SEs develop highly innovative solutions to critical social and environmental challenges, thereby advancing the SDGs. The sustainability of SEs aligns with the SDGs through a core focus on social and environmental missions embedded in their business models and the use of market mechanisms to create sustainable solutions addressing poverty (SDG 1), education (SDG 4), health (SDG 3), gender equality (SDG 5), decent work (SDG 8), and environmental objectives (SDGs 12 and 13). Through innovative, community-focused, and inclusive practices, SEs act as catalysts for achieving the 2030 Agenda (GEM, 2024; United Nations, 2023; WEF, 2024). To assess the alignment of SEs with the attainment of the SDGs, Diaz-Sarachaga and Ariza-Montes (2022) draw on expert panels and a survey of 100 Spanish social entrepreneurship practitioners to develop a three-tier hierarchical framework comprising 28 indicators across 12 SDGs. The Mondragon Corporation, a leading Spanish cooperative with a global presence, is examined as a case study for this framework.

Prior research on sustainability in the SE context has been examined primarily from three perspectives: social, economic, and environmental (Kamaludin et al., 2024; Ngumbela, 2023; Rivera-Santos et al., 2015; Urban & Bukula, 2024). The social dimension of an SE must consider not only its outputs but also the sustainability of the entire process. This requires SEs to recognize and incorporate economic, social, and environmental considerations in the selection and use of external resources (Jayawardhana et al., 2022). Understanding SE sustainability and developing appropriate measurement frameworks are essential for assessing effectiveness and guiding future development. Research indicates that SEs generating social and environmental value should be grounded in sustainability principles (Audretsch et al., 2024; Lehtimäki et al., 2021), directing attention to the creation of grassroots innovation (Agrawal et al., 2024; Bosma et al., 2016; J. G. Dees et al., 2004; Doherty et al., 2014). SEs aligned with the SDGs are better equipped to address challenges and adapt to changing circumstances, thereby contributing to resolving issues such as climate change, poverty, and inequality (Gupta et al., 2020; Kamaludin et al., 2024). In a similar vein, the Global Entrepreneurship Monitor (GEM, 2024) report on sustainability and entrepreneurship finds that this motivation is evident across diverse cultures and geographies, with little indication that purpose-driven entrepreneurship is confined to higher-income economies. Low-income economies, such as India and Guatemala, are among those with the highest proportions of new and established entrepreneurs endorsing the sustainability-driven motivation “to make a difference in the world.”

Moreover, environmental issues can be understood as inherently social and economic in nature; consequently, social entrepreneurs are well positioned to drive impactful change in relation to climate change and the transition to a sustainable circular economy (Campos et al., 2020; Kamaludin et al., 2024). Although SEs are established for social purposes, it is important to assess their efficiency and profitability in order to evaluate their overall effectiveness, which ultimately determines their sustainability (Cosa & Urban, 2025). SEs rely on economic and financial performance measures to evaluate financial accountability and sustainability objectives (Bagnoli & Megali, 2011; Salwan et al., 2025). At the same time, although financial data play a key role in achieving sustainability outcomes, social effectiveness remains essential for meeting social needs in line with the social mission (Campos et al., 2020; Kamaludin et al., 2024). Most SEs are grounded in missions based on sustainability principles, striving to deliver social benefits in ways that maximize impact (Doherty et al., 2014). Social effectiveness, as a measure of SE performance, assesses the quality and quantity of impact achieved over time (Bagnoli & Megali, 2011), which is often intangible and therefore difficult to measure (Urban & Bukula, 2024). A further challenge in measuring sustainability is the lack of consensus on appropriate indices, which creates a significant barrier to the formulation and implementation of sustainability strategies (Lubberink, 2019; Urban & Bukula, 2024).

Through a systematic review of performance measurement systems and their relevance to SEs, Cosa and Urban (2025) provide insights into the systems that best capture social entrepreneurship hybridity. These include the General PMS Model for Social Enterprises, the Multidimensional Assessment Model, the Vincentian Marketing Orientation, and the SAVE framework, all of which offer, to varying degrees, credible performance measures that enable SEs to assess and report on their overall development and sustainability. Regardless of the performance measurement system adopted, social entrepreneurs must identify and pursue opportunities that generate positive impacts on people and the planet through innovative business models, products, and services (Lubberink, 2019).

