Computers and Education: Artificial Intelligence 6 (2024) 100224

Contents lists available at ScienceDirect

Computers and Education: Artificial Intelligence

journal homepage: [Link]/journal/computers-and-education-artificial-intelligence

Teachers’ motivation and engagement to harness generative AI for teaching

and learning: The role of contextual, occupational, and background factors
Rebecca J. Collie *, Andrew J. Martin
School of Education, University of New South Wales, Australia

A R T I C L E I N F O A B S T R A C T

Keywords: Since their release in late 2022, generative AI (genAI) tools have led to widespread use, including among
Generative AI teachers. The aim of our study is to examine several factors that may be implicated in teachers’ motivation and
Teachers engagement to harness genAI in teaching and learning. We examined contextual (i.e., autonomy-supportive
Motivation
leadership, autonomy-thwarting leadership), occupational experience (i.e., professional growth striving,
Engagement
Integration
change-related stress), and background factors (i.e., gender, age, teaching experience, contract length, class size,
Autonomy-support school level) as predictors of motivation (i.e., genAI valuing) and, in turn, engagement (i.e., integration in
teaching-related work and student learning activities). Among 339 Australian teachers, our findings revealed that
perceived autonomy-supportive leadership, professional growth striving, and change-related stress were linked
with greater genAI valuing. In turn, genAI valuing was associated with greater genAI integration in both
teaching-related work and student learning activities. Perceived autonomy-thwarting leadership was directly
linked with greater genAI integration in student learning activities, and professional growth striving was directly
associated with greater genAI integration in teaching-related work. Teachers’ gender and school level were also
linked with the motivation and engagement factors, and there were several indirect associations as well. Our
results pinpoint areas of focus for future research, policy, and practice to support genAI and its application in
schools.

1. Introduction and make efforts to integrate this technology in their work. Indeed,
teachers who do not value genAI are unlikely to integrate it effectively or
Generative artificial intelligence (genAI) refers to technology that to continue to develop their skills related to this continuingly evolving
uses machine learning techniques to generate new content, including technology. Given the novelty of genAI tools at an accessible scale,
text, images, and audio (Lim et al., 2023). GenAI tools released from late however, research is currently limited on this topic.
2022 onwards (e.g., ChatGPT, Midjourney, Dall-E 2) have inspired rapid The aim of the present study, therefore, is to examine several factors
uptake among users across the globe (Lim et al., 2023), and this includes that may be implicated in teachers’ motivation and engagement to
teachers (Chiu, 2023). Because genAI has already impacted, and will harness genAI in teaching and learning. For motivation, we examined
continue to impact, the jobs of knowledge workers (Carvalho et al., the extent to which teachers value genAI as a tool for conducting their
2022), it is essential to understand what factors are associated with work, and for engagement we examined teachers’ integration of genAI
teachers’ motivation and engagement towards this technology. More­ in both their own teaching-related work and student learning activities.
over, emerging policies from educational bodies and jurisdictions sug­ Drawing together understanding from several theoretical frameworks,
gest that genAI will be embedded within schooling (e.g., Australian we focus on contextual factors, occupational experiences, and back­
Government, 2023; US Office of Educational Technology, 2023; UK ground factors assessed at the beginning of a school term as predictors of
Department for Education, 2023)—further highlighting the need to teachers’ genAI-related motivation and, in turn, genAI-related engage­
understand what factors are associated with teachers’ motivation and ment towards the end of the same school term. Specifically, we assessed
engagement to implement genAI in teaching and learning. Indeed, if two contextual predictors: autonomy-supportive leadership and
schools and educational jurisdictions want genAI to be applied—as early autonomy-thwarting leadership. We examined two occupational expe­
policy is suggesting—then it is essential that teachers see the value in rience predictors: professional growth striving and change-related

* Corresponding author. School of Education, University of New South Wales, NSW, 2052, Australia.
E-mail address: [Link]@[Link] (R.J. Collie).

[Link]
Received 11 December 2023; Received in revised form 7 April 2024; Accepted 14 April 2024
Available online 15 April 2024
2666-920X/© 2024 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license ([Link]
R.J. Collie and A.J. Martin Computers and Education: Artificial Intelligence 6 (2024) 100224

stress. Our background factors comprised personal and job characteris­ (e.g., using genAI to save time in lesson planning). Together then,
tics: gender, age, teaching experience, contract length, class size, and broaden-and-build theory (Fredrickson, 2001) and conservation of re­
school level. Fig. 1 demonstrates the hypothesized model. Findings have sources theory (Hobfoll, 1989) highlight the importance of occupational
the potential to inform understanding about why teachers may be more experiences for subsequent motivation and engagement. However, the
or less motivated and engaged when it comes to genAI, and as such are pathways vary by which occupational experiences predict motivation
relevant for guiding efforts to better support teachers to apply this new and engagement in these two frameworks. Broaden-and-build theory
technology effectively. Before going into further detail about the study’s indicates that adaptive experiences lead to more positive beliefs about
main factors and their hypothesized links, we first outline the concep­ resources and actual access of resources (and that maladaptive experi­
tual underpinnings that guided our examination. ences lead to the reverse), whereas conservation of resources theory
posits that maladaptive experiences lead to more positive beliefs about
resources and access of resources. These two conceptual models thus
1.1. Conceptual framing
provide understanding of two different processes through which
teachers’ occupational experiences may be linked to motivation and
We refer to several conceptual frameworks to understand the role of
engagement.
contextual, occupational experience, and background factors. Starting
Although background factors, such as personal and job characteris­
with contextual factors, theories highlight the importance of these for
tics, are generally not core considerations in conceptual work, there is
impacting employee functioning at work. For example, self-
emerging interest in this area. Job demands-resources theory now rec­
determination theory (Ryan & Deci, 2017) articulates the role of
ognizes that these characteristics can act as factors relevant to employee
autonomy-supportive leadership practices (that support teachers’ pro­
functioning (Demerouti & Bakker, 2023). For example, years of teaching
fessional agency) and autonomy-thwarting leadership practices (that
experience is relevant to the accumulated knowledge that a teacher has
make teachers feel pressured at work). Whereas autonomy-supportive
and may have implications for their motivation and engagement in
leadership is linked with positive employee functioning (e.g., motiva­
different ways (e.g., early career teachers may be more motivated to
tion and engagement), the reverse is true for autonomy-thwarting
learn new skills, or less engaged in tasks they have relatively little
leadership (Ryan et al., 2022).
knowledge about). As another example, job insecurity from a fixed-term
Another relevant conceptual model is job demands-resources theory
contract has been recognized as a type of demand that can lead to
(Bakker & Demerouti, 2017), which identifies that there are job re­
resource loss among employees (Demerouti & Bakker, 2023).
sources that help employees to be motivated and engaged at work, and
Together, the theoretical underpinnings described above provide
job demands that hinder employees’ motivation and engagement at
support for the role of contextual factors, occupational experiences, and
work. Indeed, there have been recent efforts aimed at integrating these
background factors as relevant to teachers’ motivation and engagement.
perspectives into unified models of human flourishing that emphasize
In the following section, we introduce the specific variables that we
the role of contextual resources and demands in impacting motivation
examined in our study. We begin with a discussion of genAI motivation
that in turn influences important outcomes (e.g., Collie & Martin,
and engagement, the study’s outcome factors.
2024b). As relevant to the present study, research among teachers
supports these premises, showing that teachers’ perceptions of
autonomy-supportive leadership are associated with enhanced motiva­ 1.2. Teachers’ motivation and engagement to use generative AI
tion and engagement at work, whereas perceptions of
autonomy-thwarting leadership are associated with poorer motivation Motivation and engagement have long been identified as important
and engagement outcomes (Collie, 2023a; Nie et al., 2015). for workplace functioning (Bakker & Demerouti, 2017; Ryan & Deci,
Conceptual work also underscores the role that occupational expe­ 2017), and the same is true for teachers (Collie, 2023b). As noted, we
riences play in relation to employees’ functioning at work—though, examined motivation and engagement in relation to genAI in this study.
there are different views on how this process occurs. Broaden-and-build For motivation, we assessed genAI valuing. Valuing is a construct that
theory (Fredrickson, 2001) establishes that adaptive experiences (e.g., appears within numerous motivation theories (e.g., Eccles & Wigfield,
professional growth striving) help individuals to be more aware of and 2023; Ryan & Deci, 2017) and refers to individuals’ subjective value
to apply the resources available to them, which then yield beneficial judgements of a particular task or resource (Wigfield et al., 2016).
outcomes. GenAI may be considered a resource that teachers can access Generally, valuing is considered to comprise four major components:
to do their work more effectively and efficiently. In contrast to that, attainment value, which reflects individuals’ sense of the importance of
conservation of resources theory (Hobfoll, 1989) states that individuals a task or resource for meeting their goals; intrinsic value, which con­
strive to reduce maladaptive experiences (e.g., change-related stress) by cerns individuals’ perceptions that the task/resource will be interesting
building and retaining resources that help them to undertake their work or enjoyable to undertake or use; utility value, which is individuals’

