Ralston et at: A meta-analytic approach to quantifying scientific uncertainty in stock assessments 
225 
Year Year 
Figure 2 (continued) 
of data-rich stock assessments that have been con- 
ducted for the PFMC. Although the approach lacks 
a theoretical basis, the method incorporates many of 
the factors that lead to model uncertainty, which we 
have shown is much greater than within-model esti- 
mation errors. One concern with the analysis is that 
calculation of uncertainty as squared deviations from 
the mean (approach 2) includes the assumption of the 
independence of the residuals, which is surely violated 
given that repeats of an assessment provide much the 
same data. Likewise, our findings pertain strictly to 
groundfish and coastal pelagic species found off the 
U.S. west coast. To the extent that the availability of 
data and the assessment “culture” in that region is 
distinctive (e.g., the use of the Stock Synthesis model- 
ing platform and a willingness to adopt meta-analytic 
results as proxy metrics), our specific findings may 
not be of general use elsewhere. Even so, our general 