A critical dimension of SE sustainability, given that SEs operate within communities, is their credibility and legitimacy, which enable them to gain community support and attract external stakeholders (Bagnoli & Megali, 2011). To achieve credibility and legitimacy, sustainable SEs may adopt strategies such as selecting “socially or environmentally certified suppliers which have ethical certification and provide employment to vulnerable groups and individuals with different types of disadvantages (mental health issues, physical handicap, disabled, etc.)” (Bagnoli & Megali, 2011; Bosma et al., 2016; Gupta et al., 2020). From this perspective, social entrepreneurs are responsible for ensuring that their SEs pursue strategies that promote long-term sustainability, environmental responsibility, social equity, and economic growth (Kamaludin et al., 2024). Achieving these long-term objectives requires the effective mobilization of resources to support the SE’s initiatives. This includes not only financial capital but also social networks and other forms of human and knowledge capital, as well as the cognitive capabilities that enable social entrepreneurs to make critical decisions within their SEs (Schaefer, Corner, et al., 2022).

This emphasis on cognitive capabilities aligns with the focus of this article on metacognition. While social entrepreneurs’ ability to adapt to changing venture environments is essential, their metacognitive processes enable them to make critical decisions within their SEs (Schaefer, Corner, et al., 2022; Tian et al., 2022).

Mair and Noboa (2006) were among the first to provide a more comprehensive understanding of the social entrepreneur by examining the antecedents of intentions to establish a social venture. Their model of social entrepreneurial intentions is based on the perceived desirability and feasibility of the enterprise. The relevance of these antecedents has been emphasized by subsequent research, particularly within social psychology, which focuses on understanding the mental processes (cognitions) that link attitudes and beliefs to action (Liñán et al., 2011). Understanding SEs from a cognitive perspective is important, given that social entrepreneurs act as change agents, and their ability to do so is grounded in their cognition (Corbett et al., 2018; Muñoz, 2018).

Various theories of human cognition and psychology have shed light on how and why entrepreneurs think differently from others (Baron, 2004; Mitchell et al., 2007; Urban, 2012). Social cognitive theory (SCT) posits that individuals’ cognitions shape and inform their behavior (Bandura, 2001), while entrepreneurial cognition refers to “the knowledge structures and simplified mental processes entrepreneurs use to make sense of previously unrelated material” (Baron, 2004; Mitchell et al., 2007). The entrepreneurial cognition literature has advanced considerably in examining cognitions related to entrepreneurial decision-making (Corbett et al., 2018; Mitchell et al., 2007; Muñoz, 2018). More recent developments in the field examine how entrepreneurs “think about thinking,” a process referred to as metacognition (Haynie et al., 2010). Metacognitive theory, introduced by psychologist John H. Flavell in the 1970s, centers on this concept. Metacognitive thinking involves self-awareness, planning, reflection, and the use of strategies based on prior knowledge to select appropriate actions aligned with environmental conditions and goals (Haynie et al., 2010; Urban, 2012).

Whereas cognition is required to complete a task, metacognition concerns awareness of one’s own cognitive activity; this awareness implies a form of monitoring that enhances individuals’ control over their cognitive processes (Norman et al., 2019). Individuals who effectively apply metacognition benefit from more accurate estimation and updating of their knowledge, as well as improved monitoring and evaluation of ongoing learning, because “metacognition triggers analytic reasoning to evaluate and refine intuitive reasoning” (Hadinejad et al., 2025). Social psychologists define metacognition as reflection on one’s thoughts, or thinking about thinking (Bandura, 2001). Similarly, in cognitive psychology, metacognition is described as an individual’s executive control, including monitoring and self-regulation (Massoni, 2014). In educational psychology, it is defined as higher-order thinking involving active control over cognitive processes and conscious awareness (Hanisch & Eirdosh, 2023; Thompson et al., 2011).

