James S. Hodges
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
James S. Hodges has a total of 4 co-authors. Here are the top co-authors by number of shared publications:
Bradley P. Carlin
Division of Biostatistics, University of Minnesota, Minneapolis, Minnesota USA 55414
2 shared publications- Qiao Fan
Nephrology Analytical Service, Minneapolis Medical Research Foundation, 600 HFA Building, 914 S. Eighth St., Minneapolis, Minnesota 55404, U.S.A.
1 shared publication Xiaoxiao Kong
Division of Biostatistics, University of Minnesota, Minneapolis, Minnesota USA 55414
1 shared publicationYue Cui
Department of Biometrics, Abbott, Abbott Park, Illinois USA 60064
1 shared publication
Latest Publications
- 1 February 2010
- Published by Taylor & Francis in Technometrics
- Vol. 52 (1) , 124-136
Partitioning Degrees of Freedom in Hierarchical and Other Richly Parameterized Models
Hodges and Sargent (2001) have developed a measure of a hierarchical model’s complexity, degrees of freedom (DF), that is consistent with definitions for scatterplot smoothers, is interpretable in terms of simple models, and enables control of a fit’s complexity by means of a prior distribution on complexity. But although DF describes the complexity of the whole fitted model, in general it remains unclear how to allocate DF to individual effects. Here we present a new definition of DF for arbitrary normal-error linear hierarchical models, consistent with that of Hodges and Sargent, that naturally partitions the n observations into DF for individual effects and for error. The new conception of an effect’s DF is the ratio of the effect’s modeled variance matrix to the total variance matrix. This provides a way to describe the sizes of different parts of a model (e.g., spatial clustering vs. heterogeneity), to place DF-based priors on smoothing parameters, and to describe how a smoothed effect competes with other effects. It also avoids difficulties with the most common definition of DF for residuals. We conclude by comparing DF with the effective number of parameters, pD, of Spiegelhalter et al. (2002). Technical appendices and a data set are available online as supplemental materials.
- 1 February 2010
- Published by Taylor & Francis in Technometrics
- Vol. 52 (1) , 124-136
Partitioning Degrees of Freedom in Hierarchical and Other Richly Parameterized Models
Hodges and Sargent (2001) have developed a measure of a hierarchical model’s complexity, degrees of freedom (DF), that is consistent with definitions for scatterplot smoothers, is interpretable in terms of simple models, and enables control of a fit’s complexity by means of a prior distribution on complexity. But although DF describes the complexity of the whole fitted model, in general it remains unclear how to allocate DF to individual effects. Here we present a new definition of DF for arbitrary normal-error linear hierarchical models, consistent with that of Hodges and Sargent, that naturally partitions the n observations into DF for individual effects and for error. The new conception of an effect’s DF is the ratio of the effect’s modeled variance matrix to the total variance matrix. This provides a way to describe the sizes of different parts of a model (e.g., spatial clustering vs. heterogeneity), to place DF-based priors on smoothing parameters, and to describe how a smoothed effect competes with other effects. It also avoids difficulties with the most common definition of DF for residuals. We conclude by comparing DF with the effective number of parameters, pD, of Spiegelhalter et al. (2002). Technical appendices and a data set are available online as supplemental materials.
- 11 June 2003
- Published by Oxford University Press (OUP) in Biometrics
- Vol. 59 (2) , 317-322
On the Precision of the Conditionally Autoregressive Prior in Spatial Models
SummaryBayesian analyses of spatial data often use a conditionally autoregressive (CAR) prior, which can be written as the kernel of an improper density that depends on a precision parameter τ that is typically unknown. To include τ in the Bayesian analysis, the kernel must be multiplied byτkfor somek. This article rigorously derivesk= (n−I)/2for theL2norm CAR prior (also called a Gaussian Markov random field model) andk=n−Ifor theL1norm CAR prior, wherenis the number of regions andIthe number of “islands” (disconnected groups of regions) in the spatial map. SinceI= 1for a spatial structure defining a connected graph, this supportsKnorr‐Held's(2002, inHighly Structured Stochastic Systems, 260–264) suggestion thatk= (n− 1)/2in theL2norm case, instead of the more commonk=n/2. We illustrate the practical significance of our results using a periodontal example.
- 11 June 2003
- Published by Oxford University Press (OUP) in Biometrics
- Vol. 59 (2) , 317-322
On the Precision of the Conditionally Autoregressive Prior in Spatial Models
SummaryBayesian analyses of spatial data often use a conditionally autoregressive (CAR) prior, which can be written as the kernel of an improper density that depends on a precision parameter τ that is typically unknown. To include τ in the Bayesian analysis, the kernel must be multiplied byτkfor somek. This article rigorously derivesk= (n−I)/2for theL2norm CAR prior (also called a Gaussian Markov random field model) andk=n−Ifor theL1norm CAR prior, wherenis the number of regions andIthe number of “islands” (disconnected groups of regions) in the spatial map. SinceI= 1for a spatial structure defining a connected graph, this supportsKnorr‐Held's(2002, inHighly Structured Stochastic Systems, 260–264) suggestion thatk= (n− 1)/2in theL2norm case, instead of the more commonk=n/2. We illustrate the practical significance of our results using a periodontal example.