Original Research

Organisational support and teachers’ performance: The moderating role of job crafting

Annette du Toit, Kleinjan Redelinghuys, Leoni van der Vaart

Received: 08 Mar. 2022; Accepted: 28 June 2022; Published: 28 Sept. 2022

Copyright: © 2022. The Author(s). Licensee: AOSIS.
This is an Open Access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Abstract

Orientation: Teachers fulfil an essential role in students’ learning and the prosperity of nations. Hence, teacher performance, and the determinants thereof, are vital to understand.

Research purpose: To extend the conversation on teacher performance in a non-WEIRD (Western, educated, industrialised and democratic) nation, the authors aimed to investigate potential factors that may influence teachers’ performance, specifically from the perspective of perceived organisational support (POS) and job crafting.

Motivation for the study: Due to the widespread impact of teacher performance and the potential of both POS and job crafting to enable it, it is valuable to investigate the collective effect of these variables on individual work performance.

Research approach/design and method: This quantitative cross-sectional study involved 207 teachers conveniently sampled from private educational organisations in Gauteng. The Survey of Perceived Organisational Support, Job Crafting Questionnaire and an Individual Work Performance Subscale were administered to assess the study variables. Structural equation modelling was employed to confirm the dimensionality of the scales, followed by moderation analysis for hypothesis testing.

Main findings: The results of the moderation analysis showed that the effect of POS on teachers’ performance is conditional upon teachers’ job crafting behaviours in the organisation. More specifically, organisational support matters for teachers’ performance but only for those with low to moderate levels of job crafting.

Practical/managerial implications: Organisations could implement interventions to enhance teachers’ perceptions of support from the organisation to improve their performance. Simultaneously, organisations can invest in interventions that teach teachers to craft their jobs and create organisational environments that foster job crafting behaviours.

Contribution/value-add: The study contributes to the limited body of literature on teachers’ performance in a developing context and literature on organisational support and job crafting.

Keywords: adaptivity; job crafting; moderation; perceived organisational support; proactivity; proficiency; teacher performance.

Introduction

‘Teachers are plausibly the most important group of professionals for any nation’s future’ (Crisci, Sepe, & Malafronte, 2019, p. 2404). Unfortunately, the treatment they receive is not always reflective of their importance. For instance, on further inspection of the 2018 OECD Teaching and Learning International Survey (TALIS) data, Schleicher (2020) noted that only one in four teachers felt that society appreciated their occupation. The TALIS data also showed that South Africa scored substantially higher than the Organisation for Economic Co-operation and Development (OECD) average on several undesirable indicators. Among others, these include (1) weekly incidences of bullying or intimidation, (2) the use or possession of illegal substances on the school premises, (3) theft and vandalism, (4) shortages related to library materials and technology and (5) a lack of physical infrastructure and support staff (OECD, 2019). Teachers’ contributions, the stumbling blocks they face and the notion that South African teachers spend less time teaching than those in other countries (OECD, 2019) highlight the importance of enabling teachers to perform optimally in delivering top-quality education (Redelinghuys, Rothmann, & Botha, 2019).

Despite its importance, teacher performance receives little attention from researchers in South Africa. Two of the few studies published in South Africa supported the positive association between teacher performance and organisational support¹ (Chinomona & Sandada, 2014; Van Der Merwe & Keyser, 2014). However, organisational support (as perceived by the teacher) only represents one side of the coin. Researchers believe that two approaches matter for improving employee outcomes: ‘top-down’ and ‘bottom-up’. Whereas organisational support is considered a top-down approach, employee job crafting is a bottom-up approach (Bakker, 2015; Van Wingerden & Van Der Vaart, 2019). Rudolph, Katza, Lavigne and Zacher’s (2017) meta-analysis shows that job crafting has a positive effect on individual work performance. Others have also demonstrated the positive outcomes of job crafting for teachers in South Africa (Peral & Geldenhuys, 2016). In the teaching context, job crafting is an adaptation process (Ghitulescu, 2007; Peral & Geldenhuys, 2016) that teachers use to tackle job demands and find meaning in their work (Ghitulescu, 2007; Van Wingerden & Poell, 2019). To extend the conversation on teacher performance in a non-WEIRD (Western, educated, industrialised and democratic) nation, the authors aimed to investigate potential factors that may influence teachers’ performance, specifically from the perspective of perceived organisational support (POS) and job crafting. In extending the conversation, the study contributes not only to the limited literature on teachers’ performance in South Africa, but also to job crafting literature in this context.