Although metacognitive approaches have gained significant attention in psychology, their application beyond this field—particularly in sustainability contexts—remains limited (Hadinejad et al., 2025). Nevertheless, interdisciplinary research integrating metacognition from psychological subfields into sustainability is evolving, with insights from environmental psychology demonstrating how metacognition can inform sustainability research. For example, Said et al. (2023) highlight the role of metacognition in shaping public responses to sustainability communication. Metacognitive strategies have also been identified as fundamental to research processes in sustainable tourism (Truong et al., 2024), while other studies apply metacognitive theory in marketing to examine sustainability-related behaviors (Vilkaite-Vaitone, 2024).

Recognizing that integrating metacognition into sustainability research can provide a more comprehensive framework—one that captures the interplay of cognitive processes, intentionality, and reflective awareness in shaping sustainable action (Hadinejad et al., 2025)—this study adopts the metacognitive awareness model (Haynie & Shepherd, 2009), which is central to the entrepreneurship cognition literature (Michaelis et al., 2021), to inform the development of the study hypotheses. Drawing on prior research integrating metacognition and sustainability, the model is used to examine how each of its dimensions may influence SE sustainability.

Kenya is the leading economic hub in East Africa and one of the continent’s main commercial centers (World Bank, 2024). Its capital, Nairobi, is home to over 5.7 million people and serves as the primary center for commerce, innovation, and enterprise in the region. With an estimated population exceeding 57 million in 2025, Kenya faces persistent socio-economic challenges, including high youth unemployment, widespread poverty, and ongoing food insecurity (United Nations, 2025; World Bank, 2024).

Social enterprise (SE) activity in Kenya can be traced back to the 1980s, when economic restructuring and declining government investment in social services created space for non-state actors, including NGOs and private organizations, to provide critical services (World Bank, 2022). The growth of SEs in Kenya has been shaped by socio-economic development, with the diversity and rapid expansion of the Kenyan SE ecosystem highlighting both the significant contribution SEs make toward tackling unemployment, fostering innovation, and supporting community development, as well as the gaps that remain in regulatory frameworks, access to finance, and sector-specific policies. SEs have increasingly played a key role in Kenya by addressing these challenges, providing innovative solutions, and creating employment opportunities (Siemens Stiftung, 2020).

Currently, Kenya hosts approximately 137,800 social enterprises, many led by women and youth (WEF, 2025). The number of SEs has more than tripled over the past decade, reflecting their growing importance in addressing community challenges. Collectively, these organizations have created around 796,000 full-time jobs, with 93% employing youth, 91% employing women, and 31% employing persons with disabilities (WEF, 2025). SEs operate predominantly in education (21%), health and well-being (18%), agriculture (15%), and civic engagement (14%), as well as in energy, water, and sanitation, demonstrating the diversity within the ecosystem (WEF, 2025).

Most SEs in Kenya are at the start-up (41%) or growth (33%) stage. However, the regulatory environment does not formally recognize SEs, resulting in most registering under general commercial or non-profit frameworks (World Bank, 2022). Recent reports indicate that 83% of SEs are formally registered, mainly as private companies (36%) or non-profits (29%), with a smaller proportion (5%) registered as benefit corporations (WEF, 2025). While institutions such as the Micro and Small-Sized Enterprise Authority and the State Department for MSME Development provide general enterprise support, no policies or legislation specifically address the unique needs of SEs in Kenya (Siemens Stiftung, 2020). Existing legislation, such as the NGO Coordination Act (1990) and the 2006 Sessional Paper, primarily governs NGOs rather than SEs themselves (British Council, 2016).

Despite these advances, SEs face persistent barriers. Key challenges include limited access to finance, cash-flow constraints, inadequate support services, and low public awareness of the social enterprise model (WEF, 2025). Access to capital remains the primary obstacle, often directed toward operational costs, expanding products and services, and investing in infrastructure. Navigating between non-profit and for-profit registration also creates practical challenges related to governance, taxation, and funding eligibility (WEF, 2025).