Fig. 1. Hypothesized model.

2
R.J. Collie and A.J. Martin Computers and Education: Artificial Intelligence 6 (2024) 100224

sense that the task/resource will be helpful for future aims; and, relative Although research examining motivation and engagement for genAI
cost, which refers to individuals’ belief that application of the task/re­ has yet to emerge, studies looking at technology more broadly are
source will take away time from other objectives (Wigfield et al., 2020). informative. This work has tended to examine transformational lead­
These four components have been examined under an overarching ership, which is similar to autonomy-supportive leadership via its focus
valuing construct (Muenks et al., 2023), which was the approach taken on recognizing and supporting the needs of each individual employee
in our study. However, given issues with the relative cost component in (Bass, 1995). Researchers have shown that teachers’ perceptions of
prior work—because it is a negatively-valenced component, whereas the transformational leadership are associated with greater technology
other valuing components are positively-valenced (Jiang et al., 2018)— valuing, but not integration of technology in student learning activities
we assessed relative gains instead of cost. We define relative gains as (Schmitz et al., 2023; see also Chiu, 2022; Vermeulen et al., 2015).
individuals’ beliefs that application of the task/resource will yield Karaca et al. (2013) examined a different leadership variable: principal
additional time or opportunities to conduct other activities. We felt this support for technological access, technical support, and professional
approach of relative gains is more appropriate to the potential yields of development for technology integration. They showed this variable was
genAI given the ease with which teachers can now access and use genAI correlated with greater technology-related valuing, as well as teachers’
tools. Taken together, in the present study, genAI valuing refers to an integration of technology in student learning activities. Emerging
overarching construct of teachers’ perceptions that genAI will be research has also examined transactional leadership, which is similar to
important for their work (attainment value), will make their work more autonomy-thwarting leadership in its focus on ensuring compliance.
interesting (intrinsic value), will be useful for conducting their work Prior work demonstrates that autonomy-thwarting leadership is asso­
(utility value), and, will help them to undertake tasks that they normally ciated with lower levels of digital transformation in schools (Ruloff &
could not do, for example, due to time constraints (relative gains). Petko, 2021) and lower innovation in teaching (Pierterse et al., 2010).
Turning to engagement, we examined a form of behavioral engage­ Taken together, research shows that leadership factors are relevant
ment relevant to genAI. Behavioral engagement features centrally in for teachers’ motivation and engagement at work generally, as well as in
definitions of engagement (e.g., Skinner & Pitzer, 2012) and is consid­ relation to technology. The aim of the present study was to extend this
ered a salient dimension of engagement because it reflects actions that understanding to ascertain whether these leadership factors are also
are put into practice (Martin et al., 2024). In the present study, we relevant to genAI motivation and engagement. Based on self-
assessed teachers’ integration of genAI in both their own determination theory (Ryan & Deci, 2017) and job demands-resources
teaching-related work and also student learning activities. GenAI inte­ theory (Bakker & Demerouti et al., 2017), it was hypothesized that
gration in teaching-related work reflects teachers’ use of this technology autonomy-supportive leadership would be associated with greater genAI
for assisting with their work tasks, such as planning, teaching, and valuing (Hypothesis 2a) and integration (Hypothesis 2b), whereas the
assessing. GenAI integration in student learning activities refers to the reverse would be true for autonomy-thwarting leadership (Hypothesis
extent to which teachers ask or require their students to use genAI tools 3a and 3b). Autonomy-supportive leaders promote teachers’ sense of
in their learning tasks. self-empowerment and agency, which should allow teachers the
Our motivation and engagement outcomes reflect teachers’ beliefs freedom and self-initiative to see the value of genAI and to apply it in
about genAI as relevant to teaching work, and their current use of genAI their work as they see fit. Conversely, genAI is a novel technology that
tools for teaching and learning. Given that motivation is generally un­ requires innovative behavior and so we assumed it would be less likely
derstood to precede engagement (e.g., Eccles & Wigfield, 2023; Ryan & to occur in an autonomy-thwarting leadership environment (e.g., see
Deci, 2017), we expected that genAI valuing would be associated with Pieterse et al., 2010). Based on the substantiation provided above and
greater levels of both genAI integration outcomes (Hypothesis 1). As prior research (Collie, 2023b), we also hypothesized that
detailed below, we were also interested in examining the role of key autonomy-supportive leadership (positive) and autonomy-thwarting
predictors in relation to these outcomes—with the aim of yielding leadership (negatively) would be indirectly associated with the
knowledge about what types of teachers and aspects of their context may engagement outcomes via motivation (Hypothesis 4).
(or may not) underpin their valuing or integration of genAI. We now
describe the contextual predictors, then occupational experiences, and 1.4. Occupational experiences
finally background factors.
Positive occupational experiences refer to teachers’ perceptions of
1.3. Contextual factors their work and are linked with adaptive outcomes among teachers
(Collie, 2023a). In our study, occupational experiences are defined in
Contextual factors have been shown to lay a foundation for teachers’ terms of “feeling good and functioning effectively” (Huppert & So,
motivation and engagement at work (Collie, 2023a; Nie et al., 2015; Wu 2013). We examined two factors relevant to this definition: one
et al., 2023). As noted earlier, we examined teachers’ perceptions of reflecting adaptive experiences (professional growth striving) and one
autonomy-supportive leadership, which reflects teachers’ sense that the reflecting maladaptive experiences (change-related stress). Professional
school principal (or equivalent) fosters their self-initiative and agency growth striving refers to teachers’ efforts to reflect on and plan for
by seeking their input, and undertakes efforts to understand their indi­ ongoing professional development (Collie, 2023a), and taps into the
vidual needs (Aelterman et al., 2019; Ryan & Deci, 2017). We also idea of functioning effectively at work. In contrast, change-related stress
examined teachers’ perceptions of autonomy-thwarting leadership, which refers to teachers’ negative psychological outcomes (i.e., sense of worry,
refers to teachers’ sense that the principal enacts controlling or pres­ pressure) resulting from recent changes at work (Collie, 2021; Putwain
suring practices by forcing compliance, or generating feelings of guilt or & von der Embse, 2019). Change-related stress taps into the idea of (not)
shame among teachers (Aelterman et al., 2019; Ryan & Deci, 2017). feeling good. These two factors were selected for examination because
Researchers have shown that teachers’ perceptions of they are relevant to opportunity and threat at work, and may be
autonomy-supportive leadership are linked with a range of general work particularly relevant to genAI. More precisely, genAI is a new techno­
motivation and engagement outcomes, including greater intrinsic logical resource that has implication for teachers’ ongoing professional
valuing (assessed as autonomous/intrinsic motivation; Collie, 2023b; development in that it poses many opportunities for development and
Nie et al., 2015; Wu et al., 2023) and greater behavioral engagement learning, as well as their ability to cope with recent stressful demands at
(Collie, 2023a; Klassen et al., 2012). The reverse is evident for teachers’ work in that it potentially offers much time-saving support.
perceptions of autonomy-thwarting leadership, which are linked with We expected that professional growth striving would be associated
negative work outcomes such as lower intrinsic valuing (Wu et al., 2023) with greater valuing (Hypothesis 5a) and integration of genAI (Hy­
and lower behavioral engagement (Collie, 2023a). pothesis 5b). This hypothesis is based on broaden-and-build theory

