How Business Incubation Dimensions Drive Entrepreneurial Outcomes: Longitudinal Evidence from a Small Island Developing State

1. Introduction

Business incubators have become central policy instruments for entrepreneurship development worldwide. Starting in the United States in the 1950s, the model has proliferated to more than 7,000 programs globally (NBIA, 2021), promising to accelerate venture creation, create employment, and stimulate regional economic development. Notwithstanding this institutional prominence, the empirical base of incubation effectiveness remains incomplete in two important respects.

First, the overwhelming majority of impact studies are cross-sectional in design (Albort-Morant & Ribeiro-Soriano, 2016; Hackett & Dilts, 2004). Cross-sectional data can ascertain associations but cannot establish whether incubation services drive changes in entrepreneurial outcomes over time or track how those effects evolve within the program period. The rare longitudinal studies in this field compare incubated and non-incubated firms at a single follow-up rather than tracking the same cohort repeatedly (Colombo & Delmastro, 2002). As a result, the temporal dynamics of incubation impact, the dimensions that drive which outcomes, and when, remain largely uncharted.

Second, the geography of incubation research has been highly concentrated. A systematic content analysis of 76 high-impact articles published between 2001 and 2020 (Meahjohn, 2023) found no peer-reviewed empirical study situated in the Caribbean or any Small Island Developing State (SIDS). This absence is striking because SIDS economies face distinctive structural vulnerabilities, such as small domestic markets, high import dependence, commodity sector reliance, and acute exposure to external shocks. These make entrepreneurship-led diversification both more urgent and hypothetically more complicated than in larger, diversified economies (Briguglio, 1995; UNCTAD, 2014). Trinidad and Tobago exemplifies these dynamics: a hydrocarbon-dependent economy in which the energy sector historically crowded out non-energy entrepreneurship, leaving the diversification imperative both well-recognized and underdelivered.

This study addresses both gaps. Findings are reported from a three-year repeated-measures longitudinal study (2017-2019) of 106 entrepreneurs enrolled in uSTART, the business incubator of the University of Trinidad and Tobago (UTT), the largest university-based incubator in Trinidad and Tobago. The study sample represented 90.6% of the total incubator population during the study period. Four OLS regression models are estimated, in which the four core dimensions of business incubation, services, mentorship, finance, and physical space predict four entrepreneurial outcome dimensions: job creation, new venture creation, entrepreneurship development, and economic growth.

The study makes three contributions. First, it provides the first peer-reviewed longitudinal evidence on incubation effectiveness from the Caribbean and SIDS, filling a documented geographic gap. Second, by disaggregating both incubation inputs and entrepreneurial outputs, the results reveal that mentorship is the dominant and consistent driver across all outcome domains, whereas the relative influence of finance, services, and physical space varies by outcome type. Third, the repeated-measures design documents the full developmental trajectory of ventures over the three-year program period, from a near-universal early-stage status at entry to a cohort predominantly earning income and pursuing expansion or international markets at program completion.

The remainder of this paper is organized as follows. Section 2 presents the theoretical framework and literature review. Section 3 describes the research context, analytical approaches, and the data. Section 4 presents the results of the study. Section 5 discusses the findings and implications of this study. Section 6 concludes.

2. Literature Review and Theoretical Framework

2.1 The Four Dimensions of Business Incubation

Business incubators are organizational environments designed to accelerate the growth and success of entrepreneurial firms through an array of bundled support resources (Bruneel et al., 2012). The theoretical rationale for incubation is based on multiple traditions. Resource-based theory posits that incubators address the resource poverty of nascent ventures by providing access to physical space, advisory expertise, financial networks, and peer communities within a single program (Hackett & Dilts, 2004). The knowledge spillover theory of entrepreneurship (Audretsch & Lehmann, 2005) highlights university-based incubators as mechanisms for commercializing academic knowledge, while the social capital theory (Greve & Salaff, 2003) emphasizes the role of incubator-mediated networks in providing entrepreneurs with information, relationships, and legitimacy.

