Self-reported competency gaps of farmer-incubatees in Philippine agribusiness cooperatives: evidence to inform a preliminary competency-based training framework

Introduction

Competency-based training (CBT) is widely used in entrepreneurship and incubation programs because it links instruction to observable performance criteria rather than to content coverage alone. In agribusiness incubation, incubatees must combine technical know-how with managerial, marketing, financial, and operational decision-making, and cooperatives play a documented role in enabling resource sharing, collective bargaining, and technology integration among smallholders (Kalogiannidis et al., 2024). Entrepreneurial marketing capability has likewise been linked to the competitiveness of agro-based micro, small, and medium enterprises in Southeast Asia (Sarma et al., 2022), and farmer-level evidence connects entrepreneurial and agribusiness competencies to production efficiency, income, and profitability (Arellano and Delos Reyes, 2019; Nain et al., 2019; Rayuddin et al., 2021; Taruc, 2025). Within cooperatives specifically, skills gaps have been documented among farmer directors (Herchenbach et al., 2026), and the managerial competencies of officers have been associated with the financial performance of agricultural micro-cooperatives in the Philippines (Bautista, 2025). In the Philippine context, an aging farm population and weak youth succession (Palis, 2020), gendered and institutional barriers to development resources (Ani and Casasola, 2020), and the promotion of capacity building and cooperatives as pathways to viable community enterprises and sustainable food marketing (Paladan, 2021; Quilloy, 2017) frame the training problem.

This study was conducted with farmer-incubatees of Department of Agriculture-supported cooperatives in Region IV-A, Philippines, who operated primarily as agriculture-based micro-enterprises under cooperative and regulatory constraints. Within the training needs assessment tradition, a training need is a gap between required and current performance, and competency assessment is an established basis for determining training needs (Ghimire et al., 2017). Because no external required-performance benchmark was administered, a training need was operationalized as a low self-reported competency level (a composite domain score within or below the Basic band of a four-point self-rating scale), triangulated against qualitative accounts of operational constraints; self-reported competency is treated as a proxy for likely training benefit, not as evidence of actual competence, and all claims are calibrated accordingly.

Positioned within performance and competency development in organizational and cooperative settings, this report condenses a larger institutional study into the Brief Research Report format to address four research questions (RQs). RQ1: What is the demographic and enterprise profile of the farmer-incubatees? RQ2: What is their self-reported competency level in management, marketing, financial, and operational domains? RQ3: Do self-reported competencies differ when respondents are grouped by profile variables, and what is the magnitude of any differences? RQ4: How do quantitative competency gaps and qualitative implementation constraints converge to inform a preliminary CBT framework?

Corresponding to RQ3, six research hypotheses were tested at the domain level:

H1: Self-reported competency differs significantly when respondents are grouped by age.

H2: Self-reported competency differs significantly when respondents are grouped by sex.

H3: Self-reported competency differs significantly when respondents are grouped by educational attainment.

H4: Self-reported competency differs significantly when respondents are grouped by business size.

H5: Self-reported competency differs significantly when respondents are grouped by ownership type.

H6: Self-reported competency differs significantly when respondents are grouped by length of business operation.

RQ1, RQ2, and RQ4 are descriptive and integrative and were therefore addressed without formal hypothesis testing.

Method

Design

A convergent mixed-methods design (Creswell and Plano Clark, 2018) was used. The quantitative component (priority strand) measured self-reported competency across four domains through a structured survey; the qualitative component explored farm-management constraints through key informant interviews. Both strands were collected during the same phase of the study. The two strands were analyzed independently and integrated at the interpretation stage through a joint display (Fetters et al., 2013; Table 1) that maps quantitative competency gaps onto qualitative constraints and derives training implications from their convergence.

Table 1. Joint display integrating quantitative competency gaps, qualitative constraints, and implications for the preliminary CBT framework.

M, mean; CBT, competency-based training; BIR, bureau of internal revenue; CDA, cooperative development authority; DTI, department of trade and industry; SEC, securities and exchange commission.

