Data Weighting

Research Summary

Mark and his collaborators have been involved in extensive research into data weighting. In 2015 he co-organized and chaired the CAPAM workshop on data weighting and acted as a guest editor for the Fisheries Research special issue (Maunder et al., 2017). He has developed several approaches to automatically weight data including indices of abundance (Deriso et al., 2007; Maunder and Starr, 2003) and composition data (Maunder, 2011), and the associated issue of estimating the variance of process error (Maunder and Deriso, 2003). He and his collaborators developed the “Law of conflicting data” that states that data conflict implies model misspecification, it should be evaluated in the context of random sampling error and have shown that data weighting may not appropriately deal with model misspecification (Maunder and Piner, 2017). Their research has emphasized the problems with using composition data, particularly length composition data, and the impact of model misspecification (Wang et al., 2009; Wang et al. 2015; Maunder and Piner, 2015; Minte-Vera et al. 2017) and recommend fixing the model not the symptoms (Wang and Maunder, 2017). They have also highlighted that inter-annual weighting for the same data set is important (Hyun et al., 2015) in addition to inter-data-set weighting and overall weighting. They have shown that model misspecification and unmodelled process variation manifests themselves in data weighting issues and have been involved in developing methods to diagnose data conflict and model misspecification including the likelihood component profile (Maunder and Starr, 2001; Lee et al., 2014; Wang et al., 2014), the age structure production model (ASPM) diagnostic (Maunder and Piner 2015; Minte-Vera et al., 2021), and the catch-curve diagnostic (Carvalho et al., 2017; Minte-Vera et al., 2021). Mark argues that it may be best to base data weighting on sampling variation and any divergence from this indicates model misspecification or unmodelled process variation, and these should be eliminated as much as possible (see Maunder and Piner 2017). They have also shown the importance of data weighting and modelling process error when conducting covariate selection (Deriso et al., 2007; Maunder et al., 2015) and more complex modelling of observation error including outliers (e.g. Maunder, 2002).  Mark has shown how integrating “raw” data into stock assessment models can change the data weighting and is likely due to the automatic accounting of correlation, which is generally not considered when using a two-step approach (Maunder, 2001).       

Relevant Papers

Minte-Vera, C.V., Maunder, M.N., Aires-da-Silva, A.M. 2021. Auxiliary diagnostic analyses used to detect model misspecification and highlight potential solutions in stock assessments: application to yellowfin tuna in the eastern Pacific Ocean. ICES Journal of Marine Science 78 (10), 3521-3537. https://academic.oup.com/icesjms/article/78/10/3521/6430633?login=true

Maunder, M. N. and Piner, K. R. 2017. Dealing with data conflicts in statistical inference of population assessment models that integrate information from multiple diverse data sets. Fisheries Research. 192: 16-27. https://www.sciencedirect.com/science/article/abs/pii/S0165783616301394

Maunder, M.N., Crone, P.R, Punt, A.E., Valero, J.L., Semmens B. X. 2017. Data conflict and weighting, likelihood functions and process error. Fisheries Research. 192: 1-4. https://www.sciencedirect.com/science/article/abs/pii/S0165783617300735

Carvalho, F., Punt, A. E., Chang, Y. J., Maunder, M. N., Piner, K. R. 2017. Can diagnostic tests help identify model misspecification in integrated stock assessments? Fisheries Research. 192: 28-40. https://www.sciencedirect.com/science/article/pii/S0165783616303113

Minte-Vera, C. V., Maunder, M. N., Aires-da-Silva, A. M., Satoh, K., Uosaki, K. Get the biology right, or use size-composition data at your own risk. 2017. Fisheries research. 192: 114-125. https://www.sciencedirect.com/science/article/abs/pii/S0165783617300231

Wang, S-P and Maunder, M. N. 2017. Is down-weighting composition data adequate for dealing with model misspecification, or do we need to fix the model? Fisheries Research. 192: 41-51. https://www.sciencedirect.com/science/article/abs/pii/S0165783616304143

Hyun, S.Y., Maunder, M.N., Rothschild, B.J. 2015. Importance of modeling heteroscedasticity of survey index data in fishery stock assessments. ICES Journal of Marine Science. 72 (1): 130-136. https://academic.oup.com/icesjms/article/72/1/130/823326