3
R.J. Collie and A.J. Martin Computers and Education: Artificial Intelligence 6 (2024) 100224

(Fredrickson, 2001), where adaptive experiences promote positive at­ 2. Study overview
titudes towards and uptake of resources in one’s environment. Put
differently, teachers who are striving to further develop as a professional We investigated the roles of different factors in relation to teachers’
are expected to gravitate towards genAI as an option for this. Research motivation (valuing) and engagement (integration) to use genAI for
on technology integration more generally provides some empirical teaching and learning, as well as the association between teachers’
support for this hypothesis: Li et al. (2019) demonstrated that teachers’ motivation and engagement. We focused on contextual (autonomy-
openness to technology, which can be considered akin to professional supportive leadership, autonomy-thwarting leadership), occupational
growth striving for technology, was associated with greater technology experience (professional growth striving, change-related stress), and
integration in teaching. background factors (gender, age, teaching experience, contract length,
For change-related stress, we advanced two opposing, alternative class size, school level) assessed at the beginning of a school term as
hypotheses and wanted to examine which was supported empirically. predictors of genAI motivation and, in turn, genAI engagement assessed
Based on broaden-and-build theory (Fredrickson, 2001), teachers who near the end of the same school term. Fig. 1 demonstrates the hypoth­
are highly stressed may be less likely to value (Hypothesis 6.1a) and esized model.
integrate genAI (Hypothesis 6.1b) because stress means they have less
time to consider and become aware of novel resources they can access. 3. Method
However, based on conservation of resources theory (Hobfoll, 1989),
teachers who are highly stressed may be more likely to value (Hypoth­ 3.1. Sample and procedure
esis 6.2a) and integrate genAI (Hypothesis 6.2b) because it serves as an
opportunity to call upon resources to reduce their work stress. Prior The sample for our study comprised 339 teachers who completed
research has only examined technology use broadly, showing that two questionnaires within one school term. The teachers were employed
teacher anxiety may predict lower integration (Henderson & Corry, in elementary (primary) schools (56%), secondary schools (39%), or at
2021); however, little is known about the role of change-related stress or both levels (6%) in Australia. Participants identified as female (87%) or
genAI—hence, the focus in our study. Based on the substantiation pro­ male (13%; none identified as non-binary or used other terms for
vided above, we also hypothesized that professional growth striving gender). Participants’ average age was 37 (SD = 10) years and they had
(positively) and change-related stress (positively or negatively; see on average 12 (SD = 9) years of teaching experience. Most of the sample
opposing, alternative Hypotheses 6.1b/6.2b) would be indirectly asso­ held a bachelor’s degree (67%) or a postgraduate degree (31%; the rest
ciated with the engagement outcomes via motivation (Hypothesis 7; held a certificate or diploma, 2%), and the vast majority (94%) spoke
Collie, 2023b). only English at home. Participants indicated their school’s location:
inner city (9%), suburban (67%), rural (23%), or remote (<1%). Par­
1.5. Background factors ticipants also reported their school’s socio-economic status: low (20%),
below average (18%), average (41%), above average (18%), or high
Teachers’ background factors are potentially relevant to their (3%). Teachers worked in government (73%), systemic Catholic (17%),
valuing and integration of genAI for teaching and learning. In the pre­ or non-government/independent schools (10%). The sample statistics
sent study, we examined six personal and job characteristics: gender, generally align with the population parameters. For example, elemen­
age, teaching experience, contract length, class size, and school level. tary schools are 82% female and secondary schools are 61% female in
Given the novelty of genAI, there does not yet appear to be research Australia (Australian Bureau of Statistics [ABS], 2022). Teachers in
investigating these factors directly. However, research on teacher Australia work in government (63%), systemic Catholic (19%), and in­
motivation and engagement broadly, or in relation to technology pro­ dependent schools (17%; ABS, 2022).
vides some preliminary indication of the potential role of these factors. Data are shared with another study (Collie & Martin, 2024a); how­
For example, some research has shown that male, younger, and less ever, none of the variable associations examined in the present study
experienced teachers are more likely to integrate technology in their have been investigated previously. Data were collected between April
work (Karaca et al., 2013; Li et al., 2019; Liu et al., 2017), whereas other and July 2023, which was Term 2 (out of four terms) in the Australian
studies have shown gender, age, and teaching experience are not school year. Time 1 was collected in April–May 2023, which corre­
significantly associated with technology integration (Inan & Lowther, sponded with week 2 in the school term for different states/territories.
2010; Liu et al., 2017). Although not related to genAI, research has Time 2 was collected in June–July 2023, which corresponded to week 9
demonstrated that fixed-term contracts are associated with heightened (towards the end) of the same term. Participants were recruited through
behavioral engagement (assessed as extra-role behaviors) among Qualtrics and their market research panel partners. People who had
teachers (Feather & Rauter, 2004). This may be because these teachers previously shown interest in participating in education-related surveys
are hoping to obtain a longer-term contract and so are putting in extra were sent the study invitation via email or app notification. These in­
effort to secure that. There is some evidence that smaller classes provide dividuals accessed the online survey, provided their consent, and
teachers with more time for instruction as they spend less time on dis­ responded to screening questions to ensure they were working as
ruptions and administrative tasks (Finn, 2019). However, other work teachers in an Australian school. Those who met the screening re­
has found no significant association between class size and technology quirements and completed the survey, but who did so hastily (<1/3
integration (Liu et al., 2017). Finally, some research has shown that median time), responded identically across most of the survey (>80%),
elementary school teachers are more likely to integrate technology over or answered alike to two oppositely phrased item pairs, were excluded
secondary school teachers (Hsu, 2010; Liu et al., 2017); however, the from the final sample (Dewitt et al., 2019). The response rate was 90% at
reverse may be evident for genAI given that elementary school is a time Time 1 (n = 492). At Time 2, the Time 1 sample was reinvited to
of concrete learning where teachers may view genAI as too abstract and participate. The final sample size represents the matched sample (n =
not so directly relevant to what elementary school students need to do 339), which was our focus given the genAI questions appeared in the
and learn in class (Collie & Martin, 2024a). Based on these mixed Time 2 survey.1 The study received Institutional Review Board ethics
findings, we felt it important to examine the background factors in approval.
relation to our substantive factors and left this as an open empirical
question: To what extent are the background factors direct and indirect
predictors of genAI valuing and integration? (Research Question 1). 1
For completeness, we conducted our analyses with the full, unmatched
sample as well, and results were very similar to those described below. Thus, we
retained our matched sample approach.