Setting and participants

Participants were farmer-incubatees of Department of Agriculture-supported agribusiness cooperatives in Region IV-A (CALABARZON, comprising the provinces of Cavite, Laguna, Batangas, Rizal, and Quezon), Philippines, enrolled in the Accelerated Research-Based Agribusiness for New and Growing Agripreneurs through Knowledgeable, Aggressive, and Dynamic Approach (ARANGKADA) Technology Business Incubation (TBI) program of the region (hereafter, the ARANGKADA TBI CALABARZON program), implemented by the Department of Agriculture Regional Field Office IV-A. A farmer-incubatee was defined as an active cooperative member-beneficiary receiving incubation support and operating a farm-based enterprise (stingless beekeeping and honey production, cacao, coconut sugar, and allied value-added lines). The program delivers assistance across five progressive stages—pre-incubation, business incubation, business readiness, business on the go, and business sustainability—covering entrepreneurial training, business and financial planning, product standardization, enterprise formalization and regulatory compliance, and sustainability and marketing arrangements, supported by shared-use facilities, product development, and market linkage. The sampling frame was the official Department of Agriculture registry of active cooperative members as of January 2025, comprising 189 eligible members across the participating cooperatives. A minimum sample of 127 was determined using the Raosoft calculator (95% confidence level, 5% margin of error, 50% response distribution) for the population of 189. Eight cooperatives and farmer organizations participated (per-cooperative allocation available from the corresponding author). Stratified random sampling (Cochran, 1977) was employed with each cooperative as a stratum, proportional allocation, and simple random sampling within strata; all 127 sampled members completed the survey (response rate = 100%). Inclusion criteria were active registry membership as of January 2025, program participation, and informed consent; inactive or unaffiliated members and incomplete questionnaires were excluded. Data were collected from May to August 2025.

Ethics

The study protocol was reviewed and approved by the St. Paul University Philippines Research Ethics Committee (approval code SPUP_2025_00196_NR). Written informed consent was obtained from all participants before data collection; participation was voluntary and responses were anonymized.

Operational definitions of variables

Consistent with competency frameworks that treat a competency as an underlying capability enabling effective performance (Boyatzis, 1982), the four conceptual variables were operationally defined as composite self-report scores on the survey instrument described in Section 2.5. Management competency was operationally defined as the mean of 14 items covering business organization and formalization, permit and licensing processes with the Bureau of Internal Revenue (BIR), Securities and Exchange Commission (SEC), Cooperative Development Authority (CDA), and Department of Trade and Industry (DTI), workplace management, strategic planning, leadership, and ethical and legal compliance. Marketing competency was defined as the mean of 8 items structured on the 7Ps of the marketing mix (product, price, place, promotion, people, process, and physical evidence) plus a global item. Financial competency was defined as the mean of 7 items covering cash management, financial reporting, budgeting, accounting, access to financial assistance, and financial planning, plus a global item. Operational competency was defined as the mean of 7 items covering process efficiency, quality management, supply chain management, production planning and control, technology utilization, and sustainability practices, plus a global item. Each composite was interpreted against the a priori bands described in Section 2.5, and a composite within or below the Basic band was treated as an indicator of training need.

Survey instrument

The survey instrument was a researcher-developed questionnaire whose content validity was established through expert validation by a panel drawn from agriculture, research and development, extension services, and higher education, with revisions made on the panel's recommendations to improve accuracy, clarity, and relevance. It comprises 36 items across the four domains operationally defined in Section 2.4 (management, 14 items; marketing, 8 items; financial, 7 items; operational, 7 items). Respondents rated each item on a four-point scale: 1 = Beginner, 2 = Basic, 3 = Proficient, 4 = Expert. Composite domain scores were computed as the mean of domain items and interpreted against a priori equal-width-style bands set before data analysis: 1.00–1.49 Beginner, 1.50–2.49 Basic, 2.50–3.49 Proficient, 3.50–4.00 Expert. These bands follow the conventional midpoint-rounding partition of a four-point scale; they were not validated against observable performance criteria, which is acknowledged as a limitation, and accordingly all scores are reported as self-reported competency.

Pilot testing with 30 comparable respondents (excluded from the main sample) yielded an overall Cronbach's alpha of 0.989, a magnitude indicating strong internal consistency but also suggesting item redundancy and possible common-method effects (Tavakol and Dennick, 2011) rather than unidimensional validity. Domain-level reliabilities in the analytical sample were management α = 0.95, marketing α = 0.95, financial α = 0.94, and operational α = 0.95.