Despite theoretical diversity, empirical research has converged on four core incubation dimensions. Incubator services, such as training, advisory support, and professional network access, provide entrepreneurs with skills and connections that are unlikely to develop independently in the early venture stage (Mian, 1996; Mahmood et al., 2017). Mentorship provides experiential guidance from practitioners and domain experts. Politis (2005) theorizes that tacit entrepreneurial knowledge is primarily transmitted through sustained mentoring relationships. Finance encompasses both direct funding and facilitated access to investors and grant programs (Al-Mubaraki & Busler, 2012; Lalkaka, 2002). Physical space provides a legitimizing professional environment, shared overhead reduction, and proximity-based peer learning, which Levakova (2012) refers to as the ‘incubator effect.’

Empirical evidence on the relative importance of these dimensions is mixed across different contexts. Colombo and Delmastro (2002) found network access and technological support to be the most influential in Italy, Thebtaranoth (2007) emphasized physical infrastructure in Thailand, and Mahmood et al. (2017) found mentorship and advisory services to be primarily in Pakistan. This cross-national variation suggests that institutional context moderates the relative salience of incubation dimensions, highlighting the value of studying understudied contexts such as SIDS.

2.2 Entrepreneurial Outcomes: A Multi-Dimensional Framework

Business incubation research has employed heterogeneous outcome measures, creating challenges for cross-study comparisons (Albort-Morant & Ribeiro-Soriano, 2016). The most common indicators are firm survival, employment generation, and revenue growth (Schwartz, 2013). However, this narrow focus has been criticized as insufficient to capture the broader developmental mandate of incubation programs in developing economies, where incubators are expected to generate not only viable firms, but also human capital, entrepreneurial culture, and macroeconomic diversification (Eshun, 2009; Allahar & Brathwaite, 2016).

Following Campbell’s (1989) framework and its application in developing country contexts (Al-Mubaraki & Busler, 2012), this study operationalizes entrepreneurial outcomes across four dimensions: job creation, new venture creation, entrepreneurship development, and economic growth. This multi-dimensional approach enables a richer assessment of incubation impact than single-metric approaches and is particularly appropriate in an SIDS context in which each dimension contributes to the overarching national goal of economic diversification.

2.3 The Longitudinal Gap

The most consequential methodological limitation in incubation research is its near-universal reliance on cross-sectional data. Albort-Morant and Ribeiro-Soriano’s (2016) bibliometric review identified longitudinal designs as rare exceptions, rather than norms. Most studies capture a single time point and rely on retrospective self-reports of incubation impact, limiting causal inference, and obscuring the temporal dynamics of venture development.

The few longitudinal studies in this literature adopted single-follow-up rather than repeated-measures designs. Colombo and Delmastro (2002) compared Italian firms at a three-year follow-up but did not administer repeated instruments to the same cohort during incubation. Schwartz (2013) examined survival at multiple intervals, but did not track within-incubation development trajectories. The present study contributes a repeated-measures design of three annual waves administered to the same 106 participants, allowing observation of the full developmental trajectory during active program participation and enabling within-period estimation of the effects of incubation dimensions on outcomes.

2.4 SIDS as a Distinctive Research Context

Small Island Developing States are characterized by high structural vulnerability: small domestic markets, openness-driven exposure to external shocks, limited factor accumulation, and dependence on narrow economic sectors (Briguglio, 1995). In resource-dependent SIDS such as Trinidad and Tobago, the ‘Dutch Disease’ mechanism, in which energy rents appreciate the exchange rate and crowd out tradeable sector development, has historically suppressed the entrepreneurship ecosystem outside the commodity sector (Meahjohn, 2023).

This context is theoretically important for several reasons. First, necessity entrepreneurship is likely to be more prevalent than opportunity entrepreneurship given that employment alternatives in diversified sectors are scarce. Second, entrepreneurial role models and experienced mentors from non-energy sectors are genuinely scarce, making the mentorship dimension of incubation potentially more impactful than in economies with dense entrepreneurial ecosystems. Third, a small domestic market means that incubation success requires rapid orientation toward export markets, making the scope and international ambition of incubated ventures a distinctive outcome of interest.