A few job crafting studies exist in the South African context (e.g. Peral & Geldenhuys, 2016, 2020; Sloan & Geldenhuys, 2021; Thomas, Du Plessis, & Thomas, 2020). Despite the value-add of these studies, several gaps exist. Firstly, these studies focused on industries (e.g. financial, manufacturing, mining) other than education (except Peral & Geldenhuys, 2016). Secondly, these studies (except Sloan & Geldenhuys, 2021) focused more on well-being (e.g. work engagement, psychological meaningfulness) than performance outcomes. This is unfortunate given the United Nations’ (2020a) sustainable goal of prioritising teacher performance. Thirdly, these studies focused more on individual determinants of job crafting (e.g. personality, emotional intelligence), than the contextual determinants. Considering that research has found that management and the broader work environment are important drivers of teacher performance (Ingusci, Callea, Chirumbolo, & Urbini, 2016), it can be argued that an understanding of the role of the organisational context on performance is important (Caesens, Stinglhamber, & Ohana, 2016). Similarly, Zhang and Parker (2019) argued that an understanding of the contextual determinants of job crafting is important as it is the basis of these proactive behaviours. Previous job crafting studies in the South African context are not only limited in scope (as illustrated here); they are also limited in number. As job crafting is context-dependent (Zhang & Parker, 2019), the highly complex South African education landscape (World Education News and Reviews, 2017) necessitates more local research.

Literature review

The support that teachers receive largely determines how educational systems function (Nkambule & Amsterdam, 2018). Perceived organisational support is a unidimensional construct encompassing employee beliefs or perceptions ‘that their organisation values their contributions and cares about their well-being’ (Eisenberger, Huntington, Hutchison, & Sowa, 1986, p. 501). Organisational support theory implies that employees are inclined to assign human-like attributes to their organisation (Eisenberger et al., 1986). This determines whether employees regard the good or bad treatment they receive from their organisation as a sign that they are favoured or disfavoured (Rhoades & Eisenberger, 2002). Gouldner (1960) advises that discretionary aid (i.e. voluntary acts by organisations to assist employees) is valued far more by employees than forced or compulsory aid (e.g. where labour unions pressure organisations to abide by their demands). Discretionary assistance indicates that organisations value and care about employees’ welfare. When employees are treated well, they may feel the need to return the ‘favour’ through enhanced performance, among other things (Kurtessis et al., 2017). Therefore, POS may have a positive impact on teachers’ performance. Apart from ‘top-down’ approaches, ‘bottom-up’ (e.g. proactive behaviours initiated by employees) approaches may also help enhance individual work performance (Rudolph et al., 2017). Job crafting is an example of this (Grant & Parker, 2009).

Job crafting reflects deliberate acts of employees to make alterations to their job and/or work environment to achieve a particular outcome (Tims, Bakker, & Derks, 2012; Wrzesniewski & Dutton, 2001). While Wrzesniewski and Dutton (2001) suggest that the attainment of meaningful work is the end goal of job crafting, Tims et al. (2012) argue that job crafting serves the purpose of finding an equilibrium between job demands and resources to enhance person-job fit. Consequently, the motives behind employee job crafting behaviours may differ (Zhang & Parker, 2019). The current study used Wrzesniewski and Dutton’s (2001) job crafting definition. External resources are not always available in non-WEIRD contexts, so meaningful work emerges as an essential intrinsic resource for teachers (Janik & Rothmann, 2015). Wrzesniewski and Dutton (2001) define ‘job crafting as the physical and cognitive changes individuals make in the task or relational boundaries of their work’ (p. 179). Three job crafting types are thus emphasised: (1) task, (2) cognitive and (3) relational crafting. Task crafting entails how employees approach their formal work duties, such as changing the nature of their tasks, spending more or less time on specific tasks or discarding certain tasks completely (Berg, Dutton, & Wrzesniewski, 2013). For instance, teachers may seek creative ways to teach and assess content, initiate new projects or engage in tasks that are not formally required (Huang, Sun, & Wang, 2022). Cognitive crafting involves changes in the way employees think about work aspects such as tasks and relationships (Berg et al., 2013). For instance, teachers may shift perspectives from ‘I often have to deal with ungrateful learners and parents’ to ‘my job enables me to change children’s lives indefinitely’, where the meaning attached to the latter perspective comfortably outweighs the former. Lastly, relational crafting refers to how employees structure their social interactions (Berg et al., 2013). It may manifest by creating stronger bonds with colleagues or purposefully decreasing toxic work interactions (Slemp, 2017). Once employees have the leeway to alter certain job aspects, they can potentially enhance their performance (e.g. Lee & Lee, 2018).