4
R.J. Collie and A.J. Martin Computers and Education: Artificial Intelligence 6 (2024) 100224

3.2. Measures Chat AI, and Bard AI.”

GenAI valuing was assessed with the following stem: “Thinking
Scales were scored from 1 (strongly disagree) to 7 (strongly agree) ahead, in the next 1–2 years I believe …” This was followed by four items
unless stated otherwise below. Prior studies have provided evidence of that tap into key aspects of valuing: attainment value (“It will be
validity for the scores of the established scales via measurement important to use generative AI in my work as a teacher”), intrinsic value
invariance tests, adequate reliability, and expected associations with (“Using generative AI will make my work more interesting”), utility
correlates (Collie, 2021, 2023b; Putwain & von der Embse, 2019), value (“Using generative AI will be useful for helping me do various
including for the adapted items (Collie, 2021). The present study pro­ teaching tasks”), and relative gains (“Using generative AI will enable me
vides the opportunity to yield validity evidence for the scores of the new to do things in my work that I normally could not do [e.g., save time that
scales (genAI motivation and engagement; see Data Analysis). Reli­ I can put towards other work-related tasks]”).
ability estimates are presented in Table 1 and were all at adequate levels GenAI integration in teaching-related work was assessed by asking
(i.e., ω > 0.70). For full lists of items, see citations below, or Supple­ teachers, “Are you using generative AI tools to aid you in your work as a
mentary Materials for the adapted scale. teacher?” Teachers responded from 1 (not at all) to 5 (everyday). Of the
sample, 49% indicated they were integrating genAI in teaching-related
3.2.1. Contextual factors activities to some degree. GenAI integration in student learning activities
The Leadership Approach to Autonomy Scale (Collie, 2021) was used asked teachers, “For what tasks are you asking your students to use
to assess perceived autonomy-supportive (5 items; e.g., “My principal generative AI?” Teachers responded to a multiple-choice question where
listens to my perspective”) and -thwarting leadership practices (5 items; the first option was “None. I am not asking my students to use generative
e.g., “My principal expresses disappointment if I don’t do things their AI in their learning.” The other options included different example tasks
way”). (e.g., learning activities in-class, homework, study/revision, other).
Responses to this item were coded 0 (for “none”) or 1 (for instances
3.2.2. Occupational experiences where teachers selected one or more learning task). Of the sample, 19%
Professional growth striving was assessed with four items from Collie’s indicated they were integrating genAI in student learning activities.
(2020) Tripartite Occupational Well-Being Scale (e.g., “I make plans for
my ongoing professional learning and development at work”). Chang­ 3.2.4. Background factors
e-related stress was assessed with items adapted from Putwain and von Gender was coded 0 (male) or 1 (female). Age and teaching experi­
der Embse’s (2019) work (e.g., “I have felt increased stress as a result of ence were continuous and coded in years. Contract length was coded
recent changes that have occurred in my work”). 0 (fixed term for less than 1 year) or 1 (permanent or fixed term for more
than 1 year). Class size was continuous. School level was coded
3.2.3. GenAI motivation and engagement 0 (elementary school) or 1 (secondary school). These factors have been
New items were developed to assess genAI motivation (valuing) and measured in prior work and have demonstrated evidence of validity (e.
genAI engagement (integration) in teaching and learning. Importantly, g., Collie, 2021; Klassen et al., 2012).
the present study provides evidence of validity for the scale scores via
several avenues: measurement invariance tests, expected correlations, 3.3. Data Analysis
sound factor structure, and adequate reliability (see Results for details).
In the survey, teachers were first given a brief definition about genAI Using Mplus 8.10 (Muthén & Muthén, 2023), we conducted confir­
before answering these items: “Generative artificial intelligence matory factor analysis (CFA) to test our measurement model (and pro­
(generative AI) refers to advanced technology that uses machine vide measurement support for our factors), followed by structural
learning to generate new content, such as text, images, music. Some equation modelling (SEM) and path analysis to ascertain structural paths
examples of these online tools include ChatGPT, GPT-4, Dall-E 2, Bing between factors as per our hypothesized model (see Fig. 1). Two sepa­
rate models were run; one model for each engagement outcome. Robust
maximum likelihood (MLR) was employed as the estimator for models
Table 1 involving the ordinal outcome (genAI integration in teaching-related
Reliabilities, descriptive statistics, and factor loadings. work) with full information maximum likelihood (FIML) to incorpo­
Omega M SD Factor rate missing data. Weighted least squares means and variances
(ω) Loadings (WLSMV) was used for models involving the binary outcome (genAI
M (Range) integration in student activities) with pairwise present for any missing
T1 Contextual factors data. Missing data were minimal (<2%). Model fit was assessed using
Perceived autonomy-supportive 0.89 4.49 1.24 0.78 the root-mean-square error of approximation (RMSEA), the comparative
leadership (0.61–0.86)
fit index (CFI), and the Tucker Lewis Index (TLI). RMSEA values ≤ 0.08
Perceived autonomy-thwarting 0.93 3.37 1.52 0.85
leadership (0.78–0.91) and CFI/TLI values ≥ 0.90 imply adequate fit (Hu & Bentler, 1999).
T1 Occupational experiences Means and standard deviations were calculated for each variable.
Professional growth striving 0.89 4.15 1.48 0.81 Following this, measurement models were run for each engagement
(0.79–0.83) outcome using CFA to obtain latent correlations among factors and to
Change-related stress 0.87 4.93 1.38 0.82
(0.68–0.91)
calculate coefficient omega as a measure of reliability. In the measure­
T2 Motivation ment model for genAI integration in teaching-related work, multiple
GenAI valuing 0.92 4.84 1.31 0.86 item scales were specified as latent factors, and the background factors
(0.79–0.96) and the engagement outcome were modelled with loading set to 1 and
T2 Engagement Outcomes
residual set to 0. The measurement model for genAI integration in stu­
GenAI integration in teaching- – 1.89 1.05 –
related work dent learning activities was the same except that the background factors
GenAI integration in student – 0.19 0.39 – and the engagement outcome were modelled as manifest variables
learning activities (given convergence errors when specifying variance/residual). To pro­
Note. Factors loadings and omegas are from the measurement model involving vide additional evidence of validity for the scale scores, we also con­
genAI integration in teaching-related work. Loadings were very similar for the ducted measurement invariance tests across a range of background
CFA involving genAI integration in student learning activities, as expected. T1 = factors (see Supplementary Materials for details).
Time 1. T2 = Time 2. To ascertain structural paths between factors, SEM was used for the

