**4. Results**

## *4.1. Measurement Model*

Table 1 shows the means and standard deviations (SDs) of each scale. The beliefs (BEL) construct had the highest mean (mean 5.147, SD 1.151), while the project characteristics (PC) construct had the lowest mean (mean 3.513, SD 1.379). These results indicate that respondents were generally positive about their change-related beliefs, organizational characteristics, and institutionalization processes, but less so about the project characteristics. To determine construct reliability and validity, indicator loadings were evaluated, with loadings greater than 0.7 (Götz et al. 2010) considered acceptable, as items with loadings of 0.6 or greater are considered acceptable (Hulland 1999) in cases where new scales are developed. After assessing their potential practical significance, seven items from the beliefs construct (Ps2, Ps3, Ps4, Ps5, Ps6, Va4, Ef3), three items from the organizational characteristics (Oc3, Oc8, Oc10), four items from the project characteristics (Pc7, Pc8, Pc9, Pc10), and two items from the institutionalization process (Ip6, IP7) were removed due to low loadings (Appendix A Table A1 shows all the construct items that were retained). The composite reliability (CR) statistics for all constructs in the model ranged from 0.958 for change-related beliefs to 0.868 for project characteristics, which were significantly higher than the recommended 0.70 (Hair et al. 2019). The internal consistency of each construct, as measured by Cronbach's alpha (α), was greater than the recommended minimum of 0.7 (Hair et al. 2019).

We used the average variance extracted (AVE) to determine the convergent validity of the measured constructs. The AVE measures the amount of variance that a construct obtains from its indicators relative to the amount due to measurement errors (Fornell and Larcker 1981). Table 1 shows that the AVEs for all the study constructs were higher than the recommended 0.5 (Hair et al. 2019). Discriminant validity was assessed using the method of Fornell and Larcker (1981) by computing the square roots of the construct AVEs and checking whether the square root was larger than the correlation between constructs (Table 2). We also used the heterotrait–monotrait (HTMT) approach, where HTMT values lower than 0.85 for all constructs establish discriminant validity (Henseler et al. 2015).

The results in Tables 1 and 2 show that the measurement model presents satisfactory indicator reliability, convergent validity, and discriminant validity, hence demonstrating satisfactory robustness needed to test the relationship between the constructs.

#### *4.2. Structural Model*

The structural model, or inner model, shows the associations between the constructs being studied. Table 3 summarizes the results for the hypothesis tests by presenting the path coefficients (β), R2-adjusted values, t-values, and model cross-validated redundancy (Q2) value. Variance inflation factors (VIFs) between the variables were first computed to determine any possible multicollinearity, and the results indicate that the VIFs were all less than the threshold of 3.0 (Hair et al. 2019). Our first prediction was that organization characteristics (OC) have a significant effect on the institutionalization process of project outcomes and the project characteristics. The results reveal a significant positive effect (β = 0.502, *p* < 0.001). However, the relationship between organizational characteristics and project characteristics was non-significant, and hence hypothesis H1a was supported, while H1b was rejected. The results also confirm our hypothesis (H2) that the characteristics of the project have a negative effect on the institutionalization process (β = −0.156, *p* < 0.05). Similarly, the findings confirm our hypothesis (H3a) that stakeholder change-related beliefs would have a positive significant effect on the institutionalization process for project outcomes (β = 0.352, *p* < 0.000). The overall institutionalization process model was significant with an adjusted R2 = 0.651 (p < 0.000). The cross-validated redundancy (Q2) value tests the predictive relevance of the model and shows how well the path model can predict the observed values. The computed Q<sup>2</sup> = 0.375 for the institutionalization process model is greater than zero, showing the predictive relevance of the model (Sarstedt et al. 2017).

**Table 3.** Summary of the structural model.


\*\* *p* < 0.01; \*\*\* *p* < 0.000; ns—non-significant.

We also predicted that the relationships between organizational characteristics (OC) and project characteristics (PC) and the institutionalization process (IP) would be mediated by the internal stakeholder's change-related beliefs. Mediation effects exist when a third variable plays an intermediate role in the relationship between the independent and dependent variables (Carrión et al. 2017; Sarstedt et al. 2020). According to Carrión et al. (2017) and Zhao et al. (2010), the only requirement to establish mediation between two variables is to test whether the indirect effect between the variables is significant even when the direct effect is not significant. Table 4 shows the mediation effects.



\* *p* < 0.05; \*\* *p* < 0.01; \*\*\* *p* < 0.000.

