Employee engagement is a positive and fulfilling, work-related state of mind whereby employees feel motivated and enthusiastic, and are actively involved in their work (Albrecht, 2010). Beyond energy and active involvement, engagement has also been defined and measured with “absorption” included as a third dimension of engagement (Schaufeli et al., 2002). Absorption refers to a flow-like such that employees feel fully concentrated on, and happily engrossed in, their work. However, engagement researchers have questioned whether absorption is a core aspect of work engagement or an outcome of energy and involvement (Bakker et al., 2011). Meta-analytic research has clearly shown employee engagement to be positively associated with important employee outcomes, such as task performance, contextual performance, proactive work behavior, innovative work behavior, and reduced turnover intentions (e.g., Crawford et al., 2010; Borst et al., 2020; Kwon and Kim, 2020; Neuber et al., 2022; Woznyj et al., 2022).

Much of the academic research focused on identifying the factors that predict engagement has been based in engagement theory (Kahn, 1990), Job Characteristics theory (Hackman and Oldham, 1980), and Job-Demands Resources theory (JD-R; Bakker and Demerouti, 2017). Kahn, for example, proposed psychologically meaningful work, psychological safety, and the availability of physical, emotional, and psychological resources as key enabling conditions for personal engagement at work. May et al. (2004) provided empirical support for Kahn’s propositions. Along similar lines, research has established that personal resources, such as meaningful work, optimism, self-efficacy, and psychological capital also have a direct influence on work engagement (Alessandri et al., 2018; Borst et al., 2020; Mazzetti and Schaufeli, 2022). Psychological conditions and personal resources differ from personality traits in that they are more context specific, labile, and amenable to change (Luthans et al., 2007.)

Beyond the well-established construct of work engagement, researchers have also argued in support of more “domain-specific” engagement constructs, such as team engagement, engaging leadership, safety engagement, pro-environmental engagement, and change engagement (e.g., Richardson and West, 2010; Schaufeli, 2015; Albrecht et al., 2022a,b). Change engagement has been defined as “an enduring and positive work-related psychological state characterized by a genuine enthusiasm and willingness to support, adopt and promote organizational change” (Albrecht et al., 2020, p. 4). Change engagement consists of two sub-dimensions: change energy and active involvement. The “energy” sub-dimension reflects employee enthusiasm for change, and the “active involvement” sub-dimension reflects an active striving toward the achievement of successful change. The two sub-dimensions are direct analogues of vigor and dedication, the primary sub-dimensions of engagement (Bakker et al., 2011). Change engagement therefore provides a more theoretically grounded and motivational expression of previously researched positive change-related constructs, such as openness to change, readiness for change, and commitment to change (Bouckenooghe et al., 2009; Arnéguy et al., 2018; Tsaousis and Vakola, 2018). Although researchers have shown that change-related organizational and job resources are associated with change engagement (Albrecht et al., 2022a), no research has yet examined the influence of a theoretically coherent set of change-related psychological resources on change engagement.

Change-related psychological conditions, or personal resources, refer to “enduring psychological states, or mindsets, that shape an individual’s ability to successfully adapt to a changing work environment” (Albrecht et al., 2020, p. 11). Change-related personal resources include constructs, such as change-related self-efficacy (e.g., Jimmieson et al., 2004), change-related resilience (e.g., Braun et al., 2017), change-related meaning-making (van den Heuvel et al., 2009), and change-related psychological safety (Singh et al., 2018; Albrecht et al., 2020). As domain-specific analogues of psychological conditions for engagement of Kahn (1990), it is here argued that change meaningfulness, change-related psychological safety, and change self-efficacy are core change-related personal resources that are prerequisite conditions for change engagement. Although researchers have suggested that consideration of change-related personal resources is essential to understanding employee attitudes to change (van den Heuvel et al., 2009), researchers have yet to examine the associations between theory-based change-related psychological resources and change engagement, nor their direct or indirect associations with proactive work behavior. Each of these constructs and their proposed relationships are shown in Figure 1 and overviewed below.

