Global model analysis by parameter space partitioning. Pitt, M. A., Kim, W., Navarro, D. J., & Myung, J. I. *Psychological Review*, 113(1):57–83, 2006.

Paper doi abstract bibtex

Paper doi abstract bibtex

To model behavior, we need to know how models behave. This means learning what other behaviors a model can produce besides the one generated by participants in an experiment. This is a difficult problem because of the complexity of psychological models (e.g., their many parameters) and because the behavioral precision of models (e.g.,interval-scale performance) often mismatches their testable precision in experiments, where qualitative, ordinal predictions are the norm. Parameter space partitioning is a solution that evaluates model performance at a qualitative level. Given a definition of a qualitative data pattern, there exists a partition on the model's parameter space that divides it into regions that correspond to each data pattern. Markov-chain Monte Carlo methods are used to discover and define these regions. Three application examples, all using connectionist models, demonstrate its potential and versatility for studying the global behavior of psychological models. Among other things, one can easily assess how central and robust the empirical data pattern is to the model, as well as the range and characteristics of its other behaviors.

@article{pitt_global_2006, title = {Global model analysis by parameter space partitioning.}, volume = {113}, issn = {1939-1471, 0033-295X}, url = {http://doi.apa.org/getdoi.cfm?doi=10.1037/0033-295X.113.1.57}, doi = {10.1037/0033-295X.113.1.57}, abstract = {To model behavior, we need to know how models behave. This means learning what other behaviors a model can produce besides the one generated by participants in an experiment. This is a difficult problem because of the complexity of psychological models (e.g., their many parameters) and because the behavioral precision of models (e.g.,interval-scale performance) often mismatches their testable precision in experiments, where qualitative, ordinal predictions are the norm. Parameter space partitioning is a solution that evaluates model performance at a qualitative level. Given a definition of a qualitative data pattern, there exists a partition on the model's parameter space that divides it into regions that correspond to each data pattern. Markov-chain Monte Carlo methods are used to discover and define these regions. Three application examples, all using connectionist models, demonstrate its potential and versatility for studying the global behavior of psychological models. Among other things, one can easily assess how central and robust the empirical data pattern is to the model, as well as the range and characteristics of its other behaviors.}, language = {en}, number = {1}, urldate = {2019-01-07}, journal = {Psychological Review}, author = {Pitt, Mark A. and Kim, Woojae and Navarro, Daniel J. and Myung, Jay I.}, year = {2006}, pages = {57--83} }

Downloads: 0

{"_id":"vAvTDaCuo3Pc537su","bibbaseid":"pitt-kim-navarro-myung-globalmodelanalysisbyparameterspacepartitioning-2006","authorIDs":[],"author_short":["Pitt, M. A.","Kim, W.","Navarro, D. J.","Myung, J. I."],"bibdata":{"bibtype":"article","type":"article","title":"Global model analysis by parameter space partitioning.","volume":"113","issn":"1939-1471, 0033-295X","url":"http://doi.apa.org/getdoi.cfm?doi=10.1037/0033-295X.113.1.57","doi":"10.1037/0033-295X.113.1.57","abstract":"To model behavior, we need to know how models behave. This means learning what other behaviors a model can produce besides the one generated by participants in an experiment. This is a difficult problem because of the complexity of psychological models (e.g., their many parameters) and because the behavioral precision of models (e.g.,interval-scale performance) often mismatches their testable precision in experiments, where qualitative, ordinal predictions are the norm. Parameter space partitioning is a solution that evaluates model performance at a qualitative level. Given a definition of a qualitative data pattern, there exists a partition on the model's parameter space that divides it into regions that correspond to each data pattern. Markov-chain Monte Carlo methods are used to discover and define these regions. Three application examples, all using connectionist models, demonstrate its potential and versatility for studying the global behavior of psychological models. Among other things, one can easily assess how central and robust the empirical data pattern is to the model, as well as the range and characteristics of its other behaviors.","language":"en","number":"1","urldate":"2019-01-07","journal":"Psychological Review","author":[{"propositions":[],"lastnames":["Pitt"],"firstnames":["Mark","A."],"suffixes":[]},{"propositions":[],"lastnames":["Kim"],"firstnames":["Woojae"],"suffixes":[]},{"propositions":[],"lastnames":["Navarro"],"firstnames":["Daniel","J."],"suffixes":[]},{"propositions":[],"lastnames":["Myung"],"firstnames":["Jay","I."],"suffixes":[]}],"year":"2006","pages":"57–83","bibtex":"@article{pitt_global_2006,\n\ttitle = {Global model analysis by parameter space partitioning.},\n\tvolume = {113},\n\tissn = {1939-1471, 0033-295X},\n\turl = {http://doi.apa.org/getdoi.cfm?doi=10.1037/0033-295X.113.1.57},\n\tdoi = {10.1037/0033-295X.113.1.57},\n\tabstract = {To model behavior, we need to know how models behave. This means learning what other behaviors a model can produce besides the one generated by participants in an experiment. This is a difficult problem because of the complexity of psychological models (e.g., their many parameters) and because the behavioral precision of models (e.g.,interval-scale performance) often mismatches their testable precision in experiments, where qualitative, ordinal predictions are the norm. Parameter space partitioning is a solution that evaluates model performance at a qualitative level. Given a definition of a qualitative data pattern, there exists a partition on the model's parameter space that divides it into regions that correspond to each data pattern. Markov-chain Monte Carlo methods are used to discover and define these regions. Three application examples, all using connectionist models, demonstrate its potential and versatility for studying the global behavior of psychological models. Among other things, one can easily assess how central and robust the empirical data pattern is to the model, as well as the range and characteristics of its other behaviors.},\n\tlanguage = {en},\n\tnumber = {1},\n\turldate = {2019-01-07},\n\tjournal = {Psychological Review},\n\tauthor = {Pitt, Mark A. and Kim, Woojae and Navarro, Daniel J. and Myung, Jay I.},\n\tyear = {2006},\n\tpages = {57--83}\n}\n\n","author_short":["Pitt, M. A.","Kim, W.","Navarro, D. J.","Myung, J. I."],"key":"pitt_global_2006","id":"pitt_global_2006","bibbaseid":"pitt-kim-navarro-myung-globalmodelanalysisbyparameterspacepartitioning-2006","role":"author","urls":{"Paper":"http://doi.apa.org/getdoi.cfm?doi=10.1037/0033-295X.113.1.57"},"downloads":0},"bibtype":"article","biburl":"https://api.zotero.org/users/3669931/collections/39JD4RHV/items?key=IdVJdX5Vfg3Y1Q3KOEpsbi2y&format=bibtex&limit=100","creationDate":"2019-06-13T18:59:44.289Z","downloads":0,"keywords":[],"search_terms":["global","model","analysis","parameter","space","partitioning","pitt","kim","navarro","myung"],"title":"Global model analysis by parameter space partitioning.","year":2006,"dataSources":["Z8J7pYW3gSpt2BMs5"]}