Table 11. Joint display comparing... | Emerald Publishing

Table 11.

Joint display comparing business support and perception of business support

Quantitative results (n = 300)	Qualitative results (n = 15)	Interpretation
Central tendency: M = 3.06 (SD = 0.76) Neutral frequency Range on Likert scale = 1–5 Correlation: r = 0.446 (medium positive correlation) Independent samples t-test for growth vs no-growth firms (t = −1.810, p > 0.05) Regression: Business Support Services (β = 0.111, t = 2.579, p = 0.010) reveals a moderate but significant network effect: support services fill ecosystem gaps through self-organisation	Participants were generally positive about the role of the support institutions in Nelson Mandela Bay. However, some participants had concerns about the impact of monitoring Quotations: “In Uitenhage we actually train SMMEs on how to conduct meetings, how to tender for projects… because a lot of small entrepreneurs don’t know how to be compliant” (P3) “I know they have programmes… but the impact, I don’t know the extent of the impact… I know there are interventions, but I don’t know the impact” (P9) “We don’t have a formal system to measure the impact of those services. It’s very fragmented and disorganised… indicative of an ecosystem that’s disintegrated” (P10)	Integration type: Divergence (discordance) with complementary and supplementary expansion Complementary: Further analysis revealed that quantitative responses showed the highest “Neutral” frequencies for access to incubators and access to competent business consultants, suggesting information asymmetries that prevent effective agent connectivity essential for ecosystem self-organisation Supplementary: Qualitative findings revealed absent measurement systems for evaluating support institution effectiveness, indicating broken feedback loops that constrain adaptive capacityCAS property demonstrated – Incomplete self-organisation: The divergence between the regression finding (β = 0.111, significant network effect) and the qualitative accounts of absent impact measurement reveals a specific and theoretically important CAS dynamic: support institutions demonstrate partial self-organisation; they coordinate activity (training, cluster formation, networking) effectively at the output level but fail to close the adaptive feedback loop between outputs and outcomes. In CAS terms, adaptive systems require feedback mechanisms that allow agents to modify their behaviour in response to system-level results. The absence of formal impact measurement means that support institutions cannot determine whether their interventions are producing the emergence conditions they are designed to create. This is not a managerial failing; it is a structural CAS property, incomplete self-organisation in which the self-organising capacity of support agents is real but bounded: agents can coordinate around activity delivery but cannot adaptively recalibrate their strategies based on ecosystem-level feedback The non-significant t-test result (p > 0.05) combined with the significant regression coefficient (p = 0.010) is itself a CAS signal: at the group level, support access does not differentiate growth from no-growth firms (t-test), but within the multivariate model, support services contribute a measurable network effect above and beyond other factors. This suggests that support institutions operate as boundary-spanning agents, filling gaps created by other ecosystem dysfunctions rather than as direct drivers of firm growth. Their significance in the regression model reflects their role in system maintenance under constraint, not transformative emergence

Quantitative results (n = 300)	Qualitative results (n = 15)	Interpretation
Central tendency: M = 3.06 (SD = 0.76) Neutral frequency Range on Likert scale = 1–5 Correlation: r = 0.446 (medium positive correlation) Independent samples t-test for growth vs no-growth firms (t = −1.810, p > 0.05) Regression: Business Support Services (β = 0.111, t = 2.579, p = 0.010) reveals a moderate but significant network effect: support services fill ecosystem gaps through self-organisation	Participants were generally positive about the role of the support institutions in Nelson Mandela Bay. However, some participants had concerns about the impact of monitoring Quotations: “In Uitenhage we actually train SMMEs on how to conduct meetings, how to tender for projects… because a lot of small entrepreneurs don’t know how to be compliant” (P3) “I know they have programmes… but the impact, I don’t know the extent of the impact… I know there are interventions, but I don’t know the impact” (P9) “We don’t have a formal system to measure the impact of those services. It’s very fragmented and disorganised… indicative of an ecosystem that’s disintegrated” (P10)	Integration type: Divergence (discordance) with complementary and supplementary expansion Complementary: Further analysis revealed that quantitative responses showed the highest “Neutral” frequencies for access to incubators and access to competent business consultants, suggesting information asymmetries that prevent effective agent connectivity essential for ecosystem self-organisation Supplementary: Qualitative findings revealed absent measurement systems for evaluating support institution effectiveness, indicating broken feedback loops that constrain adaptive capacityCAS property demonstrated – Incomplete self-organisation: The divergence between the regression finding (β = 0.111, significant network effect) and the qualitative accounts of absent impact measurement reveals a specific and theoretically important CAS dynamic: support institutions demonstrate partial self-organisation; they coordinate activity (training, cluster formation, networking) effectively at the output level but fail to close the adaptive feedback loop between outputs and outcomes. In CAS terms, adaptive systems require feedback mechanisms that allow agents to modify their behaviour in response to system-level results. The absence of formal impact measurement means that support institutions cannot determine whether their interventions are producing the emergence conditions they are designed to create. This is not a managerial failing; it is a structural CAS property, incomplete self-organisation in which the self-organising capacity of support agents is real but bounded: agents can coordinate around activity delivery but cannot adaptively recalibrate their strategies based on ecosystem-level feedback The non-significant t-test result (p > 0.05) combined with the significant regression coefficient (p = 0.010) is itself a CAS signal: at the group level, support access does not differentiate growth from no-growth firms (t-test), but within the multivariate model, support services contribute a measurable network effect above and beyond other factors. This suggests that support institutions operate as boundary-spanning agents, filling gaps created by other ecosystem dysfunctions rather than as direct drivers of firm growth. Their significance in the regression model reflects their role in system maintenance under constraint, not transformative emergence

Source(s): Authors’ own work