Griffin et al. (2007) distinguished between three forms of work performance: proficiency (task performance), adaptivity (responses to accommodate change), and proactivity (self-initiated change). Proactive performance manifests as change-oriented behaviors that reflect an employee’s willingness to initiate and adopt new and innovative ways of thinking and behaving that can lead to organizational performance and competitive advantage (Straatmann et al., 2016; Straatman et al., 2018). Griffin et al. (2007) measured proactive behavior with items assessing how often respondents have “come up with ideas to improve the way in which your core tasks are done” and “made suggestions to improve the overall effectiveness of the organization (e.g., by suggesting changes to administrative procedures).” Given that researchers have argued and demonstrated that employee engagement leads to innovative and proactive work behaviors (Hakanen et al., 2008; Salanova and Schaufeli, 2008), it is here proposed that change engagement will be positively associated with proactive work behavior (H7). Furthermore, and as previously noted as H2, H4, and H6, theory and research suggest that the influence of psychological conditions, such as meaningfulness, psychological safety, and self-efficacy will influence outcomes such as proactive behavior through their influence on engagement (Allan et al., 2019; Kwon and Kim, 2020; van den Heuvel et al., 2020).

Overall, the present study aimed to address gaps in the literature by extending well established engagement theories to the context of organizational change. The research tests a model proposing positive associations between change-related psychological resources, positive attitudes to change, and employee performance. More specifically, it is proposed that positive employee experiences of change-related meaningful work, psychological safety, and self-efficacy, will be directly associated with employee change engagement, and indirectly associated with self-reported proactive work behavior (see Figure 1).

The sample consisted of 301 paid participants sourced from Prolific. Research has shown that Prolific data have “similar psychometric properties and produces criterion validities that generally fall within the credibility intervals of existing meta-analytic results from conventionally sourced data” (Walter et al., 2019, p. 425). Beyond having a good record of responding to survey questions, the participants needed to be 18 years or older, and working full- or part-time for a minimum of 3 months in an organization of 15 or more employees that had experienced organizational change within the past year. Prior to responding, participants were informed about the purpose of the study, the confidentiality of their responses, and that the study had been granted ethics approval. Participants were also informed before commencing the survey that they needed to attend carefully to, and complete, all survey questions to receive payment. Of the 297 participants who satisfied the inclusion criteria, 49.2% reported as female and 49.8% as male. Ages ranged from 18 to 66 years, with a mean of 36.5 years. Tenure ranged from 6 months to 40 years, with a mean of 6 years. Respondents worked in organizations of varying size (10–1 million plus employees) as team members (53%), senior managers (5%), managers (17.5%), team leaders (17.8%), and executives (2%). The clear majority of participants (79%) reported experiencing a moderate to a great deal of organizational change during the previous 12 months (e.g., new structures, restructures, mergers, new technologies, new systems, new policies, relocations, changes to work schedules, and leadership).

Online calculator of Soper (2022) showed that the sample size used in the analyses (N = 297) exceeded the minimum sample size (N = 200) needed to test the proposed model at power and to detect a medium effect size. Previous studies exploring the influence of psychological conditions on engagement and change have reported medium to large effect sizes (e.g., May et al., 2004; van den Heuvel and Demerouti, 2009).

The six first order constructs shown in Figure 1 were measured on a seven-point Likert scale (1 = strongly disagree to 7 = strongly agree). Items included in each of the scales and their factor loadings are shown in Table 1. Scale means, standard deviations, alpha reliabilities, and correlations are shown in Table 2.

The two-step modeling approach of Anderson and Gerbing (1988) was applied to test the theoretically based model. As a first step, Confirmatory Factor Analysis (CFA) was used to assess the fit of the measurement model. Given the cross-sectional correlational survey design, the CFA analytic strategy also included assessing the influence of common method bias (Podsakoff et al., 2012). As a second step, structural equation modeling (SEM) was conducted to test the proposed relationships shown in Figure 1. CFA and SEM analyses were conducted using conventional or covariance based structural equation (CB-SEM) with AMOS v26 and Maximum Likelihood estimation.

Both the CFA and SEM models were assessed using a number of recommended fit indices: the chi-square statistic (non-significant), chi-square to degrees of freedom ratio (< 2 or 3; Jaccard et al., 1996), Tucker Lewis Index (TLI > 0.95), Comparative Fit Index (CFI > 0.95), the root mean square error of approximation (RMSEA < 0.05; Hu and Bentler, 1999), and the standardized root mean residual (SRMR < 0.06; Kline, 2016). Beyond assessing the overall fit and the direct effects within the proposed model, bootstrapping procedures were used to assess the indirect effect of the three change-related psychological resources on proactive work behavior via change engagement.