The absence of a dedicated legal framework has fostered the creation of networks and initiatives to support the sector. The Social Enterprise Society of Kenya (SESOK), established in 2017, provides technical support, networking opportunities, and raises awareness of social enterprises (Siemens Stiftung, 2020). Similarly, the East African Social Enterprise Network (EASEN) supports the sector through advisory services and advocacy across Kenya, Uganda, Tanzania, Rwanda, South Sudan, and Burundi (Siemens Stiftung, 2020). Kenya also benefits from a strong technical support ecosystem, with over 120 organizations offering business development services, advisory support, and networking opportunities, which have contributed to sector growth and visibility (Siemens Stiftung, 2020). By the 2010s, broader enterprise development structures, including the Micro and Small-Sized Enterprise Authority and the State Department for MSME Development, began offering programs that, while general in scope, positively influenced SE growth (Siemens Stiftung, 2020).

Overall, the SE sector in Kenya is of considerable relevance due to its significant social and economic contributions. The historical, regulatory, and institutional context underscores both the achievements and ongoing challenges that continue to shape the Kenyan SE ecosystem (World Bank, 2022).

The unit of analysis was the social entrepreneur actively involved in a social enterprise (SE), defined as a mission-driven organization that addresses social concerns while simultaneously striving to achieve financial goals (Urban & Bukula, 2024). Regarding sample size and frame, Kenya has an estimated 137,800 SEs, with Nairobi reporting the highest levels of social entrepreneurial activity (Siemens Stiftung, 2020; WEF, 2025). However, no formal or reliable database of SEs in Kenya exists. Therefore, a convenience sampling approach was employed, with the qualifying criterion being individuals “starting or currently leading any kind of activity, organization or initiative that has a particularly social, environmental or community objective” (Bosma et al., 2016, p. 9).

Based on this criterion, 550 surveys were administered across multiple locations in Nairobi to “ensure sufficient variability and a high regional representativeness” (Schindler, 2019). Participants were randomly selected from each regional subgroup to approximate the proportional distribution of SE activity across the city. Despite these efforts, the absence of a formal population database means no claims of sample representativeness can be made.

Following survey administration, and after multiple reminders over a three-week period, a total of 231 responses were received, representing an acceptable response rate for electronic surveys of this type (Schindler, 2019). Non-response bias was assessed using t-tests, which showed no significant differences between early and late respondents in terms of survey region. Ethical standards were strictly observed: all participants provided informed consent, and confidentiality of their responses was ensured.

Sample characteristics were as follows. Regarding the type of social venture, the majority (73.5%) were hybrid models, 18.1% were for-profit, and 8.5% were non-profit. In terms of gender, 61.45% of respondents were male, and 38.55% were female. Most respondents fell within the 20–30 age group (60.2%), with 33.7% in the 31–40 age range. Sector activity was concentrated in education, health and well-being, and civic engagement, with key priorities including youth, climate, and economic development. This aligns with evidence showing that SEs in Kenya operate across sectors such as education, health, agriculture, water management, and energy (WEF, 2025).

Regarding venture longevity, most SEs had been in operation for three to five years (46.8%). Geographically, the majority were located near Nairobi’s central business district, in areas including Westlands, Karen, Kilimani, Ngong Road, Ruaraka, and Embakasi. For additional details, see Table A1 in the appendix.

The survey questions were drawn from prior research and comprised two main sections, employing multi-item scales with a 7-point Likert scale ranging from 1 (Strongly Disagree) to 7 (Strongly Agree) (Schindler, 2019). Demographic data were also collected, including variables previously identified as relevant to social enterprise (SE) sustainability, such as the social entrepreneur’s age, gender, and years of SE operation (Jayawardhana et al., 2022).

For the study’s independent variables (IVs), the 36-item Generalized Measure of Adaptive Cognition (MAC; Haynie et al., 2010; Haynie & Shepherd, 2009) was adopted. This instrument has previously been used in emerging market contexts (see Urban, 2012). Drawing on existing theory and empirical work from cognitive and social psychology on metacognition, the MAC instrument’s validity has been previously confirmed, as its factor structure was consistent with five theoretically derived dimensions (Haynie & Shepherd, 2009). It has been proposed that “metacognition measures should explore heterogeneity in an individual’s performance across a wide variety of entrepreneurial tasks and situations”; accordingly, items for MAC were grouped based on the five dimensions of metacognitive awareness: goal orientation, knowledge, experience, choice, and monitoring (Haynie & Shepherd, 2009). Items were randomized during administration, and respondents were asked to rate themselves on a seven-point scale, with 1 representing “not very much like me” and 7 representing “very much like me.”