In literature, studies often focus on unidimensional accounts of individual work performance (i.e. focusing solely on in-role or task performance at the expense of other potentially helpful performance components). However, numerous authors (e.g. Carpini, Parker, & Griffin, 2017; Griffin, Neal, & Parker, 2007; Van Der Vaart, 2021; Van Lill & Taylor, 2022) advocate for a multidimensional individual work performance framework. Griffin et al.’s (2007) conceptualisation of individual work performance was used in the current study. Griffin et al.’s (2007) triadic model includes (1) task proficiency, (2) task adaptivity and (3) task proactivity. Task proficiency, which corresponds closely to task performance, reflects the abilities of employees to carry out their core work tasks in an acceptable fashion (Griffin et al., 2007). Task proficiency, in a teaching context, would refer to teachers presenting lessons and conducting assessments, for example. Task adaptivity involves the extent to which employees can manage work changes (e.g. dealing with the introduction of new apparatus) (Griffin et al., 2007). An example of task adaptability in a teaching context would be teachers adapting to learner’s different needs or changing circumstances. Adaptability is an essential asset for teachers, as the teaching profession is subject to continuous change (Collie, Granziera, Martin, Burns, & Holliman, 2020). This change is evident from the widespread school closures attributed to coronavirus disease 2019 (COVID-19), where teachers worldwide had to quickly shift from classroom teaching to online-based teaching (König, Jäger-Biela, & Glutsch, 2020). Finally, task proactivity entails self-initiated and future-directed employee (e.g. teachers) behaviours aimed at changing themselves, their work roles or their circumstances at work (Griffin et al., 2007). Task proactivity, in a teaching context, would refer to teachers anticipating student needs and responding to it, for example. Due to the widespread impact of teacher performance and the potential of both POS and job crafting to enable it, it is valuable to investigate the collective effect of these variables on individual work performance.

To the best of the authors’ knowledge, no study has focused on job crafting’s moderating role on the POS-performance relationship. One study (Cheng, Chen, Teng, & Yen, 2016) investigated the moderating effect of POS in the job crafting–job outcomes relationship (job satisfaction, organisational commitment, and individual work performance). Nevertheless, Cheng et al.’s (2016) results showed that individual job crafting is associated positively with performance and that this association depended on the tour leaders’ levels of POS. The current study deviates from the former in the following ways: (1) using a different conceptual framework of job crafting (meaning instead of fit), (2) individual work performance is not restricted to in-role performance and (3) the use of job crafting as a moderator as opposed to POS as moderator.

Job crafting has been tested as a moderator in the perceived overqualification–job boredom (Sánchez-Cardona, Vera, Martínez-Lugo, Rodríguez-Montalbán, & Marrero-Centeno, 2020), job stress–counterproductive work behaviour (Weber, 2019), job demands–burnout (Hakanen, Seppälä, & Peeters, 2017), job demands–work engagement (Hakanen et al., 2017), work engagement–team performance (Mӓkikangas et al., 2016), work conditions (coworker support, work autonomy, role ambiguity)–work engagement (Muningua, 2019), overqualification–turnover intention (Debus, Gross, & Kleinmann, 2020) and work-related boredom–bored behaviour (Van Hooff & Van Hooft, 2014) relationships. Thus, few studies have tested job crafting’s moderating effect, whereas none of them used POS as the focal predictor. Only one study included performance, but they included team performance as opposed to individual performance (Mӓkikangas et al., 2016).

Theoretical framework

Although the study focused on job crafting from a meaning-creating perspective (i.e. Wrzesniewski & Dutton’s conceptualisation of the construct) as opposed to a person–job fit perspective, the job demands–resources (JD–R) theory (Demerouti, Bakker, Nachreiner, & Schaufeli, 2001) may explain the proposed relationships. In short, the JD–R theory proposes that when employees’ job demands overpower their job resources, they may be subjected to burnout and other negative outcomes (e.g. impaired health) (Schaufeli & Bakker, 2004). However, when employees possess sufficient job resources to manage their job demands, they may experience work engagement and other positive outcomes (e.g. decreased turnover intention) (Schaufeli & Bakker, 2004). The current study variables fit into the JD–R framework, as the latter framework is not confined to specific job demands or resources, nor to specific personal resources or outcomes (Schaufeli & Taris, 2014). Consequently, the JD–R model can be customised to a broad range of work settings (Schaufeli & Taris, 2014). In layman’s terms, job demands are seen as ‘the “bad things” at work that drain energy’ (Schaufeli, 2017, p. 121). Among others, these include unfavourable work conditions, work pressure and work overload (Schaufeli & Taris, 2014). In contrast, job resources are seen as the ‘good’ aspects of one’s work (Schaufeli, 2017) and include factors like appreciation, autonomy, skill utilisation and task variety (Schaufeli & Taris, 2014). Personal resources predominantly reflect one’s ‘ability to control and impact one’s environment successfully’ (Schaufeli & Taris, 2014, p. 49).