As shown in Table 1, the proposed measurement model yielded acceptable fit. The chi-square ratio, TLI, CFI, and SRMR were at, or better than, criterion values. Although the RMSEA estimate was slightly higher than the recommended criterion, a model can be accepted with confidence when a range of different fit indices are at criterion (Hu and Bentler, 1999; Heene et al., 2011). Table 1 also shows that all fit indices were superior to the null model and a one factor model, calculated for comparison purposes. Additionally, and in support of convergent validity, the standardized loadings for all items were strong and statistically significant (p > 0.001), ranging from 0.846 to 0.958 (see Table 2).

Table 3 shows the means, standard deviations, Cronbach’s α, average variance extracted, and correlations for all first-order variables included in the measurement model. The α reliabilities exceeded the accepted criterion value of 0.80 (Nunnally, 1978). In support of the convergent validity of the scales, all six AVE values (ranging from 0.77 to 0.86) clearly exceeded the accepted criterion value of 0.5 (Fornell and Larcker, 1981). Furthermore, as shown in Table 3, the bivariate correlations were all significant and positive. Given the relatively strong correlation between change meaningfulness and the energy sub-scale of change engagement (r = 0.82), the discriminant validity for these two constructs was tested. That is, the chi-square values were compared when the correlation between the two constructs was first fixed at one, as opposed to when it was freely estimated (Anderson and Gerbing, 1988). The sizable and significant difference between the chi-square values, given one degree of freedom, supported the discriminant validity of the two constructs (Δχ² = 203.7; p < 0.001). More broadly, and also in support of discriminant validity, the square root of each construct’s AVE (ranging from 0.88 to 0.93) exceeded all corresponding construct correlations (Fornell and Larcker, 1981).

Given the cross-sectional and self-report nature of the data, tests for common method variance (CMV) were conducted by comparing CFA standardized loadings before and after the addition of a latent common method factor (Podsakoff et al., 2012). Relative to the proposed measurement model, the standardized loadings for seven of the 18 items decreased more than 0.20 (range 0.23–0.45) after including the common method factor in the model. The seven items, not surprisingly, included five change engagement items that formed part of the higher order change engagement factor (see Figure 1). The remaining two items were from the change meaningfulness scale, and therefore suggest some potential redundancy between the items. Overall, the standardized loadings decreased by a relatively modest 0.22 on average, and all factor loadings remained statistically significant (p < 0.001) after the inclusion of the common method factor. Therefore, although acknowledging the presence of CMV, the method effects appear to not be overly influential (Johnson et al., 2011; Podsakoff et al., 2012). Furthermore, and in support of modeling change-engagement as a higher order construct, the sub-scale loadings on the higher-order factor clearly exceeded the recommended value of 0.70 (Credé and Harms, 2015).

Having established the validity of the measurement model, step 2 of the broad analytic strategy (Anderson and Gerbing, 1988) involved using SEM to test the fit of the proposed structural model (see Figure 1). The structural model yielded good fit (see Table 1). In support of H1, H3, and H5, Figure 2 shows that as proposed, each of the three psychological conditions (change-related meaning, change-related psychological safety, and change-related self-efficacy) had a significant, positive, and direct effect on change-engagement. In support of H7, change-engagement had a significant, positive, and direct effect on proactive work behavior. The model explained 88.1% of the variance in change engagement and 15% of the variance in proactive work behavior.

To test the proposed mediating influence of change engagement on the relationship between the three change-related psychological resources and proactive work behavior, bias corrected bootstrapping procedures were conducted. In support of H2, H4, and H6, using 2,000 bootstrap samples, significant and positive indirect effects were found for each of the change-related psychological resources: change-related meaning (β = 0.224, p < 0.001, CI 95% = 0.146–0.312), change-related psychological safety (β = 0.096, p < 0.001, CI 95% = 0.051–0.164), and change-related self-efficacy (β = 0.115, p < 0.01, CI 95% = 0.063–0.195). As a final step in the analyses, an alternative partial mediation model was tested whereby all three psychological conditions were also specified to be directly associated with proactive work behavior. Although the partial mediation model provided equally good fit to the data (see Table 1), none of the three additional parameters were significant (p = 0.088, 0.272, 0.592). Therefore, the proposed fully mediated model was accepted as the most parsimonious and theoretically defensible model.

