This page links to some selected presentations, and the related papers.

A re-appraisal of fixed effect(s) meta-analysis

• Slides for a joint RSS/PSI webinar, September 24th 2018 [.pdf] [Video of talk and discussion]
• The slides from my 2018 JSM talk [.pdf]
• A re-appraisal of fixed effect(s) meta-analysis: Rice, Higgins and Lumley, JRSSA 2017 [paper]
• Addressing the estimation of standard errors in fixed effects meta-analysis: Dominguez-Islas and Rice, Statistics in Medicine 2018 [paper]
• A Bayesian Fixed Effects Approach to Meta-Analysis: Decoupling Exchangeability and Hierarchical Modelling: Dominguez-Islas and Rice, under review [.pdf]

Statistics for large-scale genetic studies

• Slides (PPS) giving a non-technical introduction to some statistical issues seen in high-throughput studies
• This talk on fastSKAT - a method for simultaneous association testing of extremely large numbers of variants - was selected as an oral presentation at ASHG 2017. The paper has more details.

Decision-theoretic approaches to testing

• Slides and video recording from ISBA 2018, on some work with Tyler Bonnett, describing why 2-sided p-values are not unBayesian after all
• Slides (PDF); includes material from this paper - the slides contain briefly cover some other topics, and give graphical explanations and examples

Equivalence of random-effects and conditional likelihoods

• Slides (PDF); this talk (given to various audiences) covers most of the material in my 2004 JASA paper
• An updated version with some new material (and more pointed comments about puritans)
• For a quick introduction, see this poster (JPG), displayed at the 4th Workshop on Bayesian Nonparametrics in Rome, June 2004. The content is Bayesian and semi-parametric (and colorful) but not "Bayesian semiparametric"!