The results reveal that the total effects of both organizational characteristics and project characteristics on institutionalization processes are significant (i.e., β = 0.720, *p* < 0.000, and β = −0.255, *p* < 0.000, respectively). These observed effects persist when the change-related beliefs variable is included as a mediator, and the specific indirect effects are significant. This indicates partial mediation of change-related beliefs, thereby partially confirming hypotheses H3b and H3c. Figure 1 below summarizes the structural model indicating the relationships between the dependent and independent variables of this study.

**Figure 1.** Structural model exhibiting the relationship between study constructs. \* *p* < 0.05; \*\* *p* < 0.01; \*\*\* *p* < 0.000; ns—non-significant.

## **5. Discussion**

The purpose of this study was to investigate whether micro-level project internal stakeholder change-related beliefs and macro-level organizational and project characteristics influence the institutionalization process of outcomes of a development cooperation project in an organization. Drawing insights from the organizational development, change management, and development project management literature, we tested a mediation model using survey data from a sample of 130 respondents drawn from a university implementing long-term development cooperation projects in Uganda. The study findings reveal that both organizational characteristics and internal stakeholder change-related beliefs have a positive effect on the institutionalization process, while the project characteristics have a negative effect. Additionally, internal stakeholder change-related beliefs partially mediate the relationships between organizational and project characteristics and the institutionalization process.

While a few previous studies (Buchanan et al. 2005; Cummings and Worley 2009; Jacobs 2002) developed theoretical frameworks for organizational-level change institutionalization, the current study presented an empirical examination of the institutionalization model. Consistent with our findings, Murrah-Hanson and Sandmann (2021) argued that the institutionalization process involves changes occurring both at the individual and organizational levels. At the individual level, the results suggest that institutionalization actions are likely to be enhanced as long as the internal stakeholders have positive beliefs about the project. Therefore, the more stakeholders perceive the need for change (discrepancy) and believe the project interventions match the identified organizational needs (appropriateness), the more they will explicitly commit to the project and engage in actions to integrate project

outcomes into the organization. Additionally, discrepancy beliefs are likely to reduce arbitrary perceptions of the project objectives and outcomes, among the stakeholders, and thus help legitimize the need for project-initiated changes, while perceived positive valence and self-efficacy beliefs towards the projects are expected to trigger actions that will perpetuate the project outcomes (Armenakis et al. 2007). As Murrah-Hanson and Sandmann (2021) argued, people within an organization possess beliefs and mindsets to accept and support a change intervention, which could be translated into actions to sustain the change. This means that project initiation, planning, and implementation practices must account for the non-linearity of these behavioral elements and incorporate strategies into the project cycle to nurture and constantly reinforce the identified change-related beliefs among project stakeholders. Stakeholder commitment, for example, is less linear (Murrah-Hanson and Sandmann 2021) and thus contradicts the time-constrained, mechanistic, linear outcome chain techniques that dominate development project practice (McEvoy et al. 2016). In such a case, a more iterative approach to project implementation (Baptista et al. 2019) is more appropriate to allow the targeted project internal stakeholders to interact with the project interventions and acquire the necessary beliefs, attitudes, and behaviors to sustain the changes.

Internal stakeholder views, attitudes, and behaviors regarding projects develop within the context of the organization. The organization's characteristics provide a platform for individuals to interact, develop, and express their experiences and meanings about the project. The more deeply and widely internal stakeholders exchange their beliefs and interpretations of project outcomes, the more firmly the outcomes are embedded in the organizational social system, preventing the change from decaying (Alshehhi 2014). Three illustrations follow to further explain this suggestion. First, an organization characterized by a culture that values transparency will more likely enact clear project implementation practices and processes that continuously update stakeholders about the progress of the project. As a result, the stakeholders will develop trust and look for opportunities to participate (Zakharova and Biedenbach 2013) in actions that perpetuate the project outcomes. Both a transparent, involvement-oriented organizational context (Rogiest et al. 2015) and trust in management (Smollan 2013) have been identified as critical elements in gaining commitment to organizational change. Though transparency and trust continue to be a concern in the implementation of international development cooperation projects (Agheneza 2009), more is required if the projects are to succeed and be sustained in organizations (Diallo and Thuillier 2005). Second, an organization with a well-defined philosophy, strategy, and structure will find it simpler to incorporate and sustain project interventions. To begin with, it is doubtful that such an organization will accept and implement project interventions that are not aligned with the organization, and it is not difficult to secure commitment from the organization's internal stakeholders under such conditions. Cummings and Worley (2009) and Goodman and Dean (1982) have both highlighted the importance of the relationship between intervention congruence and the persistence of change interventions. Finally, even if individuals are committed to the project change outcomes, if organizational systems do not change, the process of change institutionalization will not be realized (Murrah-Hanson and Sandmann 2021). This brings to the fore the central role of the organization's governance and leadership in supporting change institutionalization. Consistent with other studies (Grandien and Johansson 2012; Murrah-Hanson and Sandmann 2021; Pishdad and Haider 2013; Waiswa 2020; Yetano 2013), our study points to the support and championing of the project by the organization's senior management. Governance and leadership supportive actions in the form of providing the necessary resources while also creating a stable and conducive organizational environment can spur the institutionalization of project interventions. According to Alänge and Steiber (2009), the responsibility for project sustainability and change rests with the organization structure rather than individuals, who most often leave their assigned roles, yet the project continues. Similarly, the organization's governance can ensure that capable personnel are assigned to the project or change-related positions.

These findings have implications for aid-based organizational development projects' conceptualization, design, and implementation. First, no matter how carefully the project is designed, it will be difficult to implement and institutionalize if it is not clearly drawn from the organization's strategy. A well-defined organizational strategy can serve as a secure landing zone for project interventions. Consequently, internal stakeholders may easily incorporate project outcomes and related behaviors into strategic and operational planning for the organization. Second, there is need for the organization's leadership and the affected stakeholders to clearly understand the project being implemented within the organization. This calls for the expansion of the due diligence and project initiation processes to encompass a thorough analysis of both the formal organizational aspects, such as strategic plans and management structures, and the organizational environment that manifests the culture of the organization.

The standout negative relationship between project characteristics and institutionalization processes, though predicted, is a concerning result in relation to international development project practice. The negative perceptions towards the characteristics of the projects could be the result of the respondents' lack of clarity about the project objectives, which has a negative impact on commitment and integration actions. Further, the implementation of development projects is fraught with operational complexities; for example, in our case study, project activities are managed both the beneficiary organization in the South and the coordinating organization in the North. This level of operational complexity complicates local project ownership, coordination, and sponsorship. It not only jeopardizes effective monitoring, feedback, and corrective actions but also has a negative impact on commitment and the related structural actions required to institutionalize project outcomes and targeted behaviors. Furthermore, the multidimensional nature of the project, in this case touching on various aspects and changes in the organization, poses a challenge of integration in the organization's strategies and operations. These problematic characteristics of development projects have been articulated in many studies (Agheneza 2009; Gajic and Palcic 2019; Ika 2012; Ika and Hodgson 2014; Mishra 2016), and this study extends the analysis to how these same characteristics negatively affect the institutionalization of project outcomes in an organization. The results indicate a non-significant relationship between the project characteristics and organizational characteristics. This is an interesting result to note, because it may highlight the fact that a host organization can accept, without question, a project for the sake of receiving the donor funds, even if it does not fit the organization's context and characteristics or fully match the recipient organization's expectations. Hence, Picciotto (2020) advocates for the gradual adaptation of development projects to their operating context in order to meet project recipients' expectations, rather than sticking to the standard "conjectures embedded in project design" that dominate international development project design and implementation methods manifested by theories of action and change.

#### **6. Limitations and Future Research**

Since this study examined a single case of a university implementing development cooperation projects, the findings may not be generalizable. Several academics, however, have emphasized that change institutionalization is a contextual process influenced by underlying organizational factors (Buchanan et al. 2005; Martin et al. 2012; Self et al. 2007; Yetano 2013) that are socially constructed by organizational members (Clausen and Kragh 2019). Moreover, within the international development scholarship, context is a prominent factor as authors argue that one size does not fit all when it comes to project implementation (Ika 2012; Yamin and Sim 2016). An expanded study with a sample drawn from a variety of contexts could extend knowledge in the relatively less studied world of aid-based organizational development projects. Additionally, in the design of the study questionnaire, we identified items from the existing change institutionalization and development cooperation project literature, consultation with relevant experts in the field, and our own experience in implementing North–South development cooperation projects. More studies using larger samples could help confirm the three suggested categories of project outcome institutionalization actions: *explicit commitment-related, integration-related, and implicit structural-related actions*. Similarly, further studies on our suggested categorization of institutionalization actions could shed more light on how the project-targeted behaviors take root in an organization. Additionally, aid-based institutional development projects deliver outcomes at different levels: individual project level, organizational level, society level, or community level. There are no studies that examine the relationships between these outcomes. Using qualitative methods, it could be interesting to examine in depth questions such as: What factors and institutionalization actions within an organization translate individual-level project outcomes into organizational-level outcomes? How do institutionalized organizational-level project outcomes translate into community- or society-level outcomes?
