. Kingsford-Smith et al. Teaching and Teacher Education 142 (2024) 104516

Honicke, T., & Broadbent, J. (2016). The influence of academic self-efficacy on academic of Behavioral Development, 44(4), 301–312. https://doi.org/10.1177/
performance: A systematic review. Educational Research Review, 17, 63–84. https:// 0165025419885027
doi.org/10.1016/j.edurev.2015.11.002 Martin, A. J., Yu, K., Ginns, P., & Papworth, B. (2017). Young people’s academic
Huang, C. (2013). Gender differences in academic self-efficacy: A meta-analysis. buoyancy and adaptability: A cross-cultural comparison of China with north America
European Journal of Psychology of Education, 28(1), 1–35. https://doi.org/10.1007/ and the United Kingdom. Educational Psychology, 37(8), 930–946. https://doi.org/
s10212-011-0097-y 10.1080/01443410.2016.1202904
Hughes, J. N., & Chen, Q. (2011). Reciprocal effects of student-teacher and student-peer Martinek, D., & Kipman, U. (2016). Self-determination, self-efficacy and self-regulation
relatedness: Effects on academic self efficacy. Journal of Applied Developmental in school: A longitudinal intervention study with primary school pupils. Sociology
Psychology, 32(5), 278–287. https://doi.org/10.1016/j.appdev.2010.03.005. Study, 6(2), 124–133. https://doi.org/10.17265/2159-5526/2016.02.005
Reciprocal McKenzie, P., Weldon, P., Rowley, G., Murphy, M., & McMillan, J. (2014). Staff in
IBM. (2022). IBM SPSS Statistics 26 documentation. https://www.ibm.com/support/pages Australia’s schools 2013: Main report on the survey. Australian Council for Educational
/ibm-spss-statistics-26-documentation. Research. https://research.acer.edu.au/tll_misc/20/.
Jang, H., Kim, E. J., & Reeve, J. (2012). Longitudinal test of self-determination theory’s McMillan, J., Beavis, A., & Jones, F. L. (2009). The AUSEI06. Journal of Sociology, 45(2),
motivation mediation model in a naturally occurring classroom context. Journal of 123–149. https://doi.org/10.1177/1440783309103342
Educational Psychology, 104(4), 1175–1188. https://doi.org/10.1037/a0028089 Moè, A., Consiglio, P., & Katz, I. (2022). Exploring the circumplex model of motivating
Jang, H., Kim, E. J., & Reeve, J. (2016). Why students become more engaged or more and demotivating teaching styles: The role of teacher need satisfaction and need
disengaged during the semester: A self-determination theory dual-process model. frustration. Teaching and Teacher Education, 118. https://doi.org/10.1016/j.
Learning and Instruction, 43, 27–38. https://doi.org/10.1016/j. tate.2022.103823
learninstruc.2016.01.002 Multan, K. D., Brown, S. D., & Lent, R. W. (1991). Relation of self-efficacy beliefs to
Jenkins, K., Taylor, N., & Reitano, P. (2015). Listening to teachers in the ‘bush.’. In academic outcomes: A meta-analytic investigation. Journal of Counseling Psychology,
L. Graham, & J. Miller (Eds.), Bush tracks: The opportunities and challenges of rural 38(1), 30–38.
teaching and leadership (pp. 41–55). Sense Publishers. Muthén, L. K., & Muthén, B. O. (2017). Mplus user’s guide: Eighth edition. Muthén &
Karlberg-Granlund, G. (2023). The heart of the small Finnish rural school: Supporting Muthén. https://www.statmodel.com/download/usersguide/MplusUserGuideVer_8.
roots and wings, solidarity and autonomy. In K. E. Reimer, M. Kaukko, S. Windsor, pdf.
