Fishery Bulletin 100(1) 



in the form of expert opinion (e.g. Demissie et al., 1998) 

 or in the form of results of earher analyses (e.g. Viana 

 et al., 1993). Rather than assuming that estimated accu- 

 racies are "known," one can incorporate the uncertainty 

 in the estimates into the prior distributions. In addition, 

 the Bayesian approach does not rely on asymptotic results 

 that may behave poorly with small samples. We have also 

 not assessed the possible bias due to the lack of indepen- 

 dence in the readings. When suitable software becomes 

 available, this assumption should be checked. 



In our examples above, misclassification error contribut- 

 ed relatively little to the overall uncertainty. In these ap- 

 plications, where estimates of hatchery contribution were 

 used to make management decisions, the accuracy of read- 

 ings were within an acceptable range. However, the criteria 

 used to establish quality control standards in any program 

 need to be developed in the context of how the information 

 is to be used along with other sources of uncertainty. 



In conclusion, we believe that the use of agreement mea- 

 sures in combination with latent class models can con- 

 tribute significant information about both the proportions 

 of interest and the quality control aspects of an otolith- 

 marking program. Furthermore these approaches could 

 have application to similar areas in fisheries which re- 

 quire judgments that are not free of error. 



Acknowledgments 



We thank Bob Wilbur for editorial comments and three 

 anonymous reviewers for valuable suggestions. 



Literature cited 



Agresti, A. 



1990. Categorical data analysis. John Wiley, New York, 

 NY, 576 p. 

 Albert. P. S., L. M. McShane, and J. H. Shih. 



2001. Latent class modeling approached for assessing diag- 

 nostic error without a gold standard: with applications to 

 p53 immunohistochemical assays in bladder tumors. Bio- 

 metrics 57:610-619. 

 Bartholomew, D. J. 



1987. Latent variable models and factor analysis. Oxford 

 Univ. Press, New York. NY'. 427 p. 

 Christen.sen A. H., T. Gjorup, J. Hilden, C. Fenger. B. Henriksen, 

 M. Vyberg, K. Ostergaard, and B. F. Hansen. 



1992. Observer homogeneity in the histologic diagnosis of 

 Helicobacter pylori: latent class analysis, kappa coefficient, 

 and repeat frequencv Scand. J. Gastroenterol. 27:933-939. 

 Clogg, C. C. 



1995. Latent class models. Chapter 6 ;;; Handbook of sta- 

 tistical modeling for the social and behavioral sciences (G. 

 Arminger, C. C. Clogg, and M. E. Sobel, eds.), p. 311-359. 

 Plenum Press, New York, NY. 

 Croon, M. 



1990. Latent class analysis with ordered classes. Brit. J. 

 Math. Stat. Psych. 43:171-192. 

 Dawid, A. P., and A. M. Skene. 



1979. Maximum likelihood estimation of observer error- 

 rates using the EM algorithm. Appl. Statist. 28:20-28. 



Demissie, K., N. White, L. Joseph, and P. Ernst. 



1998. Bayesian estimation of asthma prevalence, and com- 

 parison of exercise and questionnaire diagnostics in the 

 absence of a gold standard. Ann. Epidemiol. 8:201-208. 

 Dempster, A.P., N.M. Laird, and D.B. Rubin. 



1977. Maximum likelihood from incomplete data via the 

 EM algorithm (with discussion). J. Royal Stat. Soc. B 39: 

 1-38. 

 Everitt, B. S. 



1984. An introduction to latent variable models. Chapman 

 and Hall. London. 107 p. 

 Fleiss, J. L. 



1981. Statistical methods for rates and proportions, 2"'' ed. 

 John Wiley, New York, NY, 352 p 

 Formann, A. K. 



1994. Measurement errors in caries diagnosis: some further 

 latent class models. Biometrics 50:865-871. 



1996. Latent class analysis in medical research. Stat. 

 Meth. Med. Res. 5:179-211. 



Hagen, P., K. Munk, B. Van Alen, and B. White. 



1995. Thermal mark technology for inseason fisheries man- 

 agement: a case study Alaska Fishery Res. Bull. 2:14.3- 

 158. 



Hui.S.L, and S.D.Walter 



1980. Estimating the error rates of diagnostic tests. Bio- 

 metrics 36:167-171. 

 Joseph, L., T. Gyorkos, and L. Coupal. 



1995. Bayesian estimation of disease prevalence and the 

 parameters of diagnostic tests in the absence of a gold 

 standard. Am. J. Epidemiol. 141:263-72. 

 Kimura, D. K.. and S. Chikuni. 



1987. Mixtures of empirical distributions: an iterative appli- 

 cation of the age-length key. Biometrics 43:23-35. 

 Ki-zanowski, W. J., and F. H. C. Marriott. 



1995. Multivariate analysis, part 2: classification, cova- 

 riance structures and repeated measurements. Arnold, 

 London, 280 p. 



McCutcheon, A. L. 



1987. Latent class analysis. Sage, Beverly Hills, CA, 96 p. 

 McLachlan, G. J., and T. Ki'ishnan. 



1997. The EM algorithm and extensions. John Wiley, New 

 York, NY, 304 p. 



Microsoft Corporation. 



1993. Microsoft Excel user's guide. Microsoft Corporation, 

 Redmond, WA. 

 Millar, R. B. 



1987. Maximum likelihood estimation of mixed stock fish- 

 ery composition. Can. J. Fish. Aquat. Sci. 44:583-590. 

 Munk, K. M., W W. Smoker, D. R. Beard, and R. W. Mattson. 



1993. A hatchery water-heating system and its application 

 to 100'7f thermal marking of incubating salmon. Progi'es- 

 sive Fish-Culturist 55:284-288. 

 Fella, J., M. Masuda, and S. Nelson. 



1996. Search algorithms for computing stock composition 

 of a mixture from traits of individuals by maximum like- 

 lihood. U.S. Dep. Commerce, NOAA Tech. Memo. NMFS- 

 AFSC-61. 



Qu, Y., M. Tan, and M.H. Kutner 



1996. Random effects models in latent class analysis for 



evaluating accuracy of diagnostic tests. Biometrics 52: 



797-810. 

 Qu.Y.andA. Hagdu. 



1998. A model for evaluating sensitivity and specificity 

 for correlated diagnostic tests in efficacy studies with an 

 imperfect reference test. J. Am. Stat. Assoc. 93:920-928. 