5
R.J. Collie and A.J. Martin Computers and Education: Artificial Intelligence 6 (2024) 100224

ordinal outcome and path analysis with probit regression was used for greater genAI valuing. Perceived autonomy-thwarting leadership was
the binary outcome. In both models, the Time 1 (T1) factors (contextual associated with greater genAI integration in student learning activities.
factors, occupational experiences, background factors) were entered as Professional growth striving was positively associated with genAI
predictors of Time 2 (T2) genAI valuing and, in turn, all factors were valuing and genAI integration in teaching-related work. Change-related
entered as predictors of the T2 outcome. The same specifications as stress was associated with greater genAI valuing (only in the probit
detailed for the measurement models were used, along with controls for regression). Male, younger, less experienced, and secondary teachers
shared variance among the T1 predictors. were more likely to value and integrate genAI. Teachers on short-term
contracts were less likely to value genAI, whereas class size was not
4. Results linked with the outcomes. Finally, teachers who reported higher genAI
valuing were more likely to integrate genAI in their own teaching-
Table 1 displays means, standard deviations for all substantive fac­ related work and their students’ learning.
tors. The measurement models demonstrated good fit for genAI inte­ Moving onto the structural modelling, which had controls for shared
gration in teaching-related work (χ2 [291] = 434.31, p < 0.001, RMSEA variance embedded, the fit indices for the two models were good: genAI
= 0.038, CFI = 0.97, TLI = 0.96) and genAI integration in student ac­ integration in teaching-related work (χ2 [291] = 434.31, p < 0.001,
tivities (χ2 [291] = 361.21, p = 0.003, RMSEA = 0.027, CFI = 0.95, TLI RMSEA = 0.038, CFI = 0.97, TLI = 0.96) and genAI integration in stu­
= 0.93). Omega coefficients are reported in Table 1 and indicate dent activities (χ2 [291] = 361.18, p = 0.003, RMSEA = 0.027, CFI =
adequate reliability. Table 1 also displays factor loading means and 0.95, TLI = 0.93). Results are displayed in Table 3 and Fig. 2. Findings
ranges for each latent factor, demonstrating sound measurement sup­ show that perceived autonomy-supportive leadership, professional
port. Supplementary Materials provide details about the measurement growth striving (only in the probit regression), and change-related stress
invariance tests, which show that invariance was attained across the were associated with greater T2 genAI valuing. For the background
background characteristics. factors, male teachers and secondary teachers were more likely to value
Table 2 displays substantive correlations (for all other correlations, genAI. In turn, T2 genAI valuing was associated with greater levels of
see Supplementary Materials). Because the models were run twice with both T2 genAI integration outcomes. There were also some direct as­
different estimators, there were some very slight differences. Here, we sociations between the T1 predictors and the engagement outcomes: T1
report correlations that were significant in at least one model. Correla­ perceived autonomy-thwarting leadership was associated with greater
tions were generally as expected, providing additional evidence of val­ T2 genAI integration in student learning activities, and T1 professional
idity. Perceived autonomy-supportive leadership was associated with

Table 3
Table 2 Standardized beta estimates from structural models.
Latent correlations for substantive associations. T2 T2 Engagement
Motivation
T2 T2 Engagement
Motivation GenAI GenAI integration GenAI integration
valuing in teaching-related in student learning
GenAI GenAI integration GenAI integration
work activities
valuing in teaching-related in student learning
work activities T1 Contextual factors
Perceived 0.21a/0.23a 0.04 0.19
T1 Contextual factors
autonomy-
Perceived 0.16b/0.17b 0.09 − 0.01
supportive
autonomy-
leadership
supportive
Perceived 0.06/0.06 0.10 0.32b
leadership
a autonomy-
Perceived − 0.03/ 0.04 0.16
thwarting
autonomy- − 0.03
leadership
thwarting
T1 Occupational experiences
leadership
Professional 0.11/0.13a 0.12a 0.04
T1 Occupational experiences
growth striving
Professional 0.19b/0.22c 0.21c
0.12
Change-related 0.15b/0.17b − 0.03 − 0.02
growth striving
stress
Change-related 0.11/0.12a 0.03 0.05
T1 Personal and job characteristics
stress
Gender (M/F) − 0.11b/ − 0.05 − 0.08
T1 Personal and job characteristics
− 0.12a
Gender (M/F) − 0.14b/ − 0.14a − 0.18b
Age − 0.16/ − 0.03 0.25
− 0.14a
− 0.14
Age − 0.24c/ − 0.15 b
0.06
Teaching − 0.07/ − 0.07 − 0.15
− 0.22c
experience − 0.07
Teaching − 0.21c/ − 0.16 b
− 0.03
Contract (short- 0.07/0.06 − 0.01 − 0.02
experience − 0.20c
term/longer-term)
Contract (short- 0.13a/0.12a 0.08 0.08
Class size 0.04/.04 0.04 0.06
term/longer-term)
School level 0.15b/0.13a 0.11a 0.28c
Class size 0.04/0.04 0.05 0.04
(elem./sec.)
School level 0.13a/0.12a 0.15b 0.28c
T2 Motivation
(elem./sec.)
GenAI valuing 0.45c 0.25b
T2 Motivation
c c 2
GenAI valuing 0.50 0.30 R 17%/17% 29% 0.24%

Note. Values before and after the slash represent those from the models for genAI Note. Values before and after the slash represent those from the genAI integra­
integration in teaching-related work and student learning activities, respec­ tion in teaching-related work and student learning activities, respectively.
tively. T1 = Time 1. T2 = Time 2. M = male. F = female. Elem. = elementary T1 = Time 1. T2 = Time 2. M = male. F = female. Elem. = elementary school.
school. Sec. = secondary school. Sec. = secondary school.
a a
p < .05. p < .05.
b b
p < .01. p < .01.
c c
p < .001. p < .001.

6
R.J. Collie and A.J. Martin Computers and Education: Artificial Intelligence 6 (2024) 100224

Fig. 2. Results from structural models examining predictors of GenAI valuing and integration.
Note. Standardized betas shown. Significant paths presented (along with some non-significant paths if part of a pair). See Table 3 for all paths. Values before and after
the slash represent those from the genAI integration in teaching-related work and student learning activities, respectively. ns = non-significant.

growth striving was associated with greater T2 genAI integration in leadership helps employees to value (or internally endorse) the norms
teaching-related work. Of the background factors, secondary teachers and goals of their working context (Ryan & Deci, 2017). Such leadership
were more likely to integrate genAI in both ways. Alongside those re­ may help teachers to internalize the importance of effective teaching,
sults, there were also three indirect associations involving genAI inte­ which genAI has the potential to enhance via its opportunities for pro­
gration in teaching-related work: Perceived autonomy-support → genAI fessional learning, adaptive teaching strategies, and assessment support
valuing → genAI integration in teaching-related work (β = 0.09, SE = (Chiu, 2023). Perhaps this result occurred, then, because
0.05, p = 0.04, 95% CI [0.01, 0.18]); change-related stress → genAI autonomy-supportive leadership means teachers are empowered to be
valuing → genAI integration in teaching-related work (β = 0.07, SE = self-determined in their work, which allows them to value genAI as a
0.03, p = 0.02, 95% CI [0.01, 0.12]); and, secondary teachers → genAI resource that can help them to be effective at work. It may also be that
valuing → genAI integration in teaching-related work (β = 0.07, SE = the empowerment provided by autonomy-supportive leadership means
0.03, p = 0.02, 95% CI [0.01, 0.12]). that teachers experience the freedom to experiment and apply genAI in
their work—which is likely vital given the novelty of this technology.
5. Discussion Professional growth striving was also linked with greater genAI
valuing, supporting Hypothesis 5a. This likely occurred because teach­
We examined the role of contextual factors, occupational experi­ ers who are striving to further develop as a professional gravitate to­
ences, and background factors in predicting teachers’ genAI valuing and wards genAI as an option for this (Li et al., 2019). The other substantive
integration. Results indicate that perceived autonomy-supportive lead­ factor linked with genAI valuing was change-related stress. Teachers
ership, professional growth striving, and change-related stress were reporting greater change-related stress at the start of the school term
linked with greater genAI valuing. In turn, genAI valuing was associated were more likely to report greater genAI valuing at the end of the term.
with greater genAI integration. Perceived autonomy-thwarting leader­ For this association, there were two opposing, alternative hypotheses.
ship was linked with greater genAI integration in student learning ac­ This result aligns with the hypothesis based on conservation of resources
tivities and professional growth striving was associated with greater theory (Hypothesis 6.2a; Hobfoll, 1989), rather than the hypothesis
genAI integration in teaching-related work. Teachers’ gender and school based on broaden-and-build theory (Hypothesis 6.1a; Fredrickson,
level were also linked with valuing and/or the integration outcomes, 2001). Conservation of resources theory specifies that stress leads in­
and there were several indirect associations as well. Notable findings are dividuals to seek out additional resources to help them overcome the
discussed below. challenges they face. Our results suggest that stressed teachers are more
likely to value genAI, potentially because it offers a lifeline to manage
5.1. Predictors of GenAI valuing their workloads—which are well-documented to be at high levels across
many countries (Australian Government, 2022; UK Department for Ed­
Perceived autonomy-supportive leadership has been linked with a ucation, 2018). Notably, it is possible this finding occurred given the
wide range of positive work outcomes for teachers (Collie, 2023a, ease with which people are able to access and start using genAI tools.
2023b; Klassen et al., 2012; Nie et al., 2015). The present research ex­ Perhaps with other technology that is more time-intensive to learn there
tends that knowledge to show that this type of leadership is also relevant would be evidence for the broaden-and-build hypothesis (e.g., Hender­
for genAI valuing. In line with Hypothesis 2a, teachers who felt their son & Corry, 2021).
principal was autonomy-supportive tended to report greater genAI Two of the background factors were linked with genAI valuing
valuing. According to self-determination theory, autonomy-supportive (Research Question 1). Male teachers and secondary school teachers

