Entrepreneurship centres (ECs) are increasingly embedded in higher education systems as institutional mechanisms to foster entrepreneurial mindsets and facilitate the transition from ideation to venture creation. ¹⁶ These centres provide experiential learning platforms where students engage in practical activities such as design thinking, opportunity recognition, mentorship programmes, and business incubation. Within the Nigerian context, ECs are a direct response to national directives, particularly the NUC’s entrepreneurship education mandate, aimed at equipping graduates with market-responsive skills. ¹⁷

Empirical studies highlight a positive association between student engagement in EC activities and entrepreneurial intention. ^{18, 19} However, institutional barriers such as infrastructural deficiencies, inadequate funding, and uneven access continue to challenge the scalability and impact of ECs on students’ entrepreneurial initiatives. ^{20, 21} In light of these, this study proposes the following hypothesis: H ₁:

Engagement with entrepreneurship centres significantly and positively influences innovative business idea generation among university students.

Innovative business idea generation refers to the capacity of individuals to conceptualise original, context-sensitive, and feasible business ventures. This process is both cognitive and environmental, relying on creativity, problem-solving, and access to supportive ecosystems. ^{22, 23} Research suggests that students exposed to dynamic and resource-rich entrepreneurial environments demonstrate higher levels of ideation and entrepreneurial activity. ²⁴ As such, entrepreneurship centres act as enabling ecosystems where idea generation is cultivated through immersion, feedback, and risk-taking.

General economic conditions (GECs), within the EST framework, are situated in the exosystem encompassing external forces that indirectly influence the individual by shaping institutional contexts and behavioural norms. ²⁵ In Nigeria, these include macroeconomic stressors such as high inflation, volatile interest rates, erratic power supply, and chronic insecurity. ²⁶

Although ECs offer students resources and learning environments conducive to entrepreneurship, the broader economic landscape can either enable or inhibit their application. For example, inflationary pressures may deter startup investments, while unstable electricity may hinder operational capacity. Conversely, GEC may also induce necessity-driven entrepreneurship, whereby students innovate out of compulsion rather than opportunity. ²⁷

The dual nature of GEC suggests a moderating influence on the efficacy of institutional interventions like ECs. ²⁸ Drawing on this understanding, two hypotheses were advanced: H ₂:

General economic conditions significantly influence innovative business idea generation among students.

This study adopted a quantitative, cross-sectional research design to investigate the influence of Entrepreneurship Centre Engagement (EnCen) on Innovative Business Idea Generation (IdGen) among students in Nigerian higher education institutions, with general economic conditions (GEC) assessed as a potential moderating variable. The research is anchored in the positivist paradigm, emphasising objective measurement, hypothesis testing, and the generalizability of findings across similar educational contexts in emerging economies. This paradigm supports the validation of theoretical propositions through empirical data and statistical inference.

Data were collected between February and March 2025 from undergraduate students enrolled in Business Administration, Accounting, and Economics programmes at a university within Nigeria’s North-Central geopolitical zone. These disciplines were purposively selected based on prior empirical evidence indicating a high propensity for entrepreneurial ideation among students in business-related fields. ^{23, 32}

A total of 269 valid responses were obtained from students who had been exposed to both entrepreneurship education and institutional entrepreneurial infrastructure, such as entrepreneurship centres. The sample comprised 105 males (39.03%) and 164 females (60.97%), all aged between 20 and 25 years, reflecting the typical age range for undergraduate students in Nigerian universities. The demographic breakdown is presented in Table 1.

A purposive sampling technique was employed to target students with exposure to entrepreneurship development studies (EDS). Participation was voluntary and anonymous, with informed consent obtained prior to questionnaire administration. Given the minimal risk involved and the observational nature of the study, formal institutional ethical clearance was not mandated under prevailing university guidelines.

Ethical approval

This study received ethical clearance from the Landmark University Institutional Research Ethics Committee (LMUIREC) under Ref. No: LMUIREC/HS/031/2025, dated 09–06–2025. The protocol was reviewed and approved under the oversight of the Landmark University Centre for Research Innovations and Discoveries (LUCRID). All data collection procedures adhered to the university’s ethical guidelines, and informed consent was obtained from all participants.

Data were collected via a structured, self-administered questionnaire adapted from previously validated scales and modified for contextual relevance. It measured three latent constructs where EnCen is the independent variable, IdGen (dependent variable) and GEC (moderating variable).

