Predicting workplace creativity of university administrative staff: the mediating role of learning goal orientation

There is increasing pressure for universities worldwide to adapt and meet rising societal expectations. The proportion of administrative staff has increased significantly to help universities accommodate their expanded and diversified missions (Baltaru and Soysal, 2018). They are in charge of helping universities to navigate complex government policy environments and meet the increasing demand for accountability and reporting requirements (Croucher and Woelert, 2022). They also play a major role in delivering a wide range of student services (Graham, 2012), while supporting the work of their academic colleagues (Gibbs and Kharouf, 2022).

Yet, despite their increasing importance, administrative staff’s contribution to university development remains largely unnoticed and undervalued (Gander et al., 2019). Research on occupational groups has largely focused on academic staff, while administrative staff report less institutional support for their development than their academic counterparts (Coomber, 2019). Only recently has the literature on university administrative staff expanded, examining their career aspirations, competencies and role in knowledge development (de Jong and del Junco, 2024; Veles et al., 2023). We seek to contribute to this growing literature by examining how they can be supported to respond effectively to the evolving demands of their roles. More specifically, this study examines the drivers of administrative staff’s workplace creativity, which refers to one’s ability to generate “novel and useful work-related ideas” (Amabile, 1996, p. 1155).

Operating at the interface between the university system and its community, administrative staff must ensure compliance with rules and regulations while also being flexible enough to meet the diverse needs of students and academics (Grinkevich et al., 2020). Such discretion requires creative thinking, which would allow administrative staff to recombine existing knowledge and deviate from existing rules in an appropriate manner to fit novel circumstances (Houtgraaf et al., 2021). Creativity can also foster a more positive attitude towards change (Liu et al., 2016) and greater openness towards new perspectives (Abu Raya et al., 2023). Both are important for understanding different stakeholders’ needs and for navigating through complex and ambiguous situations (Grinkevich et al., 2020). By reconciling rigid procedures with flexible solutions for both internal and external university stakeholders, workplace creativity can help reduce bureaucratic hindrance, improve access to services and enhance their efficiency and quality (Grinkevich et al., 2020). Given these potential benefits, it is important for university leaders to identify and cultivate drivers of administrative staff’s workplace creativity.

To achieve this goal, we draw on Amabile (1996)’s componential model of creativity as our underlying theoretical framework to examine both social and individual psychological factors influencing workplace creativity of university administrative staff. Our study also extends this model by incorporating learning goal orientation (LGO) as it has been found to be a strong predictor of employees’ creative performance (Nandi and Watts, 2025) and other elements of the componential model including intrinsic motivation, creative self-efficacy (CSE) and perceived organisation support for creativity (POSfC) (Atitumpong and Badir, 2018; Cerasoli and Ford, 2014; Hendrikx et al., 2022). We test this integrated framework using empirical evidence collected from three Vietnamese public universities. Vietnam provides a relevant study context since its public universities face growing administrative demands and have a great need for innovation to improve its educational and institutional quality, which heavily rely on their administrative staff’s capacity (Nguyen et al., 2025). Thus, this study makes its practical contribution by providing university leaders, including those from Vietnam, with insights on drivers of their administrative staff’s workplace creativity.

As a central tenet of the componential theory, intrinsic motivation refers to one’s engagement in a task out of interest in and enjoyment of the task itself (Santoro, 2022; Tremblay et al., 2009). It is theorised to play a prominent role throughout the creative process (Thuan and Thanh, 2019) and a key element of creativity (Woodman et al., 1993). When employees are intrinsically motivated, they pay more attention to the problem or an intriguing opportunity in their work, which supports the generation of novel and useful solutions (Amabile and Pratt, 2016; Nili and Tasavori, 2022). It also allows employees to sustain their efforts and engagement during the cognitively demanding implementation and evaluation stages which involve trial-and-error iterations of new ideas (Mumford et al., 2012; Thuan and Thanh, 2019). Thus, we predict that:

While intrinsic motivation represents the “will do” component, CSE focuses on the “can do” aspect of creativity, which refers to one’s belief in his/her ability to produce creative outcomes (Liu et al., 2016; Tierney and Farmer, 2002). This belief provides employees with the confidence to invest their time and effort in challenging established routines and standards and in seeking new ways to improve their work (Yang and Zhou, 2022). Such creative confidence can also foster employees’ resilience by supplying additional personal resources that buffer against the challenges and setbacks that arise during the creative process (Liu et al., 2016; Mumford et al., 2012). Hence, we predict that:

Being creative at work can be a risky decision for employees, as it involves making changes to an existing system that may result in undesirable outcomes (Nguyen et al., 2023). Perceived organisational support for creativity (POSfC) is theorised to be an important factor influencing an employee’s creative capacity (Amabile, 1996), referring to an environment where creativity is encouraged, rewarded and valued. In such a climate, employees would feel psychologically safe to voice unconventional ideas, ask questions and share knowledge without fear of negative consequences, thus promoting workplace creativity (Edmondson and Bransby, 2023; Pan et al., 2020). Thus, we predict that:

In addition, the work environment can also influence the creative outcomes via the individual components through which contextual support is translated into workplace creativity (Amabile, 1996; Woodman et al., 1993). An innovative work climate promotes the sharing of knowledge, skills and experiences across the organisation (Hendrikx et al., 2022), which provides social support resources for employees to develop their creative capacity (Yang and Zhou, 2022). Such a work environment also has higher tolerance for failures that can encourage employees to take risks and try new ideas (Edmondson and Bransby, 2023). When employees feel that their creative efforts are supported by the organisations, they tend to feel more confident in their ability to pursue these creative work endeavours (Yang and Zhou, 2022). With this rationale, we predict that:

Employee’s work efforts have to align with the broader organisational goals and thus a work climate in support of creativity would motivate employees to engage in the creative process (Amabile and Pratt, 2016). In a work climate receptive to change and new ideas, employees will have more freedom and control over how they perform their work, which is likely to enhance their interest in the task (Gumusluoglu and Ilsev, 2009; Hon, 2012). Since creative activity is extra-role behaviour (Yang and Zhou, 2022), employees are also more likely to pay attention and engage in their work when they know that their efforts are recognised and celebrated. In their recent study of innovation orientated human resource systems, Xu et al. (2023) found that employees’ perceptions of innovation culture promote their intrinsic motivation and thus their innovative work behaviour. In line with previous research, we posit that:

Employees respond differently to workplace challenges and learning opportunities, which can have major implications for their creative process and performance (Amabile and Pratt, 2016). A key explanatory factor is LGO, which depicts one’s desire to develop his/her ability in an achievement situation (Hendrikx et al., 2022; VandeWalle, 1997). The value of LGO lies in regulating employees’ attention and effort towards self-improvement by mastering new skills, adapting to new circumstances and improving personal competences (Hirst et al., 2009). Given the focus on personal learning and development, learning-orientated individuals seek out opportunities to hone their skills and are more willing to take on challenges (Cerasoli and Ford, 2014).

Given the open attitude towards learning, LGO is found to be a facilitator of learning behaviours in the workplace (Hendrikx et al., 2022). Employees holding this orientation take a more proactive approach towards their personal learning and development at work through pursuing both formal (e.g. participating in training) and informal (e.g. exchanging information with colleagues) learning opportunities (Hendrikx et al., 2022; Santoro, 2022). Moreover, their strong desire to acquire knowledge and openness to new experiences lead them to engage in exploration activities and actively search for new information in their workplace (Tanaka, 2022). These learning behaviours allow them to develop job-relevant skills and knowledge, through which it becomes easier for them to recognise creativity opportunities and solve problems at work (Hirst et al., 2009; Lee and Yang, 2015). As such, various studies have linked LGO to higher workplace creative behaviour (Atitumpong and Badir, 2018; Hirst et al., 2009; Lee and Yang, 2015). Thus, we predict that:

LGO has also been suggested by previous research to be an influential determinant of one’s creative confidence (Atitumpong and Badir, 2018; Gong et al., 2009). Firstly, LGO is grounded in an incremental conception of ability which views competences as amenable to change through effort and practice (Hendrikx et al., 2022). Thus, individuals with high LGO are more likely to view creativity as a malleable and developable skill, bolstering their self-efficacy beliefs (Atitumpong and Badir, 2018). Secondly, given its focus on self-improvement and developing successful mastery, these individuals are more likely to accumulate skills and knowledge, thereby enhancing their personal and creative confidence (Gong et al., 2009). Further, the focus on self-improvement can also shield these individuals from the evaluative judgement of others, thereby nurturing their sense of self-efficacy (Gong et al., 2009). Thus, we posit that:

However, not all individuals exhibit the same degree of LGO and one’s interest and enjoyment of an activity serves as a potential predictor of this individual difference. While learning goals provide direction for how people work, intrinsic motivation supplies the energy that sustains their effort (Cerasoli and Ford, 2014). Thus, goal orientation acts as a medium through which intrinsic motivation is translated into enhanced performance. Indeed, employees are unlikely to invest in the development of job-related competences unless they derive intrinsic satisfaction from their job in the first place (Nguyen et al., 2023). As such, when employees enjoy their job, they will be more inclined to get better at it, propelling them to search for opportunities to acquire new job-related skills. All things considered, we predict that:

Finally, since employees acquire work-related knowledge and skill in the organisational environment (Li and Tsai, 2020), work climate is likely to influence individuals’ LGO. In a positive climate supportive of creativity, employees would perceive the organisation as being more open and responsive to their ideas and suggestions for improvements (De Stobbeleir et al., 2011). Such context signals a strong commitment to learning that encourages employees’ reciprocal behaviours by developing their competence and engaging in feedback-seeking and continuous learning behaviour (Li and Tsai, 2020; De Stobbeleir et al., 2011). Moreover, by supporting creativity, the organisation fosters a psychologically safe space in which employees can try new approaches without the fear of being regarded as incompetent (Edmondson and Bransby, 2023; Li and Tsai, 2020). Such an environment would signal that personal growth and skill development are valued, encouraging employees to adopt learning-focused goals (Hendrikx et al., 2022):

The overall theoretical model along with the proposed hypotheses in the current study can be examined in Figure 1.

The current study forms part of a broader project examining the feasibility of a one-day creativity training for administrative staff working in Vietnam’s public universities. Given the restrictive institutional barriers in public organisations that often limit access for external researchers without established connections, purposeful convenience sampling was used to identify prospective participating universities. Specifically, the researchers leveraged a professional connection with an institutional gatekeeper within Vietnam’s higher education system who had extensive networks with other universities. Acting as an intermediary, this gatekeeper searched for potential university leads and facilitated connections with the research team. Three public universities located across Vietnam, representing the Northern, Central and Southern regions, agreed to host the training. Each university appointed a representative to manage participant recruitment and workshop organisation. Training brochure was circulated internally across administrative departments and interested staff registered with the appointed representative. To maintain instructional effectiveness, each class was capped at 30 participants, with registrations processed on a first-come, first-served and voluntary basis.

The data were collected using a paper-and-pen questionnaire. Because this study was part of a larger quasi-experimental investigation on the limited efficacy of one-day creativity training workshops, data for the broader study was first collected. This included a two-minute Alternative Uses Test and survey questionnaire for CSE, which are both measured again at the end of the training. The former is a widely used assessment of divergent thinking in creativity, which requires participants to generate different creative uses for a common object. Participants then answered survey questions for this study about demographic information, intrinsic motivation, workplace creativity, LGO and POSfC, in that order, without time limits.

Data was collected on the training day in the same classroom where the training was conducted at each university. The training facilitator distributed the questionnaire at the start of the workshop, immediately after explaining the study purpose and confidentiality and before content delivery. Participants completed the questionnaire individually inside the training room under the facilitator’s presence taking, on average, 9–11 min to complete. Responses were initially identifiable to enable pre-post data matching, which were subsequently pseudonymized to help protect participants’ privacy. Ninety-eight participants completed the study measures. In prioritising data integrity and internal validity of the study, a complete-case approach without imputation was adopted. Accordingly, cases with missing responses on any item of the scales were excluded from the analysis, resulting in a final sample of 94. Table 1 presents the demographic information of the participants.