For the dependent variable (DV), measures were adapted from previous studies investigating SE sustainability and consolidated from established research instruments, including the social return on investment (SROI) index, a widely used tool for SE evaluation internationally (Millar & Hall, 2013); social impact measures from the Centre for the Advancement of Social Entrepreneurship (J. G. Dees et al., 2004); the Multidimensional Assessment Model (Bellucci et al., 2012); the Benefit Impact Assessment Model (Nigri & Del Baldo, 2018); the Vincentian Marketing Orientation (Miles et al., 2014); and the SAVE framework (Bassi & Vincenti, 2019). These instruments provide credible performance measures that allow SEs to assess and report on overall sustainability.

The justification for using these measures is that sustainability is closely aligned with SEs, insofar as SEs can promote sustainable development by addressing social issues through sustainable business models (Gupta et al., 2020). SEs can develop innovative solutions to social and environmental challenges, thereby contributing to the Sustainable Development Goals (SDGs) (Kamaludin et al., 2024; Singh et al., 2024). By drawing on the theoretical scaffolding of SE theory and relying on established constructs such as SROI, this study builds on and may extend the theoretical and empirical understanding of the SE–sustainability relationship. Using established measures is advantageous because they are supported by empirically meaningful relationships established in earlier studies, allowing for more advanced knowledge development (Cosa & Urban, 2025; Urban & Bukula, 2024).

The selected measures capture the study's objectives, including items focused on social impact and reach (four items), expandability and replicability (three items), and the sustainability of the venture (four items). All items were measured using seven-point Likert scales ranging from 1 (Strongly Disagree) to 7 (Strongly Agree). See Table 1 for examples of questions related to each of the IVs and the DV.

Statistical analyses were conducted using the Statistical Package for the Social Sciences (SPSS, version 25). Descriptive, correlational, and multiple regression analyses were performed. The data were first screened for incomplete responses, including missing values. Exploratory factor analysis (EFA) and confirmatory factor analysis (CFA) were conducted in SPSS to assess the validity of the measurement scales.

Because all data were collected from the same source, it was necessary to examine common method bias (CMB). Both procedural and statistical approaches were employed to address potential CMB, as recommended by Podsakoff et al. (2012). Procedurally, the scales were piloted with 20 respondents to ensure that the items were clear and unambiguous. Statistically, Harman’s one-factor test, implemented via principal component analysis (PCA), was conducted on all measurement items (Podsakoff et al., 2012, p. 560). The PCA results identified six components with eigenvalues greater than 1.0, with the first component accounting for only 17% of the variance. Because no single factor explained the majority of the variance, the results indicated no evidence of common method bias.

The Kaiser–Meyer–Olkin (KMO) measure of sampling adequacy calculated across all items was 0.829 (approx. χ² = 964.289, df = 231, p < .001), exceeding the recommended threshold of 0.5 and indicating suitability for exploratory factor analysis (EFA). Communalities for all items were above 0.3, suggesting adequate item correlations and acceptable explanation of variance.

The initial EFA identified ten factors with eigenvalues greater than 1. Factors with zero loadings or cross-loadings were eliminated, resulting in six factors being retained. Promax rotation with Principal Axis Factoring was applied, and the rotation converged in seven iterations. The six extracted factors accounted for a total variance of 71.06%, surpassing the commonly accepted threshold of 50% cumulative variance (Schindler, 2019). Table 1 presents the factor loadings for the final pattern matrix of the six retained factors.

Internal consistency was assessed using Cronbach’s alpha. Results (Table 1) indicate excellent reliability for sustainability (α = 0.827), goal orientation (α = 0.896), and metacognitive knowledge (α = 0.843); respectable reliability for metacognitive experience (α = 0.790) and metacognitive choice (α = 0.764); and acceptable reliability for metacognitive monitoring (α = 0.688), which falls within the minimally acceptable range of 0.65–0.70 (Tonidandel et al., 2018). All inter-item correlations exceeded 0.3, confirming convergent validity for all scales (Schindler, 2019).

