Forest plots in R

http://www.r-bloggers.com/forest-plots-in-r-ggplot-with-side-table/

More on Gelman and Shalizi's philosphy of Bayesian statistics

Gelman and Shalizi: Rejoinder

Denny Borsboom
How to practise Bayesian statistics outside the Bayesian church: What philosophy for Bayesian statistical modelling?

Deborah Mayo
The error-statistical philosophy and the practice of Bayesian statistics: Comments on Gelman and Shalizi: Philosophy and the practice of Bayesian statistics

Multi-topic paper dump

 Michael Hout et al.
Multi-dimensional scaling

Pennie Dodds
Perhaps unidimensional is not unidimensional

Michael Riley
The dynamics of cognition

Ut Na Sio
Sleep on it: problem solving

Lavelle
Contrastive explanation and the many absences problem

Unbiased peer review

Very interesting peer review

Links that i need to post before I forget them

Roderick Little
The Prevention and Treatment of Missing Data in Clinical Trials

David Bodescu
Revisiting the Gain–Loss Separability Assumption in Prospect Theory

Robert Tibshirani
Scientific research in the age of omics: the good, the bad, and the sloppy

Gerd Gigerenzer
Homo Heuristicus

John Thogerson
Does Green Consumerism Increase the Acceptance of Wind Power?

Publication bias

 Gregory Francis

Publication bias and the failure of replication

Some notes on writing

Discussion sections.  I usually hate writing discussions.  However, I find they are extremely useful for two things.  First, for each result (affirming and disconfirming) sincerely asking oneself why the results occurred.  Then for each why question, generate explanations and consider the ones I believe the most in detail.  Try to find other evidence or ideas that relates to these explanations.  This process allows surprises to emerge, changing our questions and explanations, but also allows us to build stories around the evidence.  These stories can be used to push the theory further.

Future Directions.  Future directions should be based on understanding the processes and theory, pushing them forward, rather than trying to find new contexts.  Process rather than context driven.

For prescriptions.  The prescriptions should either refine existing proposals based on the results of the experiments (descriptive research), distinguish among existing proposals, or develop new ones.