The theoretical constructs in our study were operationalised and measured using self-report instruments. All the items were adopted from previous research and translated into Vietnamese using a back-translation procedure. Forward translation into Vietnamese was first completed by the first author, who is also a Vietnamese native speaker. These were then back-translated by another bilanguage lecturer in the area linguistics who was independent to the current study. Finally, the final version was reviewed to ensure semantic and conceptual equivalence by the second author, who is a senior researcher on Vietnam higher education and survey research.

CSE was measured using the three-item scale developed by Tierney and Farmer (2002) on a Likert-type agreement scale. As CSE also served as a pre-posttest outcome in our broader quasi-experimental study, it was measured on a 7-point scale, which has been shown to be more sensitive to detecting nuanced change following creativity training (Mathisen and Bronnick, 2009). All other constructs were measured on 5-point Likert scales to minimise the potential cognitive load. Sample item includes “I have confidence in my ability to solve problems creatively”. Previous studies reported 0.91 (Gong et al., 2009) and 0.82 (Atitumpong and Badir, 2018), with satisfactory CFA results.

Workplace creativity was operationalised in this study as frequency of engaging in creative behaviour and measured using Soda et al. (2019)’s four-item scale on a 5-point Likert frequency scale. Sample item includes “I provide new ideas to improve the department’s performance”. The same items were validated by Thuan and Thanh (2019), reporting a reliability of 0.85 and a convergent validity of 0.58.

Intrinsic motivation was measured using the three-item intrinsic motivation subscale from Tremblay et al.’s (2009) Work Extrinsic and Intrinsic Motivation Scale on a 5-point Likert agreement scale. Participants responded to a stem question and a sample item includes “Because I derive much pleasure from learning new things”. This measure has previously demonstrated a reliability of 0.75 and a convergent validity of 0.57 (Gupta, 2020).

LGO was measured using VandeWalle’s (1997) five-item scale on a 5-point Likert agreement scale. Sample item includes “I often look for opportunities to develop new skills and knowledge”. Using the same scale, Nguyen et al. (2023) reported a reliability of 0.74 and convergent validity of 0.63.

POSfC was measured using the four-item questionnaire developed by Zhou and George (2001) using a 5-point Likert agreement scale. Sample item includes: “Our ability to function creatively is respected by the leadership”. The reliability of this scale was reported to be 0.80 in a recent study by Pan et al. (2020) with CFA indicating a satisfactory fit.

Given the small sample size, Bayesian structural equation modelling (BSEM) was used to examine the complex relationships between the variables in this study. Bayesian inference is well suited to limited data because it uses Monte Carlo simulation to derive the distribution of the parameters rather than relying on asymptotic and large sample theory (Garnier-Villarreal and Jorgensen, 2020). It also allows direct incorporation of prior evidence into the modelling process to stabilise estimates when data is limited (van de Schoot et al., 2015). Accordingly, meta-analytic results were used to update prior knowledge on the impact of intrinsic motivation, CSE and LGO on workplace creativity with prior means of 0.16, 0.21 and 0.29, respectively (SD = 1.00 for all) (Liu et al., 2016; Nandi and Watts, 2025). We also conducted a Bayesian version of confirmatory factor analysis (BCFA) as the first step to assess the measurement model, followed by BSEM to examine the structural relations.

To evaluate Bayesian model fit, the posterior predictive p-value (PPp) is often used, with a value of 0.5 indicating adequate fit (Garnier-Villarreal and Jorgensen, 2020). However, PPp is not considered a reliable fit indicator for complex models. Accordingly, alternative fit indices are recommended, including the Bayesian adaptation of root mean square error of approximation (BRMSEA), gamma hat (B γ^) and comparative fit index (BCFI) (Garnier-Villarreal and Jorgensen, 2020). These indices are expressed as a distribution of statistics computed at each posterior draw, allowing model fit and associated uncertainty to be evaluated through variation across the posterior samples. Because the Bayesian version of these indices behaves similarly to their frequentist counterparts, traditional SEM cut-off guidelines can still be applied for descriptive purposes, but not for hypothesis-testing criteria (Garnier-Villarreal and Jorgensen, 2020).