3. Methodology

3.1 Research Context

The uSTART Business Incubator was launched at the University of Trinidad and Tobago in September 2014. It is the largest university-based incubator in Trinidad and Tobago, operating a dual model in which resident companies occupy physical space within the facility, while virtual companies receive capacity-building and business development support remotely. The program spans more than 20 economic sectors, including information and communications technology, animation and media production, agri-business, food production, engineering, environmental services, health care, and professional and technical services, all aligned with national economic diversification priorities.

uSTART’s stated mandate encompasses job creation, tax revenue generation, community development, and the cultivation of an entrepreneurial culture in communities historically dependent on large employers or public-sector employment. This multidimensional mandate makes uSTART a particularly well-suited setting for testing the multi-outcome framework employed here.

3.2 Research Design and Data Collection

A quantitative repeated-measures longitudinal design was employed, with three annual measurement waves. Wave 1 was administered between September 10 and October 25, 2017 (45 days); Wave 2, between October 1 and November 6, 2018; and Wave 3, between July 22 and August 27, 2019. A structured self-administered questionnaire with 50 items was administered to all active uSTART participants at each wave. The research philosophy was positivist and deductive, with hypotheses specified prior to the data collection.

The target population comprised all 117 active uSTART entrepreneurs across five geographic regions of Trinidad and Tobago (North, East, Central, South, and Tobago). Of the 117 eligible participants, 106 provided complete responses and consented to participate in all three waves, yielding a final sample of 106 and a response rate of 90.6% (106/117). Critically, all 106 participants completed the questionnaires at all three time points (318 total questionnaires), producing a fully balanced panel with zero attrition across the study period. The average completion time is approximately one hour per wave.

3.3 Variables and Operationalisation

The independent variable, business incubation, was disaggregated into four dimensions, each measured using a four-item Likert scale (1 = very low, 5 = very high) derived from instruments validated in prior research (Dahleez, 2009; Mehmood et al., 2017):

·       services (4 items, α = .73): range of business support services available to incubatees, including training, advisory support, and professional network access.

·       mentorship (4 items, α = .79): Quality and frequency of mentoring relationships with experienced practitioners and domain experts.

·       finance (4 items, α = .82): Access to financial support, including incubator-facilitated introductions to investors, grant assistance, and seed capital advisory.

·       physical Space (4 items, α = .84): adequacy and functionality of the physical working environment.

The dependent variable, entrepreneurship, was operationalised across four outcome dimensions:

·       job creation (4 items, α = .76): Perceived incubation impact on team building and employment generation.

·       new venture creation (5 items, α = .78): progress in establishing a new operating business entity.

·       entrepreneurship development (5 items, α = .81): Growth in entrepreneurial skills, competencies, and business acumen.

·       economic growth (5 items, α = .83): Perceived contribution of the venture to income generation, market penetration, and broader economic activity.

The 35-item instrument achieved a composite Cronbach’s alpha of .809. Factor analysis with Principal Component Analysis confirmed construct validity: KMO values ranged from .73 (Mentorship) to .92 (Services), all well above the .50 threshold, and all item loadings exceeded .30.

3.4 Analytical Approach

Descriptive statistics (frequencies, means, and standard deviations) were computed for all variables at each wave and are reported as outcome indicators in Table 3. The Pearson Product-Moment Correlation was applied to examine bivariate associations among the independent and dependent variables. Four OLS multiple regression models were then estimated for each entrepreneurial outcome dimension, with all four incubation dimensions simultaneously entered as predictors. This analytical approach follows those of Mahmood et al. (2017) and Yamockul et al. (2019). All analyses were conducted using SPSS v26.