In the current study’s context, POS thus closely aligns with the description of job resources, whereas job crafting aligns with the conceptualisaion of personal resources. Job resources (i.e. POS) typically initiate a motivational process leading to positive outcomes (i.e. individual work performance). Personal resources are positioned differently within the JD–R framework (Schaufeli & Taris, 2014). Consequently, there is no concluding evidence that prescribes the role of personal resources (e.g. solely as a mediator) (Schaufeli & Taris, 2014). Therefore, in different studies, personal resources were used successfully as mediators, moderators or in other capacities (Schaufeli & Taris, 2014). ‘The definition of personal resources implies that they may buffer negative effects of job demands on burnout and exacerbate positive effects of job resources on engagement’ (Schaufeli & Taris, 2014, p. 49). Hence, it is possible that job crafting (as personal resource) could strengthen the impact of POS (as job resource) on individual work performance (as the outcome). Apart from the JD–R theory, several empirical studies support the positive association between POS and different facets of individual work performance (Ahmed, Nawaz, Ali, & Islam, 2015; Kurtessis et al., 2017). Similarly, studies suggest that job crafting may also positively influence individual work performance (Lee & Lee, 2018; Peral & Geldenhuys, 2020; Rudolph et al., 2017). Consequently, job crafting may amplify the positive effect of POS on individual work performance. Based on this notion, it is hypothesised that job crafting moderates the relationship between POS and individual work performance.

Method

Participants and procedure

This cross-sectional study involved teachers (n = 207) from private educational organisations in Gauteng. Online questionnaires, which took 5–10 min to complete, were set up via Google Forms and sent to participating teachers’ email addresses. Links to the questionnaire were also posted on social media platforms (Facebook, LinkedIn) for further reach. All ethical requirements (i.e. anonymity, voluntary participation) were evaluated and approved by an ethical committee at a reputable institute of higher education before data collection commenced (reference number NWU-00482-19-A).

The majority of the participants were female (87.4%), between the ages of 20 and 40 years (69.1%). Participants in the resident teacher (27.5%), senior teacher (n = 59, 28.5%) and middle management (22.2%) roles were almost equally represented. Almost half of the participants had been employed for between 1 and 10 years (48.1%).

Measures

Perceived organisational support was measured by an adapted version of the Survey of Perceived Organisational Support (Eisenberger et al., 1986). Accordingly, the adapted nine-item unidimensional scale (Eisenberger, Cummings, Armeli, & Lynch, 1997) was used instead of the original 36-item measure. Each statement was rated on a scale of 1 (strongly agree) to 5 (strongly disagree). A sample item from the instrument is ‘My organisation really cares about my well-being’. A good reliability coefficient (α = 0.90) has been established for the scale (Eisenberger et al., 1997).

Job crafting was assessed by an adjusted version of the 15-item Job Crafting Questionnaire (Slemp & Vella-Brodrick, 2013). This nine-item version (three items per dimension), adapted by Kim, Im, Qu and NamKoong (2018), asked participants to indicate their task, cognitive and relational crafting behaviours. The frequency of job crafting behaviour was indicated on a scale of 1 (hardly ever) to 5 (very often). A sample item measuring relational crafting is ‘I make friends with people at work’. A sample item measuring cognitive crafting is ‘I think about how my job gives my life purpose’. Lastly, a sample item measuring task crafting includes ‘I change the scope or types of tasks that I complete at work’. A good reliability coefficient (α = 0.91) has been established for the total scale (Slemp & Vella-Brodrick, 2013).

Individual work performance was assessed by the nine-item Individual Performance Subscale developed by Griffin et al. (2007). Nine items (three items per dimension) pertain to individual-level performance, including task proficiency, task adaptivity and task proactivity. Participants rated their past month’s performance on a scale of 1 (very little) to 5 (a great deal). A sample item that measures proficiency is ‘I have carried out the core parts of my job well’. A sample item measuring adaptivity is ‘I adapted well to changes in core tasks’. Lastly, a sample item measuring proactivity is ‘I made changes to the way my core tasks are done’. Cronbach’s alpha coefficients larger than 0.80 have been established for the scale (Griffin et al., 2007).

