Sample image

Current Projects

 

Bayesian inference of age and length at 50% maturity

Jason C. Doll and Thomas E. Lauer, submitted to Transactions of the American Fisheries Society

We used logistic regression to estimate age and length at 50% maturity of yellow perch from southern Lake Michigan. Parameters were estimated using frequentist and Bayesian methods. Frequentist inference converged for only 76% of the year classes due to small sample size while Bayesian inference provided estiamtes for all year classes. We suggest that Bayesian inference of age and length at 50% maturity is an improvement over frequentist methods by providing more information about the complete distribution of parameters and are not always limited to large sample sizes.

Figure 1 Maturity
Figure 1. Female and male age at 50% maturity estimates by year class from model 3. Frequentist inference results of females (A) and males (C) display the mean estimated age at 50% maturity (black dots) and 95% confidence intervals (upper and lower lines). Bayesian inference results for females (B) and males (D) display the median estimated age at 50% maturity (black dots) and 95% credible intervals (upper and lower lines).