A full multilevel model that exploits the longitudinal panel structure offers methodological advantages over cross-sectional OLS. Given that this study constitutes the first peer-reviewed quantitative investigation of incubation in the SIDS context, OLS regression is presented as a transparent baseline model, while longitudinal trajectories are descriptively documented. This approach mirrors the precedents of pioneering incubation studies in other emerging economies (Colombo & Delmastro, 2002; Thebtaranoth, 2007).

4. Results

4.1 Sample Characteristics

The sample (N = 106) was drawn from all five geographic regions of Trinidad and Tobago: North (34.9%, n = 37), East (30.2%, n = 32), Central (24.5%, n = 26), South (8.5%, n = 9), and Tobago (1.9%, n = 2). The sex composition was 72.6% male and 27.4% female. The age distribution was skewed young, with 77.4% of participants aged 18-24 years, reflecting the university affiliation of the program. Educational attainment was high; 76.4% had completed tertiary undergraduate education, 4.7% held postgraduate qualifications, and 18.9% held secondary-level qualifications.

A large majority of participants (88.7%) had no entrepreneurship experience before entering the program, and only 13.2% had parents who owned businesses, indicating a first-generation entrepreneurship population. The primary motivation for participation was the absence of employment (55.7%), followed by continuation of studies (31.1%), and perceived market opportunity (13.2%). This necessity-entrepreneurship orientation is consistent with theoretical expectations in a resource-dependent SIDS economy, and distinguishes this sample from entrepreneurship populations in high-income economy studies.

Participants represented 14 economic sectors. The most frequently represented were information technology (19.8%), animation and media production (14.2%), professional and technical services (13.2%), food and beverage, environmental services, and small manufacturing were also well represented. This sectoral breadth aligns with the uSTART mandate to support economic diversification beyond the energy sector.

4.2 Instrument Reliability and Validity

Table 1 presents Cronbach’s alpha reliability coefficients for all eight study variables. All values exceeded the conventional .70 threshold (range: .73-.84 for incubation dimensions; .76-.83 for outcome dimensions), and the composite 35-item instrument achieved α = .809, indicating high internal consistency. Factor analysis confirmed construct validity: KMO values were .92 (Services), .73 (Mentorship), .80 (Finance), and .86 (Physical Space). All items exceeded the extraction threshold of .30 in principal component analysis. The four-component solution explained 83.4% of the total variance (Table 2).

Table 1: Reliability of the Research Instrument

Note. Total questionnaire items = 50 (35 variable items + 15 demographic items). All α values exceed the .70 threshold for acceptable internal consistency.

Table 2: Factor Analysis - Communalities (Construct Validity)

Note. Extraction Method: Principal Component Analysis. KMO values: Services = .92, Mentorship = .73, Finance = .80, Physical Space = .86 (all > .50). Four-component solution explains 83.4% of total variance. All item loadings > .30.

4.3 Longitudinal Developmental Trajectories

Before presenting the regression results, repeated-measures descriptive data are presented to document the entrepreneurial developmental trajectory across the three program years. This forms an independent layer of evidence on incubation effectiveness that complements the regression analysis.

4.3.1 Business Stage Progression

The progression of incubatees across venture development stages across the three waves is striking. In Wave 1 (2017), 91.5% of participants reported their businesses as being at an early stage and 8.5% at start-up (beginning operations); no participants earned income, pursued expansion, or sought international market entry. In Wave 2 (2018), early-stage status fell to 50.0%, active start-up operations increased to 44.3%, and 5.7% reported operating and earning income. By Wave 3 (2019), only 0.9% remained at an early stage, 41.5% were in start-up operations, 35.8% were operating and earning income, 12.3% were pursuing expansion, and 9.4% sought extra-territorial or international market entry. These trajectories demonstrate a clear, sustained progression across the full five-stage development spectrum over three years of incubation.

4.3.2 Descriptive Statistics by Outcome Indicator and Wave

Table 3 presents the means and standard deviations for the representative outcome items for each of the three annual waves. All items are measured on a five-point Likert scale (1 = very low impact, 5 = very high impact).

