Trust dynamics between extension agents and farming communities: Implications for advisory credibility

Introduction

Consider a farmer in rural Guinea who has been told by three different extension agents over five years to adopt three different soil management practices each presented as the best approach, each contradicting the last. Would that farmer trust the fourth agent who arrives with yet another recommendation? This kind of credibility erosion happens quietly across West Africa's extension systems, and it may explain why technically sound advice sometimes fails to translate into farmer action [1] . Trust between extension agents and farming communities has been described as the invisible infrastructure of advisory services [2] . Without it, even the best-designed programs struggle to gain traction. With it, communities are more willing to experiment with unfamiliar technologies, share honest feedback about their challenges, and invest their own resources in recommended practices [3] . Yet trust remains one of the least measured aspects of extension performance monitoring systems track visit counts, training attendance, and adoption rates, but rarely assess whether farmers actually believe what agents tell them [4] . Guinea's extension service faces particular trust challenges. Chronic understaffing means agents are rotated frequently, preventing sustained relationships that build credibility [5] . Language barriers between agents trained in French and communities speaking Susu, Pular, and Malinke add difficulty [6] . Previous assessments noted low adoption rates without adequately investigating whether trust deficits might be a root cause [7] . This research examined the nature and determinants of trust in Guinea's Kindia and Boke regions, and assessed how trust levels relate to advisory credibility and technology adoption [8] .

Instrument Development and Validation

The trust instrument was adapted from Mayer, Davis, and Schoorman's organizational trust model [9] for the extension context. An initial 30-item pool covered five dimensions: technical competence, reliability, benevolence, integrity, and openness. Items were reviewed by four extension specialists and two rural sociologists for content validity. Following pilot testing with 28 farmers outside the sample, the instrument was refined to 25 items with satisfactory reliability (overall α=0.91; subscale α from 0.82 to 0.89). All items used a 5-point Likert scale with descriptors in both French and the local language [10] .

Material and Methods

Material

This cross-sectional research was conducted from April to November 2021 in four prefectures across the Kindia and Boke regions of Guinea. The research was based at the Faculty of Agriculture, University of Conakry. A total of 256 farming households were surveyed using a structured questionnaire administered through face-to-face interviews in local languages (Susu and Pular). Additionally, 48 extension agents serving these communities completed a complementary questionnaire on their service delivery practices and professional background. Households were selected through multistage cluster sampling: two prefectures per region, three sub-prefectures per prefecture, www.extensionjournal.com and approximately 21 households per sub-prefecture [11] . Technology adoption was measured for five recommended practices: improved rice varieties, composting techniques, row planting, integrated pest management, and post-harvest drying methods.

Methods

Trust scores were calculated as the mean of five dimension subscale scores. K-means cluster analysis classified communities into three trust categories based on aggregate trust profiles [12] . Advisory credibility was measured through a six-item scale assessing perceived information accuracy, agent responsiveness, cultural sensitivity, follow-up consistency, conflict resolution ability, and local knowledge. One-way ANOVA compared adoption rates across trust categories. Multiple regression modeled trust as a function of agent tenure, visit frequency, language match, agent gender match with community, and prior positive experience. All interviews were conducted by trained enumerators fluent in local languages. Ethical clearance was obtained from the University of Conakry Research Ethics Board (Protocol UC-REB/2021/023, March 12, 2021) [13] .

Ethical Considerations

Informed consent was obtained verbally in local languages due to low literacy rates among participants; consent procedures were audio-recorded with participant permission. Participants were assured that individual responses would remain confidential and would not be shared with extension agencies in a way that could identify specific agents or households. Agent participants were informed that survey results would be reported at the aggregate level only. The research team included native speakers of Susu and Pular to ensure accurate communication. No incentives were offered for participation [13] . High-trust communities showed adoption rates 2.8 times higher than low-trust communities. All trust dimensions differed significantly across categories (p<0.001).

Results

Benevolence showed the widest gap between high and low trust groups (difference of 2.09 points), while integrity showed the smallest (1.55 points). Agent tenure was the strongest trust predictor (β=0.43), followed by visit frequency (β=0.31) and language match (β=0.22). Gender match between agent and community was not significant. The model explained 46.8% of variance in trust scores.