7
R.J. Collie and A.J. Martin Computers and Education: Artificial Intelligence 6 (2024) 100224

were more likely to value genAI. Given mixed findings in prior research, (Fredrickson, 2001), which states that adaptive experiences boost up­
we did not posit hypotheses about these associations. Nonetheless, these take of resources in one’s environment. At the same time, professional
findings are understandable. Male teachers have been found to imple­ growth striving was not linked with genAI integration in student
ment technology at higher degrees in some prior research (Karaca et al., learning activities (partially contradicting Hypothesis 5b). It is possible
2013; Li et al., 2019; Liu et al., 2017). Research has also shown that men that the novelty of the technology explains this finding—teachers may
tend to have more positive attitudes towards AI (Kaya et al., 2024), and want to gain expertise in genAI first before implementing it with stu­
it appears a similar phenomenon is occurring with genAI among dents. Research conducted over the coming years will be able to test this.
teachers. Efforts to support female teachers are thus needed to ensure Secondary school teachers were more likely to integrate genAI
that genAI is accessible to and accessed by these individuals to the extent (Research Question 1). Secondary school provides an opportunity for
this is important for professional opportunity and effectiveness. Sec­ advanced, abstract learning that teachers may feel is better suited to
ondary school teachers were more likely to value genAI, which differs genAI than elementary school (Collie & Martin, 2024a). There were no
from prior research on technology valuing more broadly (Hsu, 2010; Liu other significant direct findings. Thus, autonomy-supportive leadership
et al., 2017). However, as we noted earlier, there may be some hesitation and change-related stress were not associated with genAI integration
from elementary school teachers to use genAI given the focus on more (contradicting Hypothesis 2b and 6.1b/6.2b). As noted earlier, these
concrete and less technologically-oriented learning in the earlier years factors appeared more relevant to genAI valuing.
of schooling (Collie & Martin, 2024a). Alongside these findings, there were also several indirect associa­
Autonomy-thwarting leadership was not associated with genAI tions, partially supporting Hypotheses 4 and 7, and relevant to Research
valuing in our sample, contradicting Hypothesis 3a. However, as dis­ Question 1. Specifically, perceived autonomy-supportive leadership,
cussed below, this factor was directly linked with genAI integration. change-related stress, and working as a secondary teacher were all
Together, our findings highlight factors relevant to genAI valuing. These associated with greater genAI integration in teaching-related work via
factors are important to know given there will be increasing efforts to the boost these factors provided to genAI valuing. Combined, these
apply genAI in teaching and learning going forward (e.g., Australian findings yield direction for future research and practice around genAI
Government, 2023; US Office of Educational Technology, 2023; UK integration, including the types of contextual factors, occupational ex­
Department for Education, 2023). Our findings are thus pertinent to the periences, and background factors relevant to this. Indeed, our results
ongoing development of policies and practice guidelines related to suggest that efforts to boost genAI integration in schools likely need to
genAI. Indeed, if schools and educational jurisdictions want genAI to be consider support not only for individual teachers, but also for school
applied, it is essential that teachers see the value in this technology. Our leaders. Moreover, these findings are relevant in light of other educa­
results also hold implications for the implementation of genAI: they tional disruptions, like COVID-19, which have already and will continue
provide initial direction about the ways in which teachers can be better to impact and boost digital uptake in learning and teaching (Mahajan
supported to value genAI, including the types of teachers who may need et al., 2023). Future research that examines how genAI interacts with
additional support to boost their valuing (e.g., female teachers, other disruptions (e.g., pandemics, climate change) is important,
elementary teachers, teachers experiencing low autonomy-support from including whether these disruptions hasten or shift uptake and types of
leadership). integration. As a case in point, it may be that in times of disruption that
impact teacher availability or involve remote/hybrid learning, genAI
5.2. Predictors of GenAI integration can be used to help address any gaps or needs in learning activities,
content, or assessment.
GenAI valuing was a strong, positive predictor of both forms of genAI
integration. Teachers who value genAI are more likely to implement this 5.3. Limitations and future directions
technology in both their teaching and learning. This finding supports
Hypothesis 1, and is aligned with motivational theories that indicate the Limitations should be considered when interpreting our results. First,
role of valuing (and motivation more broadly) for engagement (e.g., the contextual factors reflect teachers’ perceptions. Perceived leadership
Eccles & Wigfield, 2023; Ryan & Deci, 2017). practices are distinct from leader-reported or observed practices, but are
Perceived autonomy-thwarting leadership was directly linked with important nonetheless given it is teachers’ interpretations that have a
greater genAI integration in student learning activities (contradicting strong influence on their experiences. Notwithstanding this, attempts to
Hypothesis 3b where we assumed there would be a negative associa­ gather other types of reports will be useful in future to advance
tion). This is a novel finding given relatively limited research on knowledge. Second, our hypothesized model was grounded in theory
autonomy-thwarting leadership among teachers. Autonomy-thwarting and the outcomes were measured at a later timepoint; however, our
leadership involves pressuring teachers to act, think, and feel in findings are correlational. Experimental research is needed to test cau­
certain ways (Ryan & Deci, 2017). It may be that teachers who feel their sality. Third, our study was conducted over the course of one school
principal is autonomy-thwarting look for other ways to get around that term. Although this is an important timeframe in teaching, given that it
control or pressure, such as through integrating novel technology with dictates the detailed scope and sequence of learning, research that in­
their students. GenAI integration may be an outlet for these teachers to vestigates a longer gap between waves is important such as over one
take back some degree of autonomy in their work. This explanation school year. Fourth, we were not able to account for nesting of teachers
aligns with other research showing that authoritarian leadership can within schools given the recruitment methods. Although this was un­
lead employees to pursue behavior that is outside typical expectations likely to be a concern given our sample was collected from across the
(Bodla et al., 2019). It is interesting that perceived autonomy-thwarting country, it is a relevant consideration for future research.
leadership was not consistently associated with genAI integration in Fifth, our integration outcomes involved single-item scales. Given
teaching related-work. A controlling climate means more rigid proced­ these were ascertaining actual practice, rather than a latent construct,
ures and policies, perhaps leaving teachers less opportunity for using this approach is not inappropriate. However, multi-item scales are
genAI in their own work. Future research using person-centered ap­ important in future to allow for controls for measurement error. Sixth,
proaches will be helpful for further untangling these differences. our scales for genAI valuing and integration were developed for this
Professional growth striving was directly linked with greater genAI study. As noted in Results, we were able to provide robust evidence of
integration in teaching-related work (partially supporting Hypothesis validity for the scale scores (e.g., measurement invariance, anticipated
5b). This finding likely occurred because teachers who are making plans associations with correlates, adequate reliability, sound factor struc­
to develop professionally consider genAI as one option to do this. More ture). Nonetheless, additional research is needed to provide further ev­
specifically, this finding aligns with broaden-and-build theory idence of validity. This is important for advancing knowledge on this