The pattern is consistent and striking across all outcome indicators: all items begin at M = 1.00, SD = 0.00 in Wave 1 (2017), reflecting that every participant assessed incubation impact as very low at program entry. In Wave 2 (2018), means range from 3.21 to 3.99, indicating a transition from low to moderate to high perceived impact. In Wave 3 (2019), means range from 4.27 to 4.85, indicating consistently high to very high perceived impact on all dimensions.

Table 3: Descriptive Statistics for Outcome Indicators by Wave (N = 106, 5-point Likert Scale)

Note. N = 106 at each wave. Scale: 1 = very low impact, 5 = very high impact. Wave 1 SD = .00 for all items because all 106 respondents rated impact as ‘1 - Very Low’ at program entry. Wave 3 SDs substantially smaller than Wave 2, reflecting convergence in perceived program benefit.

4.3.3 Other Longitudinal Indicators

Complementary indicators further documented the developmental trajectory. Export intention increased from 15.1% in Wave 1, to 34.0% in Wave 2, and to 46.2% in Wave 3, representing a 31.1 percentage-point in increase. This suggests that incubation builds international market orientation among the necessity-entrepreneur population in a small domestic market. Demand for parking space (a proxy for team growth and business formalization) rose from 49.1% in Wave 1 to 100% in Wave 3. Interest in all incubation services simultaneously rose from 21.7% in Wave 1 to 94.3% in Wave 3, reflecting the broadening of service awareness and utilization as entrepreneurial experience accumulated.

4.4 Regression Results

Table 4 presents the results of the four OLS regression models. Each model regresses one entrepreneurial outcome dimension for the four incubation dimensions. Standardized beta coefficients (β), t-statistics, and significance levels are reported. The overall mean R² across all the four models was .577.

Table 4: OLS Regression Results: Impact of Business Incubation Dimensions on Entrepreneurial Outcomes (N = 106)

Note. N = 106. ◆ = dominant predictor (highest β) in that model, shown bold italic. R² reported at final row of each model. n.s. = not significant. p < .05, ** p < .001.

4.4.1 Model 1: Job Creation (R² = .699)

The first model explained 69.9% of the variance in job creation, which was the strongest model fit in this study. Mentorship is the dominant predictor (β = .96, t = 8.80, p < .05), with finance as the second significant contributor (β = .80, p < .001). Services (β = .08, p < .001) and physical space (β = .06, p = .023) were also significant, although substantively modest. The large mentorship effect on job creation suggests that experiential guidance from practitioners, including guidance on team building, human resource management, and operational scaling, is the primary mechanism by which incubation drives employment generation.

4.4.2 Model 2: New Venture Creation (R² = .380)

The second model explains 38.0% of the variance in new venture creation. Mentorship is again the strongest significant predictor (β = .43, t = 2.90, p < .001) and finance is marginally significant (β = .19, p = .050), consistent with the expectation that access to capital is an enabling condition for enterprise formalization. Services (β = .02, p = .16) and physical space (β = .31, p = .08) did not reach significance. The lower R² relative to other models likely reflects that new venture creation is also shaped by factors outside the incubator’s direct control, including entrepreneurial experience, market timing, and personal risk tolerance.

4.4.3 Model 3: Entrepreneurship Development (R² = .570)

Model 3 explains 57.0% of the variance in entrepreneurship development. Mentorship is the dominant predictor with a large and highly significant effect (β = 1.08, t = 6.10, p < .001), consistent with experiential learning theory (Politis, 2005): tacit entrepreneurial knowledge is transmitted primarily through direct engagement with experienced practitioners. Physical space is significant (β = .23, p = .012), suggesting that the incubator’s professional environment contributes to entrepreneurial identity formation and competency development via peer co-location effects (McAdam & Marlow, 2007; Levakova, 2012). Services were marginally significant (β = .15, p = .044), while finance was not (β = .16, p = .199).