K. Mahon, & S. Kemmis (Eds.), Living well in a world worth living in for all. Volume 1: Nurmi, J. E., & Kiuru, N. (2015). Introduction to the special section on student effects on
Current practices of social justice, sustainability and wellbeing (pp. 47–67). Springer. teachers behaviors and attitudes. International Journal of Behavioral Development, 39
https://doi.org/10.1007/978-981-19-7985-9_4. (5), 389–390. https://doi.org/10.1177/0165025415591231
Keith, T. Z. (2015). Multiple regression and beyond: An introduction to multiple regression Orakci, Ş., & Durnali, M. (2023). The mediating effects of metacognition and creative
and structural equation modeling. Routledge. thinking on the relationship between teachers’ autonomy support and teachers’ self-
Khotimah, K., Basthomi, Y., & Eliyanah, E. (2023). EFL pre-service teacher development efficacy. Psychology in the Schools, 60(1), 162–181. https://doi.org/10.1002/
for autonomy: Rethinking future autonomy supportive teacher. Proceedings of the pits.22770
20th AsiaTEFL-68th TEFLIN-5th INELTAL Conference, 321–329. https://doi.org/ Organisation for Economic Co-operation and Development [OECD]. (2018). PISA 2018
10.2991/978-2-38476-054-1_27 results (Volume IV): Are students smart about money? (Vol. IV). Paris: PISA, OECD
Kleinkorres, R., Stang-Rabrig, J., & McElvany, N. (2023). The longitudinal development Publishing. https://doi.org/10.1787/48ebd1ba-en
of students’ well-being in adolescence: The role of perceived teacher autonomy Pajares, F., Johnson, M. J., & Usher, E. L. (2007). Sources of writing self-efficacy beliefs
support. Journal of Research on Adolescence, 33(2), 496–513. https://doi.org/ of elementary, middle, and high school students. Research in the Teaching of English,
10.1111/jora.12821 42(1), 104–120.
Lamb, S., Maire, Q., Doecke, E., Macklin, S., Noble, K., & Pilcher, S. (2020). Impact of Papatraianou, L. H., Strangeways, A., Beltman, S., & Schuberg Barnes, E. (2018).
learning from home on educational outcomes from disadvantaged children. Beginning teacher resilience in remote Australia: A place-based perspective. Teachers
Lei, H., Cui, Y., & Chiu, M. M. (2018). The relationship between teacher support and and Teaching: Theory and Practice, 24(8), 893–914. https://doi.org/10.1080/
students’ academic emotions: A meta-analysis. Frontiers in Psychology, 8(JAN), 1–12. 13540602.2018.1508430
https://doi.org/10.3389/fpsyg.2017.02288 Pelletier, L. G., Séguin-Lévesque, C., & Legault, L. (2002). Pressure from above and
Lester, L., & Cross, D. (2015). The relationship between school climate and mental and pressure from below as determinants of teachers’ motivation and teaching
emotional wellbeing over the transition from primary to secondary school. behaviors. Journal of Educational Psychology, 94(1), 186–196. https://doi.org/
Psychology of Well-Being, 5(1). https://doi.org/10.1186/s13612-015-0037-8 10.1037/0022-0663.94.1.186
Lewis, R. (2006). Classroom discipline in Australia. In C. M. Evertson, & C. S. Weinstein Pietsch, B., & Williamson, J. (2008). The loneliness of the long distance principal: The
(Eds.), Handbook of classroom management (pp. 1193–1214). Routledge. https://doi. tension between satisfying student needs and government requirements. AARE
org/10.4324/9780203874783.ch46. Annual 2008 International Education Research Conference. https://figshare.utas.edu.
Lietaert, S., Roorda, D., Laevers, F., Verschueren, K., & De Fraine, B. (2015). The gender au/articles/conference_contribution/The_loneliness_of_the_long_distance_principal_
gap in student engagement: The role of teachers’ autonomy support, structure, and The_experience_of_principals_in_isolated_central_schools/23085506.
involvement. British Journal of Educational Psychology, 85(4), 498–518. https://doi. Reeve, J. (2009). Why teachers adopt a controlling motivating style toward students and
org/10.1111/bjep.12095 how they can become more autonomy supportive. Educational Psychologist, 44(3),
Lock, G., Budgen, F., Lunay, R., & Oakley, G. (2012). Welcome to the outback: The 159–175. https://doi.org/10.1080/00461520903028990
paradoxes of living and teaching in remote Western Australian schools. Australian Reeve, J. (2012). A self-determination theory perspective on student engagement. In
and International Journal of Rural Education, 22(3), 117–134. https://search.informit. S. L. Christenson, C. Wylie, & A. L. Reschly (Eds.), Handbook of research on student
org/doi/10.3316/informit.084614878498362. engagement (pp. 601–634). Springer.
