Predictive models based on machine learning can be highly sensitive to data error. Training data are often combined from a variety of different sources, each susceptible to different types of inconsistencies, and as new data stream in during prediction time, the model may encounter previously unseen inconsistencies. An important class of such inconsistencies are domain value violations that occ...

We address the problem of reliability of independence-based causal discovery algorithms that results from unreliable statistical independence tests. We model the problem as a knowledge base containing a set of independences that are related through the well-known Pearl's axioms. Statistical tests on finite data sets may result in errors in these tests and inconsistencies in the knowledge base. ...

Experimental choice data from 881 subjects based on 40 time-tradeoff items and 32 risky choice items reveal that most subjects are time-inconsistent and most violate the axioms of expected utility theory. These inconsistencies cannot be explained by well-known theories of behavioral inconsistency, such as hyperbolic discounting