Moscarelli, Michael Benjamin (2023-05). Perception of Knowledge Retention and Behavior Change: A Function of Adult Educator-Learner Homophily. Doctoral Dissertation.
Thesis
Homophily is the propensity of individuals to be drawn to and associate with people who share similar sociodemographic, behavioral, and intrapersonal characteristics. Most scholars study homophily in the context of peer-to-peer interactions and influence in a social system. However, Rogers (1962) posited that educator-learner homophily plays a significant role when a change agent - often a non-peer individual from outside the community - introduces an innovation to a population. To date, minimal evidence supports Rogers' assertion. This dissertation, written in a three-article format, addresses this knowledge gap. The first article, an integrative literature review, synthesized 19 peer-reviewed articles from 2013 to 2019 to generate new knowledge on educator-learner homophily's effect on learner perception, attitude, learning outcomes, and behavior change. The review revealed limited research on the effect of educator-learner homophily. Nonetheless, I developed a conceptual model, the Moscarelli Model, to describe the relationship between educator-learner homophily, learner attitude, knowledge retention, and behavior change. The second article, a qualitative case study, examined the influence of educator-learner homophily on Dominican and Guatemalan small-scale farmers that receive technical assistance from U.S. volunteers. Contrary to the Moscarelli Model, farmers indicated that U.S. volunteers were more credible and reliable than local extension agents, despite seemingly lower educator-learner homophily between small-scale farmers and U.S. volunteers. The third article, a quasi-experimental study conducted in Huehuetenango, Guatemala, with 238 female small-scale farmers, tested the effect of educator-learner gender homophily and geographic origin homophily on behavior change. I used a binary logistic regression to ascertain the effects of a participant's age, literacy, and residence, combined with the training instructor's gender and nationality (a proxy for geographic origin), to determine participants' likelihoods of adopting recordkeeping logs to track costs and revenues at six- and 21 weeks post-training. The regression model measuring adoption at 21 weeks was statistically significant, ?2 (7, N = 220) = 17.93, p = .012. Participants receiving the training from a female instructor were 0.441 times as likely (i.e., less likely) to adopt recordkeeping logs. The trainer's nationality had no statistically significant effect on the likelihood of adopting recordkeeping logs.
