Prisme N°21
Michel Armatte
February Release
Booklets
The Cournot Centre’s Prismes
Is Everything Stochastic?
Prisme N°20
Glenn Shafer
Prisme N°20 December 2010 (747.1 KiB)
Kolmogorov said no, Popper said yes. My sympathies lie with Kolmogorov, the old-fashioned empiricist.
In the on-line setting, where we see previous outcomes before making the next probability forecast, we can give probabilities that have objective value because they pass statistical tests. This accounts for the success of many adaptive methods, and it is related to Leonid Levin’s notion of a universal prior probability distribution. It tells us that yes, everything is stochastic, but in a sense that is empirically empty because it is not falsifiable.
When we understand that success of adaptive methods does not depend on the world being stochastic in a falsifiable sense, we may want to be more parsimonious in causal modeling and more open to non-standard methods of probability judgement.
The Art of Spurious Regressions: Can Statistics Do without Artefacts?
Prisme N°19
Jean-Bernard Chatelain
Prisme N°19 Octobre 2010 (1005.5 KiB)
This text is currently being translated into English.
This article presents a particular case of spurious regression, where a dependent variable has a simple correlation coefficient close to zero with two other variables, which are on the contrary highly correlated with each other. In this type of regression, the parameters measuring the magnitude of effects on the dependent variable are very high. They can be “statistically significant”. The tendency of scientific journals to favour the publication of statistically significant results is one reason why spurious regressions are so numerous, especially since it is easy to build them with variables that are lagged, squared or interacting with another variable. Such regressions can enhance the reputation of researchers, by stimulating the appearance of strong effects between variables. These often surprising effects are not robust. They often depend on a limited number of observations, and this fact has fuelled scientific controversies. The resulting meta-analyses, based on statistical synthesis of the literature evaluating this effect between two variables, confirm the absence of any effect. The text provides an example of this phenomenon in the empirical literature, with the aim of evaluating the impact of development aid on economic growth.
• Order free of charge
Prisme N°18
Prisme N°17
Prisme N°16An Economic Analysis of Fair Value: Accounting as a Vector of Crisis
Prisme N°15
Vincent Bignon, Yuri Biondi & Xavier Ragot
Prisme N°15 August 2009 (537.1 KiB)
Read more »
Prisme N°14
Prisme N°13
Prisme N°12Does Fiscal Stimulus Work?
Prisme N°11
Edouard Challe
Also available in German
Prisme N°11 November 2007 (254.3 KiB)
Read more »
Prisme N°10
Prisme N°9
Prisme N°8
Prisme N°7
Prisme N°6
Prisme N°5
Prisme N°4
Prisme N°3
Prisme N°2Ramifications of the European Commission’s Directive on Takeover Bids
Prisme N°1
Jean-Louis Beffa, Leah Langenlach & Jean-Ph
ilippe Touffut
Prisme N°1 September 2003 (262.3 KiB)
Read more »