4.4.4 Model 4: Economic Growth (R² = .650)

Model 4 explains 65.0% of the variance in economic growth. Mentorship achieved its highest beta coefficient across all four models (β = 1.94, t = 2.90, p = .035), underlining its dominant and broad-based role. Finance is also highly significant (β = .08, p < .001), consistent with the expectation that capital access is important for translating entrepreneurial activity into measurable economic output. Services (β = .013) and physical space (β = .16) were not significant in this model.

5. Discussion

5.1 The Primacy of Mentorship

One of the most striking and consistent findings was the dominant role of mentorship across all the four regression models. This result builds on and provides more precise quantification than prior studies that have theorized mentorship as a core value-adding mechanism of incubation (Voisey et al., 2006; Thebtaranoth, 2007; Politis, 2005). The strength of the mentorship effect is particularly noteworthy for economic growth (β = 1.94), suggesting a multiplier-like mechanism in the SIDS context.

This amplification can be theoretically interpreted using the scarcity lens developed in Section 2.4. Experienced non-energy entrepreneurs are rare in an economy historically controlled by a single commodity sector. When incubatees, predominantly first-generation entrepreneurs (88.7% without prior experience; 86.8% without entrepreneurial parents) gain access to such mentors through the program, the informational and relational value transferred may be substantially higher than in economies with dense entrepreneurial ecosystems. The descriptive findings in Table 3 are consistent with this interpretation: The uniformly zero baseline (M = 1.00 at Wave 1) followed by dramatic increases suggests that participants arrived with minimal exposure to entrepreneurial models and resources.

The implication is direct: if mentorship explains the largest share of variance across the four outcome dimensions, then the quality, sector experience, and adequate cultural alignment of mentors should be the key focus of incubator program design and performance evaluation. Qualitative focus group data gatred as part of doctoral research (Meahjohn, 2023) found incubatees themselves identified mentorship as pivotal but systematically under-resourced, corroborating this regression finding and pointing to a structural investment need.

5.2 The Enabling Role of Finance

Finance is a significant predictor of job creation (β = .80) and economic growth (β = .08, p < .001) but not entrepreneurship development (β = .16, p = .199). This differential pattern was theoretically significant. Access to capital enables the hiring of staff and scaling of the operation, directly driving employment and economic output, but it does not build entrepreneurial competencies by itself. Skills and capabilities are developed through human interaction (mentoring and training) rather than through financial resources alone. This distinction between the enabling (finance, physical space) and developmental (mentorship, services) functions of incubation is a theoretical contribution that has not been clearly articulated in the literature.

Qualitative evidence from uSTART focus group participants gathered as part of this doctoral research (Meahjohn, 2023) identified thin angel and venture capital markets, and conservative bank lending as structural barriers in Trinidad and Tobago. The regression finding that finance significantly predicts job creation and economic growth indicates that expanding incubatees’ access to capital would likely result in high returns. Government-facilitated matching with angel investors and venture capital organizations or structured grant access should be a priority policy complement to mentorship investment.

5.3 Physical Space as a Developmental Resource

Physical space is significant in the job creation model (β = .06, p = .023) and entrepreneurship development model (β = .23, p = .012). In the former case, co-location likely functions as an enabler of professional legitimacy and team recruitment. In the latter, the ‘incubator effect,’ theorized by Levakova (2012), in which shared space creates informal peer education, and mutual motivation appears to operate empirically. These results concur with the implicit assumption that physical presence is instrumental in the rapid progression of incubation programs. For a population of young first-generation entrepreneurs in a developing economy context, a professional physical environment may be a meaningful developmental resource that virtual substitutes cannot adequately replace.

5.4 Developmental Trajectories and Longitudinal Evidence

The repeated-measures data in Table 3 provide important evidence that cross-sectional studies cannot provide. The universal M = 1.00 baseline at Wave 1, reflecting that every participant assessed incubation impact as very low at program entry, establishes a genuine pre-treatment starting point. The subsequent increases to M = 3.21-3.99 at Wave 2 and M = 4.27-4.85 at Wave 3 demonstrate a consistent, sustained improvement in perceived incubation impact across all measured dimensions. The narrowing standard deviations at Wave 3 indicate increasing homogeneity in the positive evaluation of the program as participants accumulated experience.