Entrepreneurship Centre Engagement (EnCen):

Captured students’ exposure to training, access to digital tools, and frequency of interaction with the centre. ^{8, 18} A sample statement is “I regularly participate in training sessions offered by the entrepreneurship centre”.

Innovative Business Idea Generation (IdGen):

Assessed creative thinking, confidence in idea development, and novelty frequency. ^{33, 34} A sample statement is “I frequently come up with new business ideas”.

General Economic Conditions (GEC):

Measured students’ perceptions of macroeconomic factors such as inflation, insecurity, power instability, and interest rates. ^{26, 35, 36} A sample statement is “Rising inflation and the cost of living make it difficult to finance a business idea”.

All items were rated using a five-point Likert scale (1 = Strongly Disagree to 5 = Strongly Agree). A pilot study involving 25 students was conducted to assess clarity, internal consistency, and contextual appropriateness, leading to minor refinements. Details of the scale items are presented in Appendix A.

Data were analysed using Partial Least Squares Structural Equation Modeling (PLS-SEM) in SmartPLS 4.0. ³⁷ This method was selected for its suitability in handling small-to-moderate sample sizes, tolerance for non-normal data, and its robustness in modeling complex relationships including moderation. ^{38, 39} Moreover, PLS-SEM is appropriate for exploratory theory development and prediction when using reflective measurement models.

Model evaluation proceeded in two stages, as recommended by Ref. 39. In the first stage, the measurement model was evaluated to ensure the reliability and validity of the reflective constructs. Internal consistency was assessed using Cronbach’s alpha and Composite Reliability (CR), both of which confirmed acceptable levels of reliability. Convergent validity was examined through the Average Variance Extracted (AVE), ensuring that each construct explained a substantial portion of the variance in its indicators. To establish discriminant validity, the Heterotrait-Monotrait (HTMT) ratio was employed, following the threshold guidelines proposed by Ref. 40, and results confirmed adequate distinction between the constructs.

In the second stage, the structural model was assessed to test the hypothesised relationships. This involved a bootstrapping procedure with 5,000 resamples to estimate the path coefficients, effect sizes (f ²), and statistical significance of each relationship. The model’s explanatory power was then evaluated using the coefficient of determination (R ²) to determine the proportion of variance explained in the dependent variable, and predictive relevance (Q ²) was assessed to confirm the model’s capability in forecasting outcomes.

The study utilised eighteen indicators across the three constructs, with all statistical thresholds meeting acceptable benchmarks for reliability and validity.

To enhance the predictive robustness of the structural model, the PLSpredict algorithm was applied to generate out-of-sample predictive performance. The model’s Root Mean Square Error (RMSE) and Mean Absolute Error (MAE) were benchmarked against linear regression and mean-based models. Results confirmed the model’s adequate predictive capability for the endogenous construct. ^{41, 42}

The final dataset consisted of 269 valid responses with no missing values. Descriptive analysis confirmed that the data distribution met acceptable thresholds for PLS-SEM analysis. Two measurement indicators, EnCen1 (loading = 0.188) and IdGen3 (loading = 0.596), were excluded from the analysis due to factor loadings below the acceptable threshold of 0.60. These adjustments, along with the final indicator loadings and R ² values, are presented in Figure 2. In addition, the removal decision was primarily to improve the model’s quality, as recommended by Refs. 39 and 41.

The measurement model was assessed to determine the reliability and validity of the constructs using standard PLS-SEM criteria. These assessments are presented in Table 2.

Internal consistency reliability was evaluated through Cronbach‘s alpha (α) and composite reliability (ρc), as recommended by. ⁴¹ While Cronbach’s alpha has known limitations, the use of ρc helps address underestimation concerns. Both EnCen and IdGen demonstrated high reliability, with α values of 0.952 and 0.836 respectively, and ρc values of 0.961 and 0.875, exceeding the recommended threshold of 0.70, indicating acceptable reliability.

Convergent validity was confirmed through the Average Variance Extracted (AVE), which exceeded the 0.50 benchmark for both constructs. EnCen recorded an AVE of 0.806, while IdGen had 0.541, indicating that the indicators accounted for more than half of the variance in their respective constructs, confirming convergent validity. Discriminant validity was further assessed using the Heterotrait-Monotrait (HTMT) ratio, which is considered a more robust approach than traditional methods such as the Fornell-Larcker criterion and cross-loadings. The HTMT value of 0.664 fell well below the conservative threshold of 0.85, suggesting no discriminant validity issues, in line with the criteria established by. ^{41, 43}

To examine the hypothesised relationships among constructs, the structural model was evaluated using the bootstrapping technique with 5,000 resamples. Figure 3 and Figure 4 present the structural models for both the direct and moderated relationships, respectively.