Statistical analysis

Statistical analysis of the data was conducted with Mplus 8.4 (Muthén & Muthén, 1998–2020). Descriptive statistics (i.e. means and standard deviations) were computed with SPSS 26 (IBM Corp, 2016). The measurement model with the best possible fit was determined through confirmatory factor analysis in a latent variable modelling framework (as recommended by Kline, 2016). The robust version of the maximum likelihood (MLR) estimator was used to deal with the negatively skewed distribution of the job crafting and individual work performance data (as recommended by Wang & Wang, 2020).

As the data were cross-sectional, competing measurement models were specified to identify the most optimal model. Numerous goodness-of-fit indices were used to evaluate the fit of the data to the model and to compare the fit of the competing models. Wang and Wang (2020) suggest making use of the following fit statistics to assess the model fit: the chi-square statistic (χ²), root mean square error of approximation (RMSEA), the Tucker–Lewis index (TLI), comparative fit index (CFI) and the standardised root mean square residual (SRMR). Values exceeding 0.90 were considered satisfactory for CFI and TLI, and values equal to below 0.08 for RMSEA and SRMR indicated sufficient model fit to the data. The best model was selected using the Akaike Information Criterion (AIC) and Bayes Information Criterion (BIC) values (Wang & Wang, 2020). The practical significance of correlation coefficients was determined by the following effect sizes: r = ≥ 0.10 (small effect), r = ≥ 0.30 (medium effect) and r ≥ 0.50 (large effect) (Cohen, 1988). Meaningful associations were indicated by p-values smaller than 0.05.

Model 1, considered a simple moderation model (Hayes, 2022) from the processR app (Moon, 2019), was used to determine job crafting’s moderating effect. To control for measurement errors, factor scores were saved from the best-fitting measurement model (as recommended by Skrondal & Laake, 2001). These factor scores were then used in the moderation analysis. According to Hayes (2022), moderation analysis consists of two steps. Firstly, the regression coefficient for the interaction effect is calculated to determine how a change in the moderator is related to changes in the effect of the focal predictor. Secondly, if the interaction effect is significantly different from zero in the first step, the nature of the interaction is probed. In this step, the moderation is visualised to understand how the focal predictor’s effect changes as a function of the moderator. Inferential tests, such as the Johnson–Neyman (J–N) technique, are preferred. Using this technique, one can determine a range of moderator values where the relationship between the focal predictor and the outcome variables is statistically significant or not (Hayes, 2022).

The reliability of the measuring instruments was evaluated by using Raykov’s rho coefficients (Raykov, 2009), where a reliability coefficient above 0.70 is ‘satisfactory to good’ (Hair, Risher, Sarstedt, & Ringle, 2019, p. 8).

Ethical considerations

Approval to conduct the study was obtained from the Health Research Ethics Committee (HREC), North-West University (reference number: NWU-00482-19-A1).

Results

Confirmatory factor analysis

A post-hoc power analysis was performed using the pwrSEM app (Wang & Rhemtulla, 2021). The post-hoc analyses revealed that a sample size of 187 was enough to detect an interaction effect of 0.15 at a 0.80 level of statistical power. Two competing models were tested to determine the most optimal model. In Model 1, POS was specified as a unidimensional construct (all items load onto the a-priori construct), and job crafting and individual work performance were specified as higher-order constructs (items load onto their a-priori constructs and the two lower-order constructs load onto the higher-order constructs). In both cases, the lower-order constructs were not allowed to correlate (Morin, Myers, & Lee, 2020), but POS, job crafting and individual work performance were allowed to correlate. In Model 2, all constructs were specified as unidimensional. Again, all three variables were allowed to correlate. Table 1 presents the results of the measurement models.

Compared to Model 2, Model 1 is the more complex model. Consequently, its fit statistics will be better. However, the AIC and BIC values penalise for model complexity, with lower values being preferred. Wang and Wang (2020) consider a change of more than 10 in the BIC value as strong evidence favouring the more complex model. Based on this (∆BIC Model 2 vs. Model 1 = −123.05), Model 1 was the best-fitting model despite its complexity. Factor loadings for POS ranged from 0.67 to 0.89 and were all statistically significant (p ≤ 0.001). Factor loadings for relational crafting ranged from 0.65 to 0.86, cognitive crafting from 0.65 to 0.86 and task crafting from 0.50 to 0.74. All these factor loadings were statistically significant (p ≤ 0.001). Relational crafting (β = 0.85), cognitive crafting (β = 0.78) and task crafting (β = 0.95) loaded significantly onto the higher-order job crafting factor. Factor loadings for proficiency ranged from 0.82 to 0.85, adaptivity from 0.75 to 0.82 and proactivity from 0.71 to 0.87. All these factor loadings were statistically significant (p ≤ 0.001). Proficiency (β = 0.92), adaptation (β = 0.95) and proactivity (β = 0.98) loaded significantly onto the higher-order individual work performance factor.