The findings add to the body of research supporting change engagement as a potentially useful resource that organizations can use to succeed in an environment of ongoing change. It has been well established that if employees are ready, open, and committed to change, the chance of change initiatives being successfully implemented is increased (Choi, 2011). Given the very significant investment of financial, technical, and people resources that organizational change demands, it is incumbent on organizations to optimize the return on investments through the successful realization of change. The results of the current study suggest that organizations should build interventions into change programs that are focused on developing employee change meaningfulness, change self-efficacy, and change-related psychological safety. Facilitating employee experiences of change meaningfulness, change self-efficacy, and change-related psychological safety, and particularly meaningful change, will increase the likelihood that employees will be enthusiastic about change, and actively involved in change.

Meta-analytic evidence has shown that both top-down and bottom-up, resource-developing interventions are effective in developing work engagement (Björk et al., 2021). Similar intervention strategies are likely to be effective in developing change-engagement. As such, training, coaching, job crafting, and team-based and organizational learning development interventions that have previously been used to help employees develop work engagement, self-efficacy, work meaningfulness, and psychological safety could usefully be adapted to the context of change. For example, job crafting interventions have been linked to increased employee meaningfulness of work, increased work engagement, and increased PsyCap (Sakuraya et al., 2016; Bruning and Campion, 2019). Bottom-up “change crafting” interventions, paralleling job crafting intervention designs, might involve participants identifying their own realistic and meaningful goals, breaking down key goals into manageable sub-goals, identifying and evaluating multiple pathways to achieve goals and sub-goals, identifying resources needed to achieve the goals, and the sharing of participant experiences (Luthans et al., 2006). Such intervention designs might be applied to help employees enhance change-related job resources, such as change information, change involvement, supervisor support for change, and co-worker support for change (Albrecht et al., 2020; Björk et al., 2021).

In organizational development terms, the research provides brief and reliable measures of change related personal resources and employee change engagement that can be used as “pulse-check” diagnostics of employee attitudes to change (Jolton and Klein, 2020). That is, in addition to assessing the influence of change-related job resources on employee enthusiasm for and involvement in change, survey diagnostics can be administered to assess the extent to which employees are confidently experiencing on-going organizational change as meaningful and psychologically safe. Organizations could then share survey results with employees and involve employees in developing action plans to build on strengths and address threats to employee wellbeing and performance throughout organizational change initiatives.

Despite the contributions of the study, several limitations should be noted. Firstly, the use of the prolific sample may suggest the results may not generalize to the wider world of work. However, as previously noted, researchers have shown that Prolific samples yield results with similar psychometrics and validities to conventionally sourced data (Walter et al., 2019). Secondly, because the data were single-source, and self-report, common method variance may have impacted the validity of the model (Podsakoff et al., 2012). However, as previously noted in the results, tests for common method variance showed that after the inclusion of a common method factor, the CFA standardized loadings decreased by a relatively modest amount, and all factor loadings remained statistically significant. Thirdly, because the data were collected at one time point, no strong claims for causal relations between the constructs can be made. Further research using longitudinal designs would be required to confidently establish the direct and indirect effects reported. Further research could also be conducted to integrate the current research findings within previously proposed research models that have examined the influence of organizational and change-related job resources on change engagement (Albrecht et al., 2022a). Beyond the influence of the three psychological conditions examined in the current study, future research could also examine the mediating influence of additional change-related personal resources such as change-related hope, optimism, and resilience on change engagement and proactive behavior (Youssef and Luthans, 2007). Importantly, further intervention research is needed to establish the effectiveness of top-down and bottom-up interventions such as change crafting as a means to enhance employee change engagement.

The study aimed to assess the validity of a domain-specific analogue of engagement theory (Kahn, 1990). Consistent with the theory, the results showed that change-related meaningful work, psychological safety, and self-efficacy potentially serve as important psychological preconditions for change engagement and proactive work behavior. Therefore, organizations can expect to significantly improve employee motivation to change and to innovate when they create the conditions whereby employees experience organizational change to have purpose, and whereby employees feel competent, confident, and psychologically safe. Where employees are enthusiastic and motivated about change, there will therefore be an increased likelihood of the successful implementation of organizational change, and therefore improved organizational competitiveness and sustainability.