Maunder, M.N., Deriso, R.B., and Hanson, C.H. 2015. Use of state-space population dynamics models in hypothesis testing: advantages over simple log-linear regressions for modeling survival, illustrated with application to longfin smelt (Spirinchus thaleichthys). Fisheries Research, 164: 102–111. https://www.sciencedirect.com/science/article/pii/S0165783614003105

Maunder, M.N., Piner, K.R. 2015. Contemporary fisheries stock assessment: many issues still remain. ICES Journal of Marine Science. 72 (1): 7-18. https://academic.oup.com/icesjms/article/72/1/7/819352

Wang, S. P., Maunder, M. N., Nishida, T., Chen, Y. R. 2015.  Influence of model misspecification, temporal changes, and data weighting in stock assessment models: Application to swordfish (Xiphias gladius) in the Indian Ocean. Fisheries Research, 166: 119-128. https://www.sciencedirect.com/science/article/abs/pii/S0165783614002483

Wang, S. P., Maunder, M. N., Piner, K. R., Aires-da-Silva, A. Lee, H. H. 2014. Evaluation of virgin recruitment profiling as a diagnostic for selectivity curve structure in integrated stock assessment models. Fisheries Research, 158: 158-164. https://www.sciencedirect.com/science/article/abs/pii/S0165783613003032

Lee, H. H., Piner, K. R., Methot, R. D., Maunder, M. N. 2014. Use of likelihood profiling over a global scaling parameter to structure the population dynamics model: An example using blue marlin in the Pacific Ocean. Fisheries Research, 158: 138-146. https://www.sciencedirect.com/science/article/abs/pii/S0165783613003111

Maunder, M.N. 2011. Review and evaluation of likelihood functions for composition data in stock-assessment models: Estimating the effective sample size. Fisheries Research, 109: 311–319. https://www.sciencedirect.com/science/article/abs/pii/S0165783611000890

Wang, S-.P., Maunder, M.N., and Aires-da-Silva, A. 2009. Implications of model and data assumptions: An illustration including data for the Taiwanese longline fishery into the eastern Pacific Ocean bigeye tuna (Thunnus obesus) stock assessment. Fisheries Research 97 (2009) 118–126. https://www.sciencedirect.com/science/article/abs/pii/S0165783609000344

Deriso, R.B., Maunder, M.N., and Skalski, J.R. 2007. Variance estimation in integrated assessment models and its importance for hypothesis testing. Can. J. Fish. Aquat. Sci. 64: 187-197. https://www.nrcresearchpress.com/doi/abs/10.1139/f06-178#.Xp4nR8hKjD4

Maunder, M.N. and Deriso, R.B. 2003. Estimation of recruitment in catch-at-age models. Can. J. Fish. Aquat. Sci. 60: 1204-1216. https://www.nrcresearchpress.com/doi/abs/10.1139/f03-104#.Xp40ZshKjD4

Maunder, M.N. and Starr, P.J. 2003. Fitting fisheries models to standardised CPUE abundance indices. Fisheries Research 63: 43-50. https://www.sciencedirect.com/science/article/abs/pii/S016578360300002X

Maunder, M.N. 2002. Growth of skipjack tuna (Katsuwonus pelamis) in the eastern Pacific Ocean, as estimated from tagging data. Inter American Tropical Tuna Commission Bulletin, 22(2): 93-131. (in English and Spanish). http://aquaticcommons.org/6876/

Maunder M.N. 2001. A general framework for integrating the standardization of catch-per-unit-of-effort into stock assessment models. Can. J. Fish. Aquat. Sci., 58: 795-803. https://www.nrcresearchpress.com/doi/abs/10.1139/f01-029#.Xp47wMhKjD4

Maunder, M.N. and Starr, P.J. 2001. Bayesian Assessment of the SNA1 snapper (Pagrus auratus) stock on the northeast coast of New Zealand. New Zealand Journal of Marine and Freshwater Research, 35: 87-110. https://www.tandfonline.com/doi/abs/10.1080/00288330.2001.9516980