Correlation and reliability coefficients are reported in Table 2.

The reliability coefficients are within the acceptable range (i.e. ≥ 0.70). The correlations between POS and job crafting (r = −0.03), as well as between POS and individual work performance (r = 0.12), were nonsignificant. When effects are conditional (such as in the case of POS on individual work performance), it is not surprising that the nonzero order correlations are not significant. The correlation between job crafting and individual work performance was statistically significant (r = 0.73, large effect).

Moderation analysis

The moderation analysis results, assessing whether job crafting moderates the relationship between POS and individual work performance, are reported in Table 3. It is evident from the results that the conditional effect of POS on individual work performance (i.e. the simple effect of POS when job crafting = 0) was significant (β = 0.12, p < 0.001). Similarly, the conditional effect of job crafting on individual work performance (i.e. the simple effect of job crafting when POS = 0) was significant (β = 0.73, p < 0.001). The interaction was also statistically significant (β = −0.19, p < 0.001), meaning the interaction effect is meaningfully different from zero. Therefore, one can conclude that job crafting is a significant moderator of the association between POS and individual work performance. In addition to the p-values lower than 0.05, the confidence intervals (CIs) also did not include zero. Such CIs further support the conditional effects of POS and job crafting and the interaction effect of POS and job crafting.

Simple slope analysis was performed to determine at which levels of the moderator (i.e. job crafting) the interaction occurs. The results showed that POS is related to individual work performance, only among teachers ‘low’ (θ_{X→Y|W = −0.44} = 0.20, p < 0.001, CI = 0.14, 0.27) and ‘moderate’ (θ_{X→Y|W = 0.09} = 0.10, p < 0.001, CI = 0.06, 0.15) in job crafting. Among teachers ‘high’ in job crafting, there is no statistically significant association between POS and individual work performance (θ_{X→Y|W = 0.61} = 0.00, p = 0.93, CI = −0.06, 0.07). Hence, POS matters less for individual work performance when job crafting is high. Figure 1 graphically displays the interaction effect.

FIGURE 1: The effect of job crafting on the relationship between perceived organisational support and individual work performance.

The J–N technique results are displayed in Figure 2. Results suggest that when job crafting is outside the interval [0.36, 1.12], the slope of POS is significant (p < 0.05). In other words, if 0.36 ≤ job crafting ≥ 1.12² (i.e. below the 70th but above the 99.5th percentiles of the distribution of job crafting), then the conditional effect of POS on individual work performance is statistically significant. In between these values (i.e. between the 70th and the 99.5th percentiles of the distribution of job crafting), POS’s conditional effect on individual work performance is not statistically significant. Therefore, the hypothesis is accepted.

FIGURE 2: The Johnson–Neyman plot.

Discussion

Organisations could benefit from knowing how they can manage teachers’ performance. Likewise, individuals can benefit from learning how to enhance their performance without sufficient external resources, which is often the case in developing countries. The current study’s objective was to determine if perceived support from the organisation enables teachers to perform and whether teachers’ job crafting behaviours enhance this relationship.

The results of this study support the hypothesis that job crafting moderates the relationship between POS and individual work performance. This means that POS initiates a motivational process that enhances teacher functioning but that this process depends on teachers’ level of personal resources (i.e. their job crafting behaviour). The current study supports the interactive (and amplifying) effect of ‘top-down’ (i.e. POS) and ‘bottom-up’ (i.e. job crafting) approaches on performance proposed in the literature (Bakker, 2015; Van Wingerden & Van Der Vaart, 2019). Furthermore, this study supports the flexibility and cross-context applicability of the JD–R model and how different job resources and personal resources may interact to produce a certain organisational outcome. The moderating effect of job crafting also corresponds with previous empirical studies (Debus et al., 2020; Hakanen et al., 2017; Van Hooff & Van Hooft, 2014), in which job crafting moderated the effect of focal predictors. The findings of the current study are encouraging in that it also shows that organisational and personal resources help teachers achieve their predetermined job tasks (i.e. task performance), adapt their performance to changing demands and act as change provocateurs by proactively performing tasks. The rapidly evolving (hybrid) school environment can benefit from teachers’ adaptive and proactive performance behaviours (Collie et al., 2020), as such an environment is characterised by uncertainty. In uncertain or unpredictable environments, adaptability and proactivity matter more than proficiency (Gagné et al., 2022; Griffin et al., 2007), and the current study illustrates how this can be achieved.

