Pluralistic Extension-Service Exposure and Perceived Knowledge, Skills, Adoption and Livelihood Outcomes among Smallholder Farmers in Liberia

Abstract

Background: Pluralistic extension may widen farmers’ learning opportunities, but cross-sectional associations between exposure and outcomes can be distorted by selection, geography and common-method reporting. This study examined the relationship between named-provider exposure and perceived knowledge, attitude, skills, adoption and livelihood-improvement breadth among smallholders in five districts of Bong County, Liberia. Methods: The verified analytical sample contained 232 respondents. Exposure was the count of seven named providers from which services were reportedly received. Outcomes were respondent-level percentages of 37 coded items endorsed as improved across four parallel dimensions. Ordinary least-squares models with HC3 robust standard errors adjusted for age, farming experience, education, farmer-organization membership, bank-account ownership and district. Results: Mean scores rose from approximately 36 among respondents reporting no provider to approximately 50 with one provider and 91 with two to three providers, before plateauing near 86 with four or more. In adjusted models (n = 230), each additional provider was associated with 6.29 points higher knowledge breadth (95% CI [4.68, 7.90]), 6.31 higher attitude breadth (95% CI [4.69, 7.92]), 6.21 higher skills breadth (95% CI [4.60, 7.82]) and 6.30 higher adoption breadth (95% CI [4.67, 7.94]); all p < .001. Results persisted after excluding BRAC-Liberia and when exposure was represented as any-provider contact. Conclusions: Broader provider exposure was consistently associated with broader perceived improvement, particularly in human-capital and adoption-related domains. The findings do not demonstrate causal impact because exposure and outcomes were measured simultaneously and by self-report. Longitudinal, practice-labelled and independently measured outcomes are required.