Figure 3
From: Interactions with humans shape coyote responses to hazing
Effect of different treatments on the proportion of time coyotes spent avoiding the tester in adult (a) and child (b) trials. Bars and error values are produced from the output of a generalized linear mixed effect model, and take into account other significant fixed factors (i.e., cumulative previous hazings and number of times hazed per day) and random effects (coyote pair ID). Error bars denote standard errors. The same letters above bars denote treatments that are not significantly different from each other.
