TY - JOUR
T1 - Recognition memory models and binary-response ROCs
T2 - A comparison by minimum description length
AU - Kellen, David
AU - Klauer, Karl Christoph
AU - Bröder, Arndt
N1 - Funding Information:
We thank Chad Dube and Jeff Starns for providing their original data sets. The research reported in this article was supported by grant Kl 614/32-1 from the Deutsche Forschungsgemeinschaft to Karl Christoph Klauer. Model-fitting routines used in this article can be obtained upon request
PY - 2013
Y1 - 2013
N2 - Model comparison in recognition memory has frequently relied on receiver operating characteristics (ROC) data. We present a meta-analysis of binary-response ROC data that builds on previous such meta-analyses and extends them in several ways. Specifically, we include more data and consider a much more comprehensive set of candidate models. Moreover, we bring to bear modern developments in model selection on the current selection problem. The new methods are based on the minimum description length framework, leading to the normalized maximum likelihood (NML) index for assessing model performance, taking into account differences between the models in flexibility due to functional form. Overall, NML results for individual ROC data indicate a preference for a discrete-state model that assumes a mixture of detection and guessing states.
AB - Model comparison in recognition memory has frequently relied on receiver operating characteristics (ROC) data. We present a meta-analysis of binary-response ROC data that builds on previous such meta-analyses and extends them in several ways. Specifically, we include more data and consider a much more comprehensive set of candidate models. Moreover, we bring to bear modern developments in model selection on the current selection problem. The new methods are based on the minimum description length framework, leading to the normalized maximum likelihood (NML) index for assessing model performance, taking into account differences between the models in flexibility due to functional form. Overall, NML results for individual ROC data indicate a preference for a discrete-state model that assumes a mixture of detection and guessing states.
KW - Discrete-state models
KW - Hybrid models
KW - Minimum description length
KW - Model selection
KW - Normalized maximum likelihood
KW - Recognition memory
KW - Signal detection
UR - http://www.scopus.com/inward/record.url?scp=84880738547&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84880738547&partnerID=8YFLogxK
U2 - 10.3758/s13423-013-0407-2
DO - 10.3758/s13423-013-0407-2
M3 - Article
C2 - 23504915
AN - SCOPUS:84880738547
SN - 1069-9384
VL - 20
SP - 693
EP - 719
JO - Psychonomic Bulletin and Review
JF - Psychonomic Bulletin and Review
IS - 4
ER -