All Bayesian analyses were conducted using the blavaan and stan packages in R (Merkle and Rosseel, 2018). Models were estimated using Hamiltonian Monte Carlo with three chains, 5,000 iterations per chain and a burn-in phase of 2,000 iterations. To enhance transparency, replicability and validation, Depaoli and van de Schoot (2017)’s WAMBS-checklist was applied (see  Appendix). To facilitate comparison with frequentist null hypothesis testing, the probability of direction (PD) was calculated for each path, representing the proportion of posterior distribution that aligns with the reported parameter estimate (Makowski et al., 2019). PD is strongly related to the frequentist p-value, with PD > 0.975 corresponding to a two-sided p-value < 0.05 and PD > 0.95 corresponding to a one-sided p-value < 0.05 (Makowski et al., 2019).

Table 2 displays the results of the descriptive statistics and bivariate correlations. All predictors demonstrated a positive correlation with workplace creativity, with the exception of POSfC (r = 0.12, ns). Among the predictors, significant positive correlations were observed for three pairs, which are CSE and intrinsic motivation (r = 0.24, p < 0.05), LGO and intrinsic motivation (r = 0.33, p < 0.05), LGO and POSfC (r = 0.24, p < 0.01).

The initial BCFA of the hypothesised measurement model indicated good fit statistics for B γ^ (M = 0.926, SD = 0.011) and BCFI (M = 0.922, SD = 0.012), but not for BRMSEA (M = 0.081, SD = 0.005) and PPp (0.004). Item 5 of LGO was removed due to its low factor loading and Item 4 of POSfC was removed based on high residual correlation identified through the modification indices. The final measurement model indicated an adequate fit to the data with BRMSEA (M = 0.067, SD = 0.008), B γ^ (M = 0.949, SD = 0.011) and BCFI (M = 0.943, SD = 0.013) all within the acceptable range, apart from PPp (0.075). Table 3 presents the factor loadings for items, along with the consistency and validity measures for each theoretical construct. All retained items showed sufficient factor loadings and all constructs met the recommended thresholds for reliability and convergent validity.

The structural model successfully converged, with all Rhat values at the optimal value of 1 and effective sample sizes > 1,000. BRMSEA (M = 0.067, SD = 0.008), B γ^ (M = 0.948, SD = 0.011) and BCFI (M = 0.942, SD = 0.013) indicated a good fit of the model with the data, but not for PPp (0.072). The WAMBS checklist procedures also supported the model convergence (see  Appendix).

The model predicted 44.9% of variance in administrative staff’s workplace creativity. Posterior means, standard deviations, credible intervals for all paths are displayed in Table 4 and their 89% posterior interval estimates are plotted in Figure 2.

Regarding predictors of workplace creativity, we found positive non-zero direct effects for CSE (H2), LGO (H6) and intrinsic motivation (H1), but not for POSfC (H3). Among the individual predictors, intrinsic motivation had a positive non-zero direct effect on LGO (H8a) and LGO had a positive non-zero direct effect on CSE (H7a). Regarding the indirect influence on workplace creativity, LGO exhibited a positive non-zero indirect effect via CSE (H7b), while intrinsic motivation also exhibited a non-zero indirect effect via LGO (H8b). Finally, the influence of POSfC on the three remaining predictors was examined. Non-zero positive direct effect was found for intrinsic motivation (H5a) and LGO (H9a), but not via CSE (H4a). Consistent with this pattern, POSfC only showed a positive non-zero indirect effect on workplace creativity via intrinsic motivation (H5b) and LGO (H9b), but not CSE (H4b).