Mahatmya, D., Lohman, B. J., Matjasko, J. L., & Farb, A. F. (2012). Engagement across Reeve, J. (2016). Autonomy-supportive teaching: What it is, how to do it. In W. C. Liu,
developmental periods. In S. L. Christenson, C. Wylie, & A. L. Reschly (Eds.), J. C. K. Wang, & R. M. Ryan (Eds.), Building autonomous learners (pp. 129–152).
Handbook of research on student engagement (pp. 45–64). New York, USA: Springer. Singapore: Springer.
Martin, A. J. (2009). Motivation and engagement across the academic life span: A Reeve, J., Cheon, S. H., & Jang, H.-R. (2019). A teacher-focused intervention to enhance
developmental construct validity study of elementary school, high school, and students’ classroom engagement. In J. A. Fredricks, A. L. Reschly, & S. L. Christenson
university/college students. Educational and Psychological Measurement, 69(5), (Eds.), Handbook of student engagement interventions. Academic Press.
794–824. https://doi.org/10.1177/0013164409332214 Reeve, J., & Shin, S. H. (2020). How teachers can support students’ agentic engagement.
Martin, A. J., Colmar, S. H., Davey, L. A., & Marsh, H. W. (2010). Longitudinal modelling Theory Into Practice, 59(2), 150–161. https://doi.org/10.1080/
of academic buoyancy and motivation: Do the “5Cs” hold up over time? British 00405841.2019.1702451
Journal of Educational Psychology, 80(3), 473–496. https://doi.org/10.1348/ Roberts, P. (2013). The role of an authentic curriculum and pedagogy for rural schools
000709910X486376 and the professional satisfaction of rural teachers. Australian and International Journal
Martin, A. J., Ginns, P., Papworth, B., & Nejad, H. G. (2013). The role of academic of Rural Education, 23(2), 89–99. https://search.informit.org/doi/abs/10.3316/aei
buoyancy in Aboriginal/Indigenous students’ educational intentions: Sowing the pt.199280.
early seeds of success for post-school education and training. Vol. 14). In Roberts, P. (2014). Education for rural Australia. In A. Hogan, & M. Young (Eds.), Rural
R. G. Mooney, & J. C (Eds.). Seeding success in indigenous Australian higher education and regional futures (pp. 117–134). New York: Routledge. https://doi.org/10.4324/
(diversity in higher education (Vol. 14). Emerald Gr, 0–32. 9781315775333-20.
Martin, A. J., & Marsh, H. W. (2008a). Academic buoyancy: Towards an understanding of Roberts, P., & Fuqua, M. (2021). Ruraling education research. In P. Roberts, & M. Fuqua
students’ everyday academic resilience. Journal of School Psychology, 46(1), 53–83. (Eds.), Ruraling education research: Connections between rurality and the disciplines of
https://doi.org/10.1016/j.jsp.2007.01.002 educational research. Springer.
Martin, A. J., & Marsh, H. W. (2008b). Workplace and academic buoyancy: Psychometric Roberts, P., & Guenther, J. (2021). Framing rural and remote: Key issues, debates,
assessment and construct validity amongst school personnel and students. Journal of definitions, and positions in constructing rural and remote disadvantage. In
Psychoeducational Assessment, 26(2), 168–184. https://doi.org/10.1177/ P. Roberts, & M. Fuqua (Eds.), Ruraling education research: Connections between
0734282907313767 rurality and the disciplines of educational research. Springer.
