Séminaire n°29 + Atelier

Seminar: The New Statistics: Why, How, and Where Next?

Speaker:  Geoff Cumming, Statistical Cognition Laboratory, School of Psychological Science, La Trobe University, Melbourne, Australia

Contact : g.cumming (at) latrobe.edu.au                

Date : 11/05/12, 10-12h

Abstract : 
Many disciplines rely on null hypothesis significance testing (NHST) and p values, despite their deep flaws having been known for more than half a century. I will explain why estimation, based on effect sizes and confidence intervals, is a much more informative approach. 'The Dance of the p Values' demonstrates one dramatic shortcoming of p values: A replication experiment is likely to give a very different value of p, so p simply cannot be trusted.
I refer to estimation - and its extension, meta-analysis - as The New Statistics. The techniques themselves are not new, but for most researchers it would be new, and a highly beneficial change, to switch from NHST to these techniques. I will describe practical ways to use the new statistics. I will use ESCI (Exploratory Software for Confidence Intervals), which runs under Excel, to illustrate concepts and calculate confidence intervals. ESCI is a free download here.
That website also gives information about my book: Cumming, G. (2012). Understanding The New Statistics: Effect Sizes, Confidence Intervals, and Meta-Analysis. New York: Routledge.
I gave a short radio talk that summarises the main argument for the new statistics. The podcast and transcript are available here.

Talk :  downloadable here (pps file)

Workshop: The New Statistics in Practice

Speaker :  Geoff Cumming

Date : 11/05/12, 14-17h

Abstract :
In this workshop I will discuss how to use the new statistics in practice. Choice of topics will be responsive to the interests of people attending. I could consider various measures, including correlations, proportions, and the standardized effect size Cohen’s d. I could consider a range of simple experimental designs. I will also discuss meta-analysis. ESCI will serve to illustrate many of the ideas, and calculate confidence intervals in the different situations. I will consider statistical power, but will emphasize the advantages of an alternative approach to planning experiments: Precision for planning. This approach calculates the N required for our planned experiment to be likely to give a confidence interval that is not greater than some specified target length. There will be ample time for discussion, and for considering data and situations that are of particular interest to participants.

Talk :  downloadable here (pps file)

Entrée libre sur inscription auprès de claudia.fritz (at) upmc.fr