Other than a formal goodness-of-fit test, one analysis that provides a sense of whether a particular GLM is plausible is the following: Suppose the ML fitted equation were the true equation. At the observed x values for the n observations, randomly generate n variates with distributions specified by the fitted GLM. Construct scatterplots. Do they look like the scatterplots that were actually observed? Do this for a Poisson loglinear model for the horseshoe crab data, with y = number of satellites and x = width. Does the variability about the fit resemble that in the actual data, including a similar number of 0’s and large values? Repeat this a few times to get a better sense of how the scatterplot observed differs from what you would observe if the Poisson GLM truly held.