Martin, A. J., & Marsh, H. W. (2009). Academic resilience and academic buoyancy: Rojas Flórez, L. F. (2015). Factors affecting academic resilience in middle school
Multidimensional and hierarchical conceptual framing of causes, correlates and students: A case study. GiST Education and Learning Research Journal, 11(11), 63–78.
cognate constructs. Oxford Review of Education, 35(3), 353–370. https://doi.org/ https://doi.org/10.26817/16925777.286
10.1080/03054980902934639 Roorda, D. L., Koomen, H. M. Y., Spilt, J. L., & Oort, F. J. (2011). The influence of
Martin, A. J., & Marsh, H. W. (2020). Investigating the reciprocal relations between affective teacher-student relationships on students’ school engagement and
academic buoyancy and academic adversity: Evidence for the protective role of achievement: A meta-analytic approach. Review of Educational Research, 81(4),
academic buoyancy in reducing academic adversity over time. International Journal 493–529. https://doi.org/10.3102/0034654311421793

12
A.A. Kingsford-Smith et al. Teaching and Teacher Education 142 (2024) 104516

Rosenthal, B. S. (1998). Non-school correlates of dropout: An integrative review of the Thomson, S., De Bortoli, L., Underwood, C., & Schmid, M. (2020). Pisa 2018: Reporting
literature. Children and Youth Services Review, 20(5), 413–433. https://doi.org/ Australia’s results. Volume II student and school characteristics. Australian Council
10.1016/S0190-7409(98)00015-2 for Educational Research [ACER]. https://research.acer.edu.au/ozpisa/49/.
Ryan, R. M., & Deci, E. L. (2000). Self-determination theory and the facilitation of Uçar, F. M., & Sungur, S. (2017). The role of perceived classroom goal structures, self-
intrinsic motivation, social development, and well-being. American Psychologist, 55 efficacy, and engagement in student science achievement. Research in Science &
(1), 68–78. https://doi.org/10.1037/0003-066X.55.1.68. Technological Education, 35(2), 149–168. https://doi.org/10.1080/
Ryan, R. M., & Deci, E. L. (2017). Self-determination theory: Basic psychological needs in 02635143.2017.1278684
motivation, development, and wellness. Guilford Publications. Valentine, J. C., DuBois, D. L., & Cooper, H. (2004). The relation between self-beliefs and
Sarrazin, P. G., Tessier, D. P., Pelletier, L. G., Trouilloud, D. O., & Chanal, J. P. (2006). academic achievement: A meta-analytic review. Educational Psychologist, 39(2),
The effects of teachers’ expectations about students’ motivation on teachers’ 111–133. https://doi.org/10.1207/s15326985ep3902_3
autonomy-supportive and controlling behaviors. International Journal of Sport and van de Schoot, R., Lugtig, P., & Hox, J. (2012). A checklist for testing measurement
Exercise Psychology, 4(3), 283–301. https://doi.org/10.1080/ invariance. European Journal of Developmental Psychology, 9(4), 486–492. https://
1612197X.2006.9671799 doi.org/10.1080/17405629.2012.686740
Schuitema, J., Peetsma, T., & Van der Veen, I. (2016). Longitudinal relations between Watt, H. M. G. (2004). Development of adolescents’ self-perceptions, values, and task
perceived autonomy and social support from teachers and students’ self-regulated perceptions according to gender and domain in 7th- through 11th-grade Australian
learning and achievement. Learning and Individual Differences, 49, 32–45. https://doi. students. Child Development, 75(5), 1556–1574. https://doi.org/10.1111/j.1467-
org/10.1016/j.lindif.2016.05.006 8624.2004.00757.x
Schunk, D. H., & Mullen, C. A. (2012). Self-efficacy as an engaged learner. In Waxman, H., & Huang, S.-Y. (1998). Classroom learning environments in urban
S. L. Christenson, C. Wylie, & A. L. Reschly (Eds.), Handbook of research on student elementary, middle, and high schools. Learning Environments Research, 1(1), 95–113.
engagement (pp. 219–235). New York, USA: Springer. https://doi.org/10.1007/978- https://doi.org/10.1023/A:1009940816549
1-4614-2018-7. Weißenfels, M., Hoffmann, D., Dörrenbächer-Ulrich, L., & Perels, F. (2023). Linking
Simões, F., & Calheiros, M. M. (2023). Multiple autonomy support attunement academic buoyancy and math achievement in secondary school students: Does
connections with perceived competence in learning and school grades among rural academic self-efficacy play a role? Current Psychology, 42(27), 23422–23436.
