Abstract
Scientific buffalo husbandry management practices play a pivotal role in enhancing rural livelihoods and milk productivity in India. Despite widespread training initiatives, there is limited empirical evidence quantifying the cognitive impact of such programs on smallholder farmers. This study, grounded in Adult Learning Theory (Andragogy), adopted a quasi-experimental one-group pre-test-post-test design involving 518 farmers trained at the Indian Council of Agricultural Research-Central Institute for Research on Buffaloes (ICAR-CIRB). A validated 15-item knowledge test was used to assess learning across five thematic domains: breeding, feeding and nutrition, animal health, milk quality, and milk marketing. The training intervention resulted in a substantial improvement in farmers' knowledge, as evidenced by both absolute and normalized learning gains. Subject-wise analysis revealed feeding and nutrition as the most improved domain, followed by milk marketing and breeding. Multiple linear regression analysis identified specific socio-demographic and experiential factors-such as gender, livestock holding, and prior experience-as significant predictors of knowledge gain, while age and baseline knowledge showed negative associations. Disaggregated profiling further showed that young, moderately experienced female farmers with medium-to-large herd sizes achieved the highest learning gains. These findings suggest that training outcomes vary considerably across learner profiles and subject areas. The study highlights the importance of targeted, learner-sensitive extension models not only for enhancing knowledge transfer, but also for promoting the adoption of best practices, and for enabling systematic monitoring and evaluation of how the trained subject matter areas are implemented at the field level.