Our study examined predictors of administrative staff’s workplace creativity in Vietnamese public universities. As theorised, CSE (β = 0.296; PD = 100%; 89% CI [0.173, 0.428]) and intrinsic motivation (β = 0.228; PD = 97%; [0.036, 0.428]) both positively predicted workplace creativity, with a stronger effect for the former. Contrary to our expectation, POSfC did not predict workplace creativity (β = 0.024; PD = 63.8%; [−0.093, 0.138]). This may be explained by the proximal focus argument, in which employees respond more strongly to the nearer sources of influence in the work environment (Diliello et al., 2011), such as the influence of a more immediate supervisor or manager (Suifan et al., 2018). As such, organisational support for creativity likely reflects generic, distal support that administrative staff do not directly experience, limiting its translation into creative work behaviours. On the other hand, POSfC can positively predict intrinsic motivation (β = 0.152; PD = 98%; [0.033, 0.283]), while showing an inconsistent positive association with CSE (β = 0.240; PD = 63.8%; [−0.032, 0.363]). This pattern supports POSfC’s function as an autonomy-supportive environment that can enhance staff’s freedom and interest, encouraging them to seek improvement in their jobs and engage in creative behaviours (Nili and Tasavori, 2022). In line with our expectations, POSfC appears to affect workplace creativity indirectly via intrinsic motivation and CSE.

We further extend the componential model by incorporating LGO as a new explanatory variable. Consistent with prior research (Atitumpong and Badir, 2018; Gong et al., 2009), LGO predicted CSE (β = 0.533; 96.6%; [0.059, 1.091]) and workplace creativity (β = 0.392; 98.8%; [0.106, 0.721]), with CSE also partially mediating the relationship. However, the magnitudes were imprecise, with values ranging from minimal to substantial. We also found that intrinsic motivation predicted LGO (β = 0.237; 99.5%; [0.082, 0.419]) and exerts a positive indirect effect on workplace creativity via LGO, supporting the self-determination perspective that intrinsic motivation precedes LGO (Cerasoli and Ford, 2014). According to this view, intrinsic motivation fuels one’s energy and focus, while learning goals operate like a steering wheel directing such drive and cognition towards competence-relevant behaviours that foster workplace creativity (Cerasoli and Ford, 2014). Finally, POSfC may also predict LGO (β = 0.089; 95%; [0.002, 0.189]), thereby indirectly influencing workplace creativity. This observation supports the role of innovative work climate in communicating and signalling a learning climate that encourages staff to engage in competencies-building activities that foster creative behaviour (De Stobbeleir et al., 2011; Hendrikx et al., 2022). However, this interpretation should be treated with caution, as both the direct and indirect effects of POSfC are marginal and include negative estimates.

Taken together, our findings provide preliminary evidence extending Amabile (1996)’s componential model to a non-Western context characterised by collectivism and high power distance, such as Vietnam. Cross-cultural creativity research suggests that key individual drivers of creativity are often relevant across cultural contexts (Xie and Paik, 2019). As such, the individual predictors identified in this study are not only applicable within the study context but may also have potential relevance beyond it. Consistent with this perspective, the observed absence of a direct relationship between POSfC and workplace creativity may be best understood primarily through a proximal focus lens. The cultural differences in our study may operate as a contextual backdrop that potentially heightens employees’ sensitivity to more immediate supervisors’ reactions while limiting the influence of more distal organisational support (Xie and Paik, 2019). Nevertheless, we acknowledge the limited generalisability of our findings, as the way in which creativity is expressed and evaluated may differ depending on broader national culture and local management norms across study sites (Xie and Paik, 2019).

Our results suggest that leaders from the surveyed universities and in other institutions operating under similar institutional and cultural characteristics, may be able to promote the creative behaviour of their administrative staff by targeting the “will do” (intrinsic motivation), “can do” (CSE) and “want to grow” (LGO) attitudes. As a rapid measure, leaders may consider brief capacity-building training workshops focused on developing learning goals and creative confidence, both of which are trainable (Mathisen and Bronnick, 2009; Tanaka, 2022). Echoing prior research (Grinkevich et al., 2020; Holmes, 2020), our findings on LGO also suggest the value of a more structured approach to continuing professional development (CPD) for university administrative staff that nurtures learning and competence-building behaviours conducive to workplace creativity.