Qualitative component

Qualitative data were collected through 16 key informant interviews with active cooperative members representing all participating cooperatives. Participants were purposively selected from among the survey respondents, with recruitment coordinated through Department of Agriculture personnel; only members who were informed of the study's purpose and voluntarily consented were interviewed. Interviews lasted 60–70 min, were conducted in Filipino, and, with consent, were transcribed verbatim and reviewed for accuracy before coding. Transcripts were analyzed thematically following Braun and Clarke (2006): familiarization, initial coding, theme development, review, definition, and reporting. Two researchers coded independently and resolved discrepancies after each interview; thematic adequacy was judged on an information-power rationale (Malterud et al., 2016). Illustrative quotations were translated into English and are attributed using identifiers that combine the participant's cooperative acronym with an interview number (e.g., MUFAI-03). The cooperatives quoted are the MountView Upland Farmers Association Inc. (MUFAI), the Mabini Unique Beekeeper Association (MUBA), the Batas Dao Agrarian Reform Beneficiaries Association Inc. (BatasDao), and the Bailen Coffee Growers Association (BCGA); participants' personal identities are withheld to preserve confidentiality.

Statistical analysis

Quantitative data were analyzed in IBM SPSS Statistics (Version 26; IBM Corp, 2019). Frequencies, percentages, means, and standard deviations described the profile and competency levels. Differences by sex were tested with independent-samples t-tests; homogeneity of variance was evaluated with Levene's test (all p ≥ 0.13; equal-variance results are reported). For age, educational attainment, ownership type, and length of business operation, group sizes were markedly unequal (e.g., 6 vs. 44 within age; 13 vs. 88 within ownership type), so Welch's heteroscedasticity-robust analysis of variance (ANOVA; Welch, 1951) was used as the primary omnibus test, followed by Games–Howell post hoc comparisons (Games and Howell, 1976) for significant effects; the single-respondent partnership category was excluded because a within-group variance could not be estimated, so the ownership-type analysis compares the four remaining arrangements. Classical one-way ANOVA results are reported in Supplementary Table S3 as a sensitivity analysis; divergences from the Welch-based decisions are addressed in Section 3.3. Business size contained degenerate cells (small, n = 2; medium, n = 1) and was retained as a classical ANOVA, reported for completeness but not interpreted. Effect sizes are reported as eta-squared (η²) and omega-squared (ω²), computed on the analytic set of each primary test (for ownership type, the four groups entering the Welch test), and Cohen's d (Cohen, 1988) for t-tests, with 95% confidence intervals for mean differences. Because 30 inferential tests were conducted (five outcomes × six grouping variables), the Benjamini–Hochberg procedure (Benjamini and Hochberg, 1995) was applied at a false discovery rate (FDR) of q = 0.05 across the full family of primary tests; both unadjusted and FDR-adjusted p-values are reported, and decisions on the six research hypotheses (H1–H6) were based on the FDR-adjusted values, with a hypothesis considered supported when the adjusted p was below 0.05. Complete descriptive and inferential outputs are provided as Supplementary material.

Results

Profile of the farmer-incubatees

The 127 respondents were predominantly older adults: 34.6% were 50–59 years old, 22.8% were 40–49, and 21.3% were 60 or older, while only 4.7% were 29 or younger. Women constituted 58.3% of the sample and men 41.7%. Educational attainment was mixed: 26.8% were high school graduates, 23.6% college graduates, 14.2% elementary graduates, 14.2% college undergraduates, 8.7% vocational/technical graduates, 7.1% high school undergraduates, and 5.5% reported other attainment. All ventures were agricultural, and nearly all (97.6%) were micro-enterprises. By ownership type, 69.3% operated as cooperative members, 12.6% as sole proprietors, 10.2% as registered Barangay Micro Business Enterprises (BMBE, a Philippine registration category under Republic Act 9178), 7.1% as corporations, and 0.8% (one respondent) as a partnership. The largest group had operated for 3–5 years (35.4%), followed by more than 5 years (26.0%), less than 1 year (22.8%), and 1–2 years (15.7%). The age distribution mirrors broader concerns about aging farm populations (Palis, 2020), while the prevalence of cooperative membership underscores the relevance of collective governance to any training response (Kalogiannidis et al., 2024).

Self-reported competency levels

All four composite means fell within the Basic band, with management lowest (M = 1.83), driven by formalization and permit items, and operational relatively highest (M = 2.20), with technology utilization its weakest item (Table 2). This pattern is consistent with reports that agro-based micro-enterprises acquire initial exposure to business concepts without sufficient practice-based reinforcement (Sarma et al., 2022).

Table 2. Self-reported competency by domain (n = 127; scale 1–4; basic = 1.50–2.49).