In the current study, it is worth noting that POS exerts influence at low and moderate levels of job crafting. More specifically, the J–N plots reveal that the effect of POS on individual work performance decreases as teachers’ level of job crafting increases, and at a certain threshold of job crafting, the positive impact of POS on individual work performance becomes nonsignificant. The more prominent effect of job crafting behaviours is not surprising and may be explained by distinguishing between proximal and distal processes, similar to Kanfer’s (1990) work with motivational processes. According to this differentiation, distal factors are those factors that have a more distant influence on behaviour. In contrast, proximal factors are those factors that sustain an action while engaging in the task (Kanfer, 1990). Hence, the distal factor (i.e. POS) sets the stage for the behaviour (i.e. performance), while the proximal factor (i.e. job crafting) sustains the behaviour. Job resources are also positioned as activators or enablers of personal resources (Sepeng, Stander, Van Der Vaart, & Coxen, 2021; Van Veldhoven et al., 2020). Peral and Geldenhuys (2020) also confirmed strong relationships between personal characteristics and performance. The current study extends the literature by illustrating that personal resources may matter more for teacher performance than organisational resources and may even override the effect of the latter. This is an important finding for resource-constrained (developing) countries, such as South Africa, and opens new avenues for intervention research to demonstrate the value-add of personal resources in these contexts.

In achieving its aim, the study contributed to the limited literature on teachers’ performance in the South African context by presenting the contributory factors (i.e. POS and job crafting). It also provides a nuanced understanding of the interaction between ‘top-down’ (i.e. POS) and ‘bottom-up’ (i.e. job crafting) approaches when enhancing (multidimensional) performance. The results indicate that it is a delicate dance between providing just enough POS while encouraging some initiative to alter jobs. The study also contributed to the limited job crafting literature in the South African context by presenting the important role it fulfils for teachers’ proficiency, adaptivity and proactivity, perhaps even more so than organisational support.

Practical implications

Despite the nonsignificant association between POS and individual work performance (as illustrated by the nonzero order correlations), this study’s results confirm the role of POS in enhancing teachers’ performance as it interacts with job crafting, and it matters for those who do not engage in much job crafting. Consequently, educational organisations are encouraged to design purposeful intervention programmes that build POS. Organisations could start with baseline assessments to determine the levels of POS within their organisation, followed by implementing policies that will improve POS among the teachers. These could include creating favourable working conditions by implementing discretionary support services, developing social networks, ensuring organisational fairness and providing effective leadership (Eisenberger, Malone, & Presson, 2016).

Perhaps even more important, educational organisations should design purposeful intervention programmes that teach employees to craft their jobs. Job crafting interventions such as the Michigan Job Crafting Exercise have been shown to encourage job crafting behaviour among teachers up to one year after implementation. Such interventions start with informing teachers about job crafting, its benefits and how to craft their teaching tasks. The training also includes identifying barriers to and resources for crafting (Van Wingerden, Bakker, & Derks, 2017). Based on the job crafting model by Wrzesniewski and Dutton (2001), interventions should focus on all three dimensions. For example, task crafting interventions could include offering free classes to disadvantaged communities (Coleman, 2017) or taking part in other creative and challenging teaching projects (Huang et al., 2022). Their task crafting would likely entail more approach rather than avoidance task crafting (Bruning & Campion, 2018; Zhang & Parker, 2019), as teachers are often subjected to more formalised task agreements which they cannot avoid, evade or minimise. Teachers could build and strengthen relationships with colleagues by volunteering to coach or mentor recruits or recently qualified teachers, offering enrichment workshops to support teachers who might benefit from such activities, and organising social events. Keeping a daily journal of the positive contributions and reflecting on the reasons for being a teacher (e.g. developing others) is helpful to reframe the meaning of one’s work into something more positive (Coleman, 2017). All practical training sessions should be followed up with reflection or evaluation sessions. On a managerial level, managers should also be encouraged to retain employees who willingly partake in job crafting behaviours (Van Wingerden et al., 2017). On an organisational level, policymakers should consider how they can enable job crafting behaviours. Here, a supportive climate (e.g. openness and collaboration) and an enabling job design (e.g. autonomy, discretion, and flexibility) could be helpful (Lazazzara, Tims, & De Gennaro, 2020).