8
R.J. Collie and A.J. Martin Computers and Education: Artificial Intelligence 6 (2024) 100224

topic, and also for ongoing development of strong measurement tools. Declaration of competing interest
Relatedly, we examined relative gains, rather than the traditional
approach of considering relative costs. The rationale for this was based The authors declare that they have no known competing financial
on measurement issues with a negatively-valenced cost factor (Jiang interests or personal relationships that could have appeared to influence
et al., 2018) and because genAI is expected to yield comparatively more the work reported in this paper.
gains than costs in terms of teachers’ time. In future, it will be important
to consider relative costs alongside relative gains. Acknowledgements
Seventh, although we suggest that genAI acts as a resource for
teachers, we did not position genAI as a contextual factor in our study. RJC would like to thank the Australian Research Council for funding
Instead, we were more interested in teachers’ valuing of and integration this research (DE200100436).
of this novel resource. In future, once genAI is applied more widely, it
will be important to examine teachers’ perceptions of genAI directly as a Appendix A. Supplementary data
contextual resource. Eighth, we considered leadership support in this
study, but going forward it is also important to consider collegial sup­ Supplementary data to this article can be found online at [Link]
port, given that collaboration has been identified as a major theme in org/10.1016/[Link].2024.100224.
dealing with challenges posed by educational disruptions (e.g., COVID-
10; Mahajan et al., 2023). It may be that collaboration helps teachers to References
identify novel and helpful ways to integrate genAI in their teaching. Or,
indeed, perhaps increasing reliance on genAI to develop learning ac­ Aelterman, N., Vansteenkiste, M., Haerens, L., Soenens, B., Fontaine, J. R. J., & Reeve, J.
(2019). Toward an integrative and fine-grained insight in motivating and
tivities, content, and assessment tasks may lead to less collaboration that demotivating teaching styles: The merits of a circumplex approach. Journal of
would have otherwise occurred in developing these resources. Ninth, we Educational Psychology, 111(3), 497–521. [Link]
focused on genAI valuing and integration, but did not consider teachers’ Ahmad Bodla, A., Tang, N., Van Dick, R., & Mir, U. R. (2019). Authoritarian leadership,
organizational citizenship behavior, and organizational deviance: Curvilinear
perceptions of accuracy in genAI content. This is a critical issue going
relationships. The Leadership & Organization Development Journal, 40(5), 583–599.
forward given misinformation is common in content created by genAI, [Link]
and can also be deliberately created by actors endeavoring to sow un­ Australian Bureau of Statistics [ABS]. (2022). Schools: Data on students, staff, schools, rates
and ratios for government and non-government schools, for all Australian states and
truths (Lim, 2023). Research that examines how teachers can navigate
territories. Author. [Link]
this issue in using genAI in their work and students’ work is vital (Lim, 2022.
2023)—such as by using critical literacy to check content and ensure Australian Government. (2022). National teacher workforce action plan. Department of
truthful sources are cited, and explicitly teaching these skills to students Education. [Link]
Australian Government. (2023). Australian framework for generative artificial intelligence in
(Casal-Otero et al., 2023). Finally, although generally aligned with the schools. Author. [Link]
gender breakdown of the Australian teaching population (ABS, 2022), mework-generative-artificial-intelligence-ai-schools.
the sample was largely female. Investigations that gather data from a Bakker, A. B., & Demerouti, E. (2017). Job demands–resources theory: Taking stock and
looking forward. Journal of Occupational Health Psychology, 22(3), 273–285. https://
greater percentage of male teachers are necessary to further advance [Link]/10.1037/ocp0000056
understanding. Bass, B. M. (1995). Theory of transformational leadership redux. The Leadership
Quarterly, 6(4), 463–478. [Link]
Carvalho, L., Martinez-Maldonado, R., Tsai, Y.-S., Markauskaite, L., & De Laat, M.
6. Conclusion (2022). How can we design for learning in an AI world? Computers & Education:
Artificial Intelligence, 3, Article 100053. [Link]
Educational jurisdictions are currently developing policy and prac­ caeai.2022.100053
Casal-Otero, L., Catala, A., Fernández-Morante, C., Taboada, M., Cebreiro, B., & Barro, S.
tice guidelines around the use of genAI in teaching and learning (e.g.,
(2023). AI literacy in K-12: A systematic literature review. International Journal of
Australian Government, 2023; US Office of Educational Technology, STEM Education, 10(1), 29. [Link]
2023; UK Department for Education, 2023), and indications are that Chiu, T. K. F. (2022). School learning support for teacher technology integration from a
self-determination theory perspective. Educational Technology Research &
schools will be expected to employ this technology (e.g., Australian
Development, 70(3), 931–949. [Link]
Government, 2023). Our findings are relevant to this changing space Chiu, T. K. F. (2023). The impact of generative AI (GenAI) on practices, policies, and
given they highlight some key factors relevant to teachers’ motivation research direction in education: A case of ChatGPT and Midjourney. Interactive
(valuing) and engagement (integration) relating to genAI. In particular, learning environments. Advance online publication. [Link]
10494820.2023.2253861
contextual factors, occupational experiences, and background factors all Collie, R. J. (2020). The tripartite occupational well-being scale. University of New South
appear to have pertinence in this regard. Our results, thus, pinpoint Wales. [Link]
areas to focus on in future research, policy, and practice regarding Collie, R. J. (2021). COVID-19 and teachers’ somatic burden, stress, and emotional
exhaustion: Examining the role of principal leadership and workplace buoyancy.
supporting genAI and its application in schools. AERA Open, 7. [Link]
Collie, R. J. (2023a). Teacher well-being and turnover intentions: Investigating the roles
Statement on open data and ethics of job resources and job demands. British Journal of Educational Psychology, 93(3),
712–726. [Link]
Collie, R. J. (2023b). Teachers’ work motivation: Examining perceived leadership
The Institutional Research Board ethics approval received for this practices and salient outcomes. Teaching and Teacher Education, 135, Article 104348.
study does not allow for data sharing. Ethical approval obtained from [Link]
Collie, R. J., & Martin, A. J. (2024a). Teachers’ uptake of generative AI in teaching and
the UNSW Human Research Ethics Committee. Participants provided
learning: Important questions and early answers. Manuscript submitted for publication.
implied consent for their participation by completing the online survey. Collie, R. J., & Martin, A. J. (2024b). The Academic and Social-Emotional Flourishing
Framework. Manuscript submitted for publication.
Demerouti, E., & Bakker, A. B. (2023). Job demands-resources theory in times of crises:
CRediT authorship contribution statement
New propositions. Organizational Psychology Review, 13(3), 209–236. [Link]
org/10.1177/20413866221135022
Rebecca J. Collie: Writing – review & editing, Writing – original Department for Education. (2018). Making data work: Report of the teacher workload
draft, Project administration, Methodology, Funding acquisition, Formal advisory group. Retrieved from [Link]
orkload-advisory-group.
analysis, Data curation, Conceptualization. Andrew J. Martin: Writing Department for Education. (2023). Policy paper: Generative artificial intelligence in
– review & editing, Methodology, Conceptualization. education. [Link]
telligence-in-education.
Dewitt, B., Fischhoff, B., Davis, A. L., Broomell, S. B., Roberts, M. S., & Hanmer, J.
(2019). Exclusion criteria as measurements II: Effects on utility functions. Medical
Decision Making, 39(6), 704–716. [Link]