These descriptive trajectory data, combined with the business stage progression (from 91.5% early stage in Wave 1 to 35.8% earning income and 21.7% pursuing expansion or international markets in Wave 3) and the increase in export orientation (from 15.1% to 46.2%), provide converging evidence of incubation effectiveness. The increase in export orientation is especially significant in the SIDS context, in which small domestic markets make international market orientation a prerequisite for scaling.

5.5 Limitations

This study has some limitations that must be acknowledged. The single-program design (uSTART only) restricts generalizability within Trinidad and Tobago and across the Caribbean region. uSTART is the largest and most resource-endowed university incubator in the country, and programs with fewer resources may yield different outcomes. The findings are framed as an exploratory case study that establishes the first empirical ground following the precedent of Colombo and Delmastro’s (2002) Italian case.

The OLS regression approach to pooled data does not adequately exploit the longitudinal panel structure. Future research should employ multilevel modelling (entrepreneurs nested within time periods) or fixed-effects estimators to control for time-invariant unobserved heterogeneity. Outcome measures were self-reported perceptions and not objective administrative data, and the absence of a non-incubated comparison group prevented causal attribution of developmental trajectories to incubation per se.

6. Conclusion

This study provides the first peer-reviewed longitudinal evidence of business incubation effectiveness in the Caribbean and Small Island Developing States. Drawing on a three-year repeated-measures panel (N = 106, 2017-2019) representing 90.6% of the uSTART business incubator population at the University of Trinidad and Tobago, four OLS regression models were estimated, and the full developmental trajectory of incubated ventures across the program period was documented.

The main conclusion is that mentorship is the dominant driver of entrepreneurial outcomes across all four dimensions tested: job creation (β = .96), new venture creation (β = .43), entrepreneurship development (β = 1.08), and economic growth (β = 1.94), with the strongest effects observed in the models directly relevant to economic policy. Finance is the second-most consistently significant predictor of output-oriented outcomes. The four-dimensional model explains 57% of the variance in entrepreneurial outcomes overall, with job creation showing the highest single-model fit (R² = .699).

The longitudinal panel reveals a striking and coherent developmental trajectory. At program entry, all 106 participants uniformly assessed the incubation impact as very low (M = 1.00) across all outcome indicators. By program completion, means had risen to 4.27-4.85 on a five-point scale, early-stage businesses had been almost entirely eliminated, and nearly a quarter of the cohort was actively pursuing expansion or international market penetration.

The findings strongly argue for treating mentor quality as the highest-return investment in incubation program design. For policymakers in Trinidad and Tobago and analogous SIDS economies, this study offers evidence that university-based incubators can function as genuine instruments of economic diversification among a necessity-entrepreneur population, provided that high-quality, sector-experienced mentors are available and that financial access barriers are addressed.

Multi-program comparative studies throughout the Caribbean are among the future research priorities, multilevel modelling to exploit the panel structure, integration of administrative outcome data, matched-comparison group designs, and cross-SIDS comparisons extending to Jamaica, Barbados, and Pacific Island economies.

Funding: This research received no external funding.

Conflicts of Interest: The author declares no conflict of interest. The funding sponsors had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, and in the decision to publish the results.

Informed Consent Statement/Ethics Approval: All participants gave their verbal informed consent for inclusion before they participated in the study. Anonymity was assured, participants were informed of the research purpose, and data were collected, anonymized, and handled in accordance with the ethical guidelines of the host institution. No interventional procedures were conducted.

Data Availability Statement: Data supporting the reported results are available from the corresponding author upon reasonable request.

Declaration of Generative AI and AI-assisted Technologies: This study has not used any generative AI tools or technologies in the preparation of this manuscript.