Table 3 presents the coefficient of determination(R ²⁾ and effect size (F ²). The model’s explanatory power was determined using the coefficient of determination (R ²). The R ² value for IdGen was 0.416, indicating that approximately 41.6% of the variance in innovative business idea generation was explained by the main effects and the interaction term. According to Ref. 38, this represents a moderate level of explanatory power. Furthermore, the effect size (f ²) for EnCen was 0.711, signifying a substantial effect on IdGen.

To evaluate the predictive validity of the model, the PLSpredict procedure was employed using 10-fold cross-validation, as proposed by Refs. 37, 42 and 44. The predictive accuracy of the PLS model was benchmarked against two reference models: the individual-mean (IA) model and the linear (LM) model. Assessment metrics included Root Mean Squared Error (RMSE), Mean Absolute Error (MAE), and average prediction loss differences. The Q ² value for IdGen was 0.466, indicating moderate-to-strong predictive relevance ( Table 4). The PLS model significantly outperformed the individual-mean (IA) model (p < 0.001) but did not demonstrate superiority over the linear model (LM) in this context.

The structural model was employed to test the proposed hypotheses, with the results presented in Table 5. All three hypothesised relationships were statistically significant. Engagement with the Entrepreneurship Centre (EnCen) exerted a strong positive effect on Innovative Business Idea Generation (IdGen), reflected by a path coefficient (β) of 0.645. General Economic Conditions (GEC) also demonstrated a significant positive impact on IdGen (β = 0.193). Interestingly, the interaction effect between EnCen and GEC on IdGen was negative and significant (β = –0.148), suggesting a dampening moderation effect of economic challenges on the positive influence of entrepreneurship centre engagement among students.

The moderation analysis showed that general economic challenges (GEC) significantly moderated the relationship between entrepreneurship centre engagement (EnCen) and innovative business idea generation (IdGen) (β = -0.148, t = 3.252, p = 0.001). Given that both the direct effect of EnCen on IdGen and the interaction effect were statistically significant, this result reflects a partial moderation. This implies that while EnCen positively influences innovative idea generation, its effect is weakened in the presence of adverse economic conditions.

By integrating ecological systems theory into the analysis, the study highlights the significance of both institutional (microsystem) and economic (exosystem) influences on students’ entrepreneurial development. The finding of partial moderation provides empirical support for the idea that institutional innovation must be contextualised within its broader socio-economic setting. It contributes to entrepreneurship literature by underscoring the contingent effectiveness of university-based interventions in economically strained environments.

For higher education leaders, this study highlights the urgency of institutional resilience. Entrepreneurship centres must go beyond offering standardised training to develop adaptive, resource-efficient models that are responsive to the realities of Nigeria’s economic environment. Institutions should also build stronger linkages with industry, financial institutions, and government agencies to buffer the external shocks that limit student ventures.

University managements should invest in digital infrastructure, mentorship networks, and incubator platforms while lobbying for greater policy alignment and funding from government bodies such as the Tertiary Education Trust Fund (TETFund). By aligning centre activities with national development goals, entrepreneurship centres can also become more strategic and better positioned to attract external funding.

From a policy perspective, this study underscores the need for a coherent entrepreneurship development framework that integrates educational reform with macroeconomic support. Government agencies must recognise that entrepreneurship development in universities cannot thrive in isolation from broader socio-economic reforms. This includes expanding access to credit facilities, reducing the cost of doing business, and providing stable infrastructure to support student-led enterprises.

While the study draws from a representative sample of Nigerian university students, the findings may not be directly generalizable to other developing countries with differing educational and economic conditions. However, the underlying theoretical framework remains applicable, and replication studies in other African or global south contexts (typically low- and middle-income nations facing systemic economic and institutional challenges) could help establish broader relevance and comparative insights.

In conclusion, this study affirms that university-based entrepreneurship centres are instrumental in promoting innovation and venture readiness among students in Nigeria. However, their impact is moderated, though not negated, by prevailing economic challenges. Thus, a dual strategy that reinforces institutional interventions with economic policy support is essential for fostering sustainable student entrepreneurship and national development.

All participants voluntarily consented to participate in the study after being informed about its purpose and their rights. No personally identifiable data was collected, and confidentiality was strictly maintained throughout.

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