Limitations and recommendations

The study has a number of methodological and theoretical limitations. Firstly, a cross-sectional design was used. Although the hypothesis was grounded in theory, no causal relationship could be established, and a reverse relationship cannot be excluded. For example, should a person receive a high performance review, they may perceive the organisation as supportive and caring. Cross-sectional designs are also infamous for common method bias, especially when self-report surveys are used (Spector, Rosen, Richardson, Williams, & Johnson, 2019). This leads to the second methodological limitation. All three measuring instruments were self-reported. This could be problematic in performance’s case, considering the Dunning–Kruger effect (Kruger & Dunning, 1999), where certain people may likely score their abilities higher than their actual performance level. Although POS and job crafting are best reflected in one’s own perceptions thereof, more objective measuring of performance would be preferred. This could include ‘hard’ organisational performance data (e.g. student retention and throughput), which is more objective than either self or other reports. Alternatively, one could control for self-deception enhancement, as recommended by Spector et al. (2019). Thirdly, research has also found that POS levels could fluctuate within short periods, depending on work experiences (Caesens et al., 2016). This is not captured by cross-sectional or the usual longitudinal designs. Intensive longitudinal designs (e.g. diary studies or experience sampling methods) allow dynamic changes in relationships between variables to be explored (Nezlek, 2020). The use of such designs is recommended for future research for ‘real-life’ experiences.

Lastly, the timing of the study could have contributed to different work experiences. The data were collected during the COVID-19 pandemic, when most institutions were compelled to find new ways of operating. Although the ‘new (future) normal’ may be more reflective of the current pandemic situation than previously thought, one cannot ignore the additional stress placed on teachers suddenly switching to online remote teaching. Teachers might have perceived support from the organisation differently than if they had been working from the office. Job crafting behaviour could also have played out differently. For example, teachers might have had to engage in various job crafting behaviours to navigate reduced organisational resources while working remotely. Unclear job role expectations, poor change communication and job autonomy encourage job crafting behaviours (Petrou, Demerouti, & Schaufeli, 2018), which may have been evident during this period. A comparative study after the pandemic may be valuable to confirm the current research results and evaluate the current study findings’ transferability.

From a theoretical perspective, numerous other organisational and individual factors also influence job crafting behaviours, including job crafting motives, contextual variables (such as work climate and job design) and personal factors (such as personality traits) (Lazazzara et al., 2020). This applies to performance as well, where proximal capacity (knowledge and skills) and distal capacity (such as ability) influence performance (Griffin et al., 2007). These factors were not measured in this study and could have influenced performance during the COVID-19 pandemic. For example, a person who does not have the technological skills or ability to deliver the required support to students could perform more poorly. It is recommended that future studies include these variables to gain a more holistic picture of teachers’ performance. In the current study, one teacher experienced POS as detrimental to their individual work performance because of their high levels of job crafting (≥ 1.12). Although this is a negligible finding in the current study, it is the opposite of what one would expect, and the finding may be more prevalent in other contexts. Future research should monitor whether POS could be a hindrance (rather than a resource) for employees with high levels of personal resources, as some researchers (Biron & Van Veldhoven, 2016; Deelstra et al., 2003; Van Veldhoven et al., 2020) have previously questioned the universality of job resources.

Conclusion

The current study aimed to determine if a relationship exists between the contextual factor (i.e. POS) and individual work performance and whether job crafting plays a role in moderating this relationship among a sample of teachers in South Africa. The results indicated that job crafting moderates the relationship between POS and performance, especially for those with low to moderate levels of job crafting. Organisations may use the findings to develop interventions to create supportive organisational contexts and encourage job crafting behaviour among teachers.

Acknowledgements

The authors would like to acknowledge the people who participated in the study.

The authors declare that they have no financial or personal relationships that may have inappropriately influenced them in writing this article.

A.D.T. acted as the primary researcher, as this study formed part of her master’s research. She conceptualised the article, collected the data, interpreted the research results and shared in the writing of the article. K.R. and L.V.D.V. acted as the supervisor and co-supervisor, respectively. They played an advisory role, assisting in the conceptualisation of the study, interpretation of the research results and refining the research article. K.R. also assisted with data collection, and L.V.D.V. analysed the data.

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Data sharing will be done in line with ethics requirements. The data is the intellectual property of the North-West University and will be provided upon reasonable requests to the corresponding author.

The views and opinions expressed in this article are those of the authors and do not necessarily reflect the official policy or position of any affiliated agency of the authors.

References

Footnotes

1. In the current study’s context, such support would be from the schools’ managing boards.

2. It is important to note that less than 0.5% of the data are in the upper range (i.e. ≥ 1.12) of the moderator. In a sample size of 207, that is one case.