9
R.J. Collie and A.J. Martin Computers and Education: Artificial Intelligence 6 (2024) 100224

Eccles, J. S., & Wigfield, A. (2023). Expectancy-value theory to situated expectancy-value Technology Research & Development, 65(4), 795–813. [Link]
theory: Reflections on the legacy of 40+ years of working together. Motivation s11423-016-9487-9
Science, 9(1), 1–12. [Link] Mahajan, R., Lim, W. M., Kumar, S., & Sareen, M. (2023). COVID-19 and management
Feather, N. T., & Rauter, K. A. (2004). Organizational citizenship behaviours in relation education: From pandemic to endemic. International Journal of Management in
to job status, job insecurity, organizational commitment and identification, job Education, 21(2), Article 100801. [Link]
satisfaction and work values. Journal of Occupational and Organizational Psychology, Martin, A. J., Collie, R. J., Stephan, M., Flesken, A., Halcrow, F., & McCourt, B. (2024).
77(1), 81–94. [Link] What is the role of teaching support in students’ motivation and engagement
Finn, J. D. (2019). Academic and non-cognitive effects of small classes. International trajectories during adolescence? A four-year latent growth modeling study. Learning
Journal of Educational Research, 96, 125–135. [Link] and Instruction.
ijer.2019.05.006 Muenks, K., Miller, J. E., Schuetze, B. A., & Whittaker, T. A. (2023). Is cost separate from
Fredrickson, B. L. (2001). The role of positive emotions in positive psychology: The or part of subjective task value? An empirical examination of expectancy-value
broaden-and-build theory of positive emotions. American Psychologist, 56(3), versus expectancy-value-cost perspectives. Contemporary Educational Psychology, 72,
218–226. [Link] Article 102149. [Link]
Henderson, J., & Corry, M. (2021). Teacher anxiety and technology change: A review of Muthén, L. K., & Muthén, B. (2023). Mplus user’s guide. Author.
the literature. Technology, Pedagogy and Education, 30(4), 573–587. [Link] Nie, Y., Chua, B. L., Yeung, A. S., Ryan, R. M., & Chan, W. Y. (2015). The importance of
10.1080/1475939X.2021.1931426 autonomy support and the mediating role of work motivation for well-being: Testing
Hobfoll, S. E. (1989). Conservation of resources: A new attempt at conceptualizing stress. self-determination theory in a Chinese work organisation. International Journal of
American Psychologist, 44(3), 513–524. [Link] Psychology, 50(4), 245–255. [Link]
066X.44.3.513 Office of Educational Technology. (2023). Artificial intelligence and the future of
Hsu, S. (2010). Developing a scale for teacher integration of information and teaching and learning: Insights and recommendations. [Link]
communication technology in grades 1–9. Journal of Computer Assisted Learning, 26 Pieterse, A. N., van Knippenberg, D., Schippers, M., & Stam, D. (2010). Transformational
(3), 175–189. [Link] and transactional leadership and innovative behavior: The moderating role of
Hu, L.-T., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure psychological empowerment. Journal of Organizational Behavior, 31(4), 609–623.
analysis: Conventional criteria versus new alternatives. Structural Equation Modeling: [Link]
A Multidisciplinary Journal, 6(1), 1–55. Putwain, D. W., & von der Embse, N. P. (2019). Teacher self-efficacy moderates the
Huppert, F., & So, T. C. (2013). Flourishing across Europe: Application of a new relations between imposed pressure from imposed curriculum changes and teacher
conceptual framework for defining well-being. Social Indicators Research, 110(3), stress. Educational Psychology, 39(1), 51–64. [Link]
837–861. [Link] 01443410.2018.1500681
Inan, F. A., & Lowther, D. L. (2010). Factors affecting technology integration in K-12 Ruloff, M., & Petko, D. (2021). School principals’ educational goals and leadership styles
classrooms: A path model. Educational Technology Research & Development, 58(2), for digital transformation: Results from case studies in upper secondary schools.
137–154. [Link] International Journal of leadership in education. Advance online publication. https://
Jiang, Y., Rosenzweig, E. Q., & Gaspard, H. (2018). An expectancy-value-cost approach [Link]/10.1080/13603124.2021.2014979
in predicting adolescent students’ academic motivation and achievement. Ryan, R. M., & Deci, E. L. (2017). Self-determination theory: Basic psychological needs in
Contemporary Educational Psychology, 54, 139–152. [Link] motivation, development, and wellness. Guilford Press.
cedpsych.2018.06.005 Ryan, R. M., Duineveld, J. J., Di Domenico, S. I., Ryan, W. S., Steward, B. A., &
Karaca, F., Can, G., & Yildirim, S. (2013). A path model for technology integration into Bradshaw, E. L. (2022). We know this much is (meta-analytically) true: A meta-
elementary school settings in Turkey. Computers & Education, 68, 353–365. https:// review of meta-analytic findings evaluating self-determination theory. Psychological
[Link]/10.1016/[Link].2013.05.017 Bulletin, 148(11–12), 813–842. [Link]
Kaya, F., Aydin, F., Schepman, A., Rodway, P., Yetişensoy, O., & Demir Kaya, M. (2024). Schmitz, M.-L., Antonietti, C., Consoli, T., Cattaneo, A., Gonon, P., & Petko, D. (2023).
The roles of personality traits, AI anxiety, and demographic factors in attitudes Transformational leadership for technology integration in schools: Empowering
toward Artificial Intelligence. International Journal of Human–Computer Interaction, teachers to use technology in a more demanding way. Computers & Education, 204,
40(2), 497–514. [Link] Article 104880. [Link]
Klassen, R. M., Perry, N. E., & Frenzel, A. C. (2012). Teachers’ relatedness with students: Skinner, E. A., & Pitzer, J. R. (2012). Developmental dynamics of student engagement,
An underemphasized component of teachers’ basic psychological needs. Journal of coping, and everyday resilience. In S. L. Christenson, A. L. Reschly, & C. Wylie (Eds.),
Educational Psychology, 104(1), 150–165. [Link] Handbook of research on student engagement (pp. 21–44). Springer.
Li, Y., Garza, V., Keicher, A., & Popov, V. (2019). Predicting high school teacher use of Vermeulen, M., Van Acker, F., Kreijns, K., & van Buuren, H. (2015). Does
technology: Pedagogical beliefs, technological beliefs and attitudes, and teacher transformational leadership encourage teachers’ use of digital learning materials.
training. Technology, Knowledge and Learning, 24(3), 501–518. [Link] Educational Management Administration & Leadership, 43(6), 1006–1025. [Link]
10.1007/s10758-018-9355-2 org/10.1177/1741143214535749
Lim, W. M. (2023). Fact or fake? The search for truth in an infodemic of disinformation, Wigfield, A., Eccles, J. S., & Möller, J. (2020). How dimensional comparisons help to
misinformation, and malinformation with deepfake and fake news. Journal of understand linkages between expectancies, values, performance, and choice.
Strategic Marketing. [Link] Educational Psychology Review, 32(3), 657–680. [Link]
Lim, W. M., Gunasekara, A., Pallant, J. L., Pallant, J. I., & Pechenkina, E. (2023). 020-09524-2
Generative AI and the future of education: Ragnarök or reformation? A paradoxical Wigfield, A., Tonks, S. M., & Klauda, S. L. (2016). Expectancy-value theory. In
perspective from management educators. International Journal of Management in K. R. Wentzel, & D. B. Miele (Eds.), Handbook of motivation at school (pp. 55–74).
Education, 21(2), Article 100790. [Link] Routledge.
Liu, F., Ritzhaupt, A. D., Dawson, K., & Barron, A. E. (2017). Explaining technology Wu, G., Zhang, L., Liu, X., & Liang, Y. (2023). How school principals’ motivating style
integration in K-12 classrooms: A multilevel path analysis model. Educational stimulates teachers’ job crafting: A self-determination theory approach. Current
Psychology, 42(24), 20833–20848. [Link]

10