M, mean; SD, standard deviation; both are composites of domain scores across respondents. BMBE, barangay micro business enterprise; SEC, securities and exchange commission; DTI, department of trade and industry.

Item-level means and standard deviations for all 36 items are reported in Supplementary Table S1.

Differences by profile variables

Hypothesis-testing results are presented in the sequence H1 (age), H2 (sex), H3 (educational attainment), H4 (business size), H5 (ownership type), and H6 (length of business operation), matching the row order of Table 3; pairwise comparisons for supported hypotheses follow in Table 4.

Table 3. Differences in self-reported competency by profile variable: primary test statistics, exact p-values, FDR-adjusted p-values, and effect sizes.

Welch's heteroscedasticity-robust analysis of variance (ANOVA) is the primary omnibus test for age, educational attainment, ownership type, and length of business operation, given markedly unequal group sizes; df = degrees of freedom, and the denominator df of the Welch F are shown in parentheses. p (FDR) = Benjamini–Hochberg false discovery rate-adjusted p-value across the family of 30 primary tests.

In the decision column, sig, statistically significant after FDR adjustment (adjusted p <0.05, q = 0.05), indicating support for the corresponding hypothesis, and ns, not significant.

Overall = mean of the four domain composites.

η² and ω² are computed on the analytic set of each primary test; for ownership type this is the four groups entering the Welch test (partnership, n = 1, excluded because a within-group variance could not be estimated).

For sex, values in brackets are 95% confidence intervals (CIs) for Cohen's d; group means (standard deviations, SDs) were 1.80 (0.60) vs. 1.85 (0.58) for management, 1.97 (0.60) vs. 2.07 (0.70) for marketing, 2.01 (0.63) vs. 2.13 (0.67) for financial, 2.24 (0.66) vs. 2.18 (0.69) for operational, and 1.96 (0.56) vs. 2.02 (0.60) overall (male, n = 53; female, n = 74); Levene's tests were non-significant (all p ≥ 0.13). Business-size results involve cells of n = 2 and n = 1 and are shown for completeness only.

Classical one-way ANOVA results for all factors are reported in Supplementary Table S3 as a sensitivity analysis; classical decisions were more conservative for several age and length-of-operation contrasts (see Section 3.3). Group sizes, means, and SDs for all strata appear in Supplementary Table S2.

Table 4. Significant Games–Howell pairwise comparisons for FDR-significant omnibus effects.

Games–Howell comparisons do not assume equal variances and use Welch-adjusted degrees of freedom per pair; they were computed for all FDR-significant omnibus effects. Full pairwise matrices are reported in Supplementary Table S4. The single-respondent partnership category was excluded because a within-group variance could not be estimated.

H1 (age) was supported in every domain after FDR correction, with small-to-moderate effects (η² = 0.07–0.10); however, the classical ANOVA sensitivity analysis (Supplementary Table S3) reached more conservative decisions (management, marketing, and operational competency non-significant after correction), and because the age pattern rests heavily on the very small youngest cell (n = 6), the age findings should be interpreted cautiously despite surviving correction. H2 (sex) was not supported in any domain, and effect sizes were trivial (|d| = 0.09–0.18); however, because the 95% confidence intervals for d span values up to about ±0.5, these results do not establish equivalence between men and women, only the absence of detectable differences at this sample size. H3 (educational attainment) was supported in every domain, with the largest effects in the study (η² = 0.21–0.34, i.e., roughly a fifth to a third of variance). H4 (business size) was not supported in any domain; because this factor contained degenerate cells (n = 2 and n = 1), the result is reported for completeness but not interpreted. H5 (ownership type) was supported in every domain, with moderate effects (η² = 0.10–0.15). H6 (length of business operation) was supported in every domain except financial (η² =0.06–0.09); the classical sensitivity analysis was again more conservative (marketing and operational competency non-significant after correction), a divergence that reflects the Welch test's greater efficiency when smaller groups have smaller variances.