Given intrinsic motivation’s role in fuelling LGO, formal CPD programmes may benefit from moving beyond a one-size-fits-all approach and use a more personalised approach focusing on individual staff development needs. When CPD aligns with personal interest and career goals, it is more likely to promote better learning outcomes (Jiang et al., 2023) and translate into workplace creativity. CPD need not be limited to formal courses, but can also include informal and on-the-job initiatives such as peer mentoring, peer showcase and cross-unit projects (Holmes, 2020). These learning opportunities may help build creative confidence, which is the strongest predictor of workplace creativity in our model. This is because they cultivate mastery through an expanded knowledge base and provide vicarious experience from observing workplace role models, which are the two known sources of developing self-efficacy (Mathisen and Bronnick, 2009).

Building on the direct impact of an innovative climate on intrinsic motivation, universities may make organisational support more tangible through line-managers working with staff to identify meaningful career goals and development opportunities. Managers should set learning-orientated goals tailored to individual needs, thereby strengthening motivation, engagement in CPD and the likelihood that learning translates into workplace creativity. At the institutional level, universities may consider implementing a formal CPD framework with clearly defined competences, transparent pathways and links to career progression with regular check-ins to sustain engagement (Holmes, 2020). Such a framework may also benefit from allocating protected time for CPD to encourage participation (Coomber, 2019).

To a limited extent, the implications of our study may also extend to other public institutions operating under similar institutional and cultural conditions. Indeed, a growing body of research on public sector creativity has emphasised its role in fostering innovation and modernising the public services (Houtgraaf et al., 2021). Given the bureaucratic nature of administrative roles, which prioritise accuracy, standardisation and adherence to formal rules and procedures, staff may become fixated on existing knowledge and are more resistant to change. As such, access to formal and structured CPD programmes may support employees to unlearn outdated practices by engaging in exploration activities to acquire new knowledge and skills that can promote greater efficiency and innovation capacity in services (Tanaka, 2022). Further, when employees feel that their institution is making an investment in their future, they are more likely to be satisfied, engaged and motivated to enhance their contribution (Holmes, 2020).

Our study contains several limitations restricting the conclusion that it can make. Firstly and foremost, the study was conducted on a small sample of administrative staff in Vietnamese public universities using convenience sampling. Generalisation of the findings should therefore be approached with caution beyond culturally and institutionally similar contexts, particularly where cultural and socioeconomic conditions differ substantially. Moreover, our methods of locating participating universities and the study participants may have introduced self-selection bias, further limiting its external validity. Secondly, our cross-sectional design does not allow for causal inference, especially with the relationship between intrinsic motivation and LGO since Cerasoli and Ford (2014) found a reciprocal effect in their study. Future studies can use experimental and longitudinal studies in different contexts to test these causal links. Thirdly, we rely on administrative staff’s self-rating as only our single method of measuring their creativity, which might not be entirely reliable due to the risk of self-bias and social desirability.

In challenging the negative portrayal of administrative staff as sources of universities’ administrative bloat (de Jong and del Junco, 2024), we position them as catalysts of de-bureaucratisation who apply creativity to better support students and academics while streamlining work. Empirically, we show that their workplace creativity can be nurtured through a learning and development mediated pathway. Intrinsic motivation fuels engagement, LGO directs effort towards mastery and CSE converts mastery experiences into confident creative action. Perceived organisational support influences creativity primarily through these individual pathways, implying that universities should prioritise placing a stronger and more structured focus on CPD to cultivate them, rather than expecting climate alone to produce creative behaviour. Together, these patterns extend the componential model with a workplace learning lens, reinforcing the developmental aspect of creativity.

The authors would like to thank the leaders and managers from all of the three public universities in Vietnam for their approval and support to conduct the data collection in this study.

No funding was received for conducting this study.