adolescents. Current Psychology, 42(3), 1687–1700. https://doi.org/10.1007/ https://doi.org/10.1007/s12144-022-03488-y
s12144-021-01557-2 Wettersten, K. B., Guilmino, A., Herrick, C. G., Hunter, P. J., Kim, G. Y., Jagow, D., …
Skinner, E. A., Zimmer-Gembeck, M. J., Connell, J. P., Eccles, J. S., & Wellborn, J. G. McCormick, J. (2005). Predicting educational and vocational attitudes among rural
(1998). Individual differences and the development of perceived control. high school students. Journal of Counseling Psychology, 52(4), 658–663. https://doi.
Monographs of the Society for Research in Child Development, 63(2). https://doi.org/ org/10.1037/0022-0167.52.4.658.
10.2307/1166220 What Works National Office National Curriculum Services. (2012). Success in remote
Smith, L., & Sinclair, K. (2000). Transforming the HSC: Affective implications. Change: schools: A research study of eleven improving remote schools. Commonweath of
Transformations in Education, 3(2), 67–79. http://hdl.handle.net/2123/4376. Australia. https://www.nintione.com.au/resources/nol/success-in-remote-schools
Solomon, D., Battistich, V., & Hom, A. (1996). Teacher beliefs and practices in schools -a-research-study-of-eleven-improving-remote-schools/.
serving communities that differ in socioeconomic level. The Journal of Experimental Wigfîeld, A., & Eccles, J. S. (1994). Children’s competence beliefs, achievement values,
Education, 64(4), 327–347. https://doi.org/10.1080/00220973.1996.10806602 and general self-esteem: Change across elementary and middle school. In The Journal
Stockard, J. (2011). Increasing reading skills in rural areas: An analysis of three school of Early Adolescence, 14 pp. 107–138). https://doi.org/10.1177/
districts. Journal of Research in Rural Education, 26(8), 1–19. http://www.cac.psu. 027243169401400203
edu/dept/jrre/articles/26-8.pdf%5Cn. Williams, G., & Deci, E. L. (1996). Internalisation of biopsychosocial values by medical
Talsma, K., Schüz, B., Schwarzer, R., & Norris, K. (2018). I believe, therefore I achieve students: A test of self determination theory. Journal of Personality and Social
(and vice versa): A meta-analytic cross-lagged panel analysis of self-efficacy and Psychology, 70(4), 767–779. https://doi.org/10.1037//0022-3514.70.4.767
academic performance. Learning and Individual Differences, 61(April 2017), 136–150. Young, D. J. (2000). Rural differences in student achievement: The effect of student
https://doi.org/10.1016/j.lindif.2017.11.015 perceptions. Educational Research and Evaluation, 6(3), 207–228. https://doi.org/
Tarbetsky, A. L., Martin, A. J., & Collie, R. J. (2017). Social and emotional learning, 10.1076/1380-3611(200009)6:3;1-a;ft207
social and emotional competence, and students’ academic outcomes: The roles of Zhao, J., & Qin, Y. (2021). Perceived teacher autonomy support and students’ deep
psychological need satisfaction, adaptability, and buoyancy. In E. Frydenberg, learning: The mediating role of self-efficacy and the moderating role of perceived
A. J. Martin, & R. J. Collie (Eds.), Social and emotional learning in Australia and the peer support. Frontiers in Psychology, 12(June), 1–11. https://doi.org/10.3389/
asia-pacific: Perspectives, programs and approaches (pp. 1–471). https://doi.org/ fpsyg.2021.652796
10.1007/978-981-10-3394-0 Zhou, M., Ma, W. J., & Deci, E. L. (2009). The importance of autonomy for rural Chinese
Tawfik, A. A., Rong, H., & Choi, I. (2015). Failing to learn: Towards a unified design children’s motivation for learning. Learning and Individual Differences, 19(4),
approach for failure-based learning. Educational Technology Research & Development, 492–498. https://doi.org/10.1016/j.lindif.2009.05.003
63(6), 975–994. https://doi.org/10.1007/s11423-015-9399-0

13