Qualitative findings

Thematic analysis identified six recurring constraints (quotations translated from Filipino). First, farming was perceived as unattractive to younger people, raising succession concerns: “Honestly, few young people want to farm anymore. They would rather work in an office or abroad than toil in the fields” (MUFAI-03). Second, regulatory ambiguity was described as a barrier to growth in stingless beekeeping: “We want to expand our beekeeping, but it is not clear to us which permits are required” (MUBA-01). Third, honey quality and packaging varied because processing was individual rather than collective: “We know consistent quality and proper packaging matter, but we still lack the equipment and guidance to achieve them” (MUBA-02). Fourth, natural calamities and pest pressure disrupted production stability: “Our biggest challenge is the changing weather, because it directly affects the health of the bees and our livelihood” (MUBA-04). Fifth, low member participation reduced cooperative efficiency: “Low member participation in meetings and trainings hinders proper coordination, accountability, and the development of the cooperative” (BatasDao-01). Sixth, equipment gaps and transport problems constrained cacao, coconut sugar, and other value-added lines: “The lack of proper transportation, drying facilities, and processing equipment causes major post-harvest losses, lowers product quality, and keeps farmers' incomes down” (BCGA-04). These experiential accounts echo documented struggles of Philippine agrarian smallholders (Cadusale, 2023).

Mixed-methods integration

The joint display shows predominantly convergent evidence, with the lowest-rated survey items corresponding closely to the narrated constraints; the qualitative strand also surfaced participation, succession, and climate-related stressors that a competency survey cannot detect, motivating the cross-cutting design principles rather than an additional content module.

Discussion

Farmer-incubatees in these cooperatives rated themselves at a Basic level across management, marketing, financial, and operational domains. Because respondents were active micro-enterprise operators rather than prospective trainees, self-assessed basic competency at this stage suggests that support should move beyond awareness-building toward coached, practice-based application, in line with capacity-building models that pair training with commercialization support (Padilla et al., 2020).

The inferential results, discussed here in the hypothesis-testing sequence, indicate that self-reported competency was associated with structural characteristics rather than with sex. Consistent with the support for H1, the youngest incubatees ( ≤ 29 years) rated themselves lowest across domains, although this age pattern rests on six respondents and should be read with corresponding caution. H2 was not supported: with trivial point estimates, no male–female competency gap was detectable in this sample. This should not be read as evidence that the training environment is gender-inclusive: no equivalence framework was applied, and inclusivity concerns access, participation, and voice rather than mean self-ratings (Ani and Casasola, 2020). Practically, the data provide no basis for separate male and female curricula, while gender-responsive delivery remains a design consideration on other grounds.

For H3, educational attainment showed the largest associations (η² = 0.21–0.34), with college graduates rating themselves substantially higher than respondents with elementary or high school attainment; business size (H4) yielded no interpretable differences given degenerate cells; for H5, ownership type showed moderate associations (η² = 0.10–0.15), with BMBE registrants consistently lowest and cooperative members higher across domains; and for H6, incubatees in their second year of operation rated themselves lower than both newer and more established peers across most domains. These are cross-sectional associations among self-reports and do not establish that age, education, ownership arrangement, or business maturity cause competency differences; selection into ownership categories and education-related response styles are plausible alternative explanations. Nevertheless, the magnitude of the education and ownership effects supports differentiation: a uniform package would pitch content too low for college-educated cooperative members and too high for BMBE registrants and less formally educated incubatees, and it is the size of these differences, not significance alone, that grounds tiered delivery.

Integrating the two strands, the proposed response is a preliminary evidence-informed CBT framework rather than a fully specified curriculum: four modules—management, marketing, financial, and operational, with the contents summarized in Table 1—delivered through experiential tasks, mentoring, and cooperative-based peer learning, and tiered by educational background and enterprise arrangement. Full specification of outcomes, performance criteria, assessment methods, and sequencing is the subject of the parent project's training-manual phase and remains to be validated against observed performance.

This report has limitations. Competencies were self-reported and not benchmarked against required-performance standards; the design was cross-sectional and confined to a single regional incubation program; several grouping cells were very small (notably the single partnership respondent and the six respondents aged 29 or below), making the age results sensitive to the choice of omnibus test; the near-ceiling pilot alpha suggests item redundancy; and the qualitative strand was condensed for the brief format. Future work should validate the competency classifications against observable performance, apply a required-proficiency gap analysis, and evaluate whether the tiered framework improves measured competence and enterprise outcomes over time.

Generative AI statement

The author(s) declared that Generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher's note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/forgp.2026.1861460/full#supplementary-material

Supplementary material Supplementary material at the source

Footnotes

Edited by: M. B. Pasco, San Beda University, Philippines

Reviewed by: Carlos Alberto Echeverría Mayorga, Universidad Alberto Masferrer, El Salvador

Krishna Alejandrino, JA Philippines, Philippines