Methods: Time series analysis and panel models were used to analyze the effect of predictors on member engagement, over time. Cross sectional analyses of member engagement in Mexican and US coalitions were also conducted. Multinomial logistic regression was used to evaluate participant roles. A dynamically complete time series analysis was conducted.
Results: There were 3061 participants in the U.S. coalitions, and 210 participants in Mexican coalitions. A dot plot indicates that member engagement has increased over time, from spring to fall 2013, in Mexican based coalitions. Results from the cross sectional analysis of Mexican coalitions indicated that participant perceived benefit has a strong and significant impact on time invested (p<0.05), as well as perceived cohesion (p<0.05). In general, the coalition’s use of a participatory leadership style was associated with members taking on leadership roles in the coalition. Time series analysis showed statistically significant results on the impact of efficiency (p<0.05), participatory leadership (p<0.001), benefits (p<0.01), perceived difficulties (p<0.01), and cohesion (p<0.1) on participant role, in U.S. coalitions. The time series of the Mexican coalitions show similar results to the cross sectional analyses. Participatory leadership (p<0.1) and cohesion (p<0.1) had a strong impact on time invested in the coalition, while cohesion and participatory leadership impacts participant role. When lagged variables are included, the marginal effect is more statistically significant in U.S. coalitions.
Conclusions: This study evaluated theoretically relevant predictors of member engagement in community coalitions over time. Differences in member engagement in the U.S. and Mexico are compared. These results inform practice, as the predictors that impact member engagement are identified. Practitioners will be able to accordingly develop culturally competent coalition practices and adjust activities to maximize engagement.