Discussion

The 2.8-fold adoption difference between high and low-trust communities provides clear evidence that trust is not a peripheral factor but a central determinant of whether advice translates into action [2] . This is consistent with health communication research showing source credibility mediates persuasive impact regardless of message quality [14] . An agent can deliver perfectly accurate information, but if the farmer doesn't trust the messenger, the message doesn't land. Agent tenure as the strongest trust predictor aligns with social exchange theory trust accumulates through repeated positive interactions over time [9] . Guinea's practice of frequently rotating agents works directly against this [5] . The significance of language match (β=0.22) points to a solvable barrier through recruitment policies favoring local language speakers [6] . The radar chart reveals senior agents' credibility advantage is most pronounced in local knowledge and cultural sensitivity dimensions requiring extended community immersion rather than formal training. Junior agents scored well on information accuracy, but lag on relational dimensions, suggesting mentoring programs that accelerate community integration would be more effective than additional technical coursework [15,16] .

Conclusions

This research demonstrates that trust between extension agents and farming communities has a measurable and substantial effect on technology adoption in Guinea. Hightrust communities adopted technologies at nearly three times the rate of low-trust communities. Agent tenure and visit frequency were the strongest trust predictors, suggesting staffing stability and regular contact matter more than any single programmatic intervention. Advisory credibility is built through sustained presence, linguistic accessibility, and demonstrated concern for farmer welfare. Extension managers should treat trust as a trackable performance indicator and design posting policies supporting long-term agent-community relationships [17,18] .

References

1. Feder G, Anderson JR, Birner R, Deininger K. Promises and realities of community-based agricultural extension. IFPRI Discuss Pap 00959. 2010:1-27.
2. Swanson BE, Rajalahti R. Strengthening agricultural extension and advisory systems. Washington (DC): World Bank; 2010.
3. Pannell DJ, Marshall GR, Barr N, Curtis A, Vanclay F, Wilkinson R. Understanding and promoting adoption of conservation practices by rural landholders. Aust J Exp Agric. 2006;46(11):1407-1424.
4. Davis K, Swanson B. New directions for agricultural extension. In: Christoplos I, editor. Mobilizing the potential of rural extension. Rome: FAO; 2010. p. 1-17.
5. Ministry of Agriculture Guinea. National agricultural extension strategy review. Conakry: MAG; 2019.
6. Diallo MC, Barry AB. Language barriers in agricultural service delivery in Guinea. West Afr J Appl Ecol. 2018;26(2):87-98.
7. Fofana I, Goundan A. Agricultural sector performance in Guinea: an assessment. IFPRI Country Rep. 2016:142.
8. Leeuwis C. Communication for rural innovation. 3rd ed. Oxford: Blackwell Science; 2004.
9. Mayer RC, Davis JH, Schoorman FD. An integrative model of organizational trust. Acad Manage Rev. 1995;20(3):709-734.
10. DeVellis RF. Scale development: theory and applications. 4th ed. Thousand Oaks (CA): SAGE; 2017.
11. Cochran WG. Sampling techniques. 3rd ed. New York: John Wiley; 1977.
12. Hair JF, Black WC, Babin BJ, Anderson RE. Multivariate data analysis. 8th ed. Andover: Cengage; 2019.
13. World Medical Association. Declaration of Helsinki. JAMA. 2013;310(20):2191-2194.
14. Hovland CI, Janis IL, Kelley HH. Communication and persuasion. New Haven (CT): Yale University Press; www.extensionjournal.com 1953.
15. Rogers EM. Diffusion of innovations. 5th ed. New York: Free Press; 2003.
16. Oakley P, Garforth C. Guide to extension training. Rome: FAO; 1985.
17. Christoplos I. Mobilizing the potential of rural and agricultural extension. Rome: FAO; 2010.
18. Birner R, Davis K, Pender J, Nkonya E, Anandajayasekeram P, Ekboir J, et al. From best practice to best fit. J Agric Educ Ext. 2009;15(4):341355.