Confirmatory factor analysis (CFA) was conducted to further validate the factor structure through cross-validation. Convergent validity at the construct level was assessed using the average variance extracted (AVE), with all constructs exceeding the recommended threshold of 0.50 (Hair et al., 2010). Discriminant validity was evaluated using the Fornell–Larcker criterion, confirming that the square root of each construct’s AVE exceeded its highest correlation with any other construct. Additionally, the heterotrait–monotrait (HTMT) ratio of correlations was examined to assess the true correlation between latent constructs under the assumption of perfect reliability (Hair et al., 2010). All HTMT ratios were below 0.70, well under the maximum recommended threshold of 0.90, thereby supporting discriminant validity. Cross-loading analyses further confirmed discriminant validity, as indicator variables loaded higher on their intended constructs than on any other constructs (Hair et al., 2010). Model fit was also supported by a standardized root mean square residual (SRMR) value ≤ 0.10, indicating acceptable fit. Composite reliability (CR) was assessed alongside Cronbach’s α values (see Table 1), with all CR scores exceeding the recommended threshold for internal consistency reliability (Hair et al., 2010).

Table 2 presents the descriptive statistics and Pearson’s correlation coefficients. Mean scores (M) for all metacognitive dimensions exceeded the scale midpoint, indicating generally high agreement on these factors. Sustainability returned a mean score slightly below the scale midpoint (M = 3.463). Several significant correlations were observed. According to Schindler (2019), correlation magnitudes can be interpreted as small (≥ 0.10), medium (≥ 0.30), or large (≥ 0.50). Results indicate positive relationships between sustainability and goal orientation (r = 0.269, p < .05), metacognitive experience (r = 0.335, p < .01), and metacognitive choice (r = 0.252, p < .05). Non-significant correlations were observed between sustainability and metacognitive knowledge and monitoring. Significant intercorrelations were also observed among the factors; however, all coefficients were below 0.9, indicating no evidence of multicollinearity at this stage of the analysis (Schindler, 2019).

The influence of the independent variables (IVs) on the dependent variable (DV), sustainability, was examined using multiple regression analysis. First, the normality of the data was assessed to confirm that the distribution of the data was appropriate for modeling (Schindler, 2019). Assumption testing for all extracted variables ensured that the data were normally distributed, contained no outliers, and met the linearity criterion. Table A2 and Figures A1-A3 in the appendix present the results of these analyses. According to the regression results shown in Table 3 and Table 4, the base model indicates an adjusted R² of 0.097, with all IVs together explaining approximately 15.3% of the variance in the DV (R²= 0.153). The Durbin–Watson statistic (d = 1.864) falls between the critical values of 1.5 and 2.5, indicating no evidence of first-order linear autocorrelation in the data (Schindler, 2019). The ANOVA results (not shown) indicate a statistically significant model fit, F = 2.771, p = .024.

Table 3 and Table 4 present the standardized regression β coefficients for each of the study hypotheses. For H1, the results show that goal orientation is positively associated with the sustainability of social enterprises; however, this relationship is not statistically significant (β = 0.160, p = .199), and H1 is therefore not supported. For H2, metacognitive knowledge is negatively associated with sustainability, but the effect is also non-significant (β = −0.086, p = .444); thus, H2 is not supported. For H3, metacognitive experience is positively and significantly related to the sustainability of social enterprises (β = 0.265, p = .044), supporting H3. For H4, metacognitive choice shows a positive but non-significant association with sustainability (β = 0.126, p = .322), and H4 is therefore not supported. Finally, metacognitive monitoring is negatively associated with sustainability, although the effect is not statistically significant (β = −0.102, p = .399), leading to the rejection of H5.

The collinearity of predictor variables was assessed using tolerance (TOL) values and variance inflation factors (VIF). When TOL values exceed 0.20 and VIF values are below 5, predictor variables do not exhibit elevated collinearity. In cases of elevated collinearity, strategies such as creating higher-order constructs, eliminating indicators, or merging indicators would be considered. As shown in the collinearity statistics in Table 4, all VIF values were below 10, indicating no problem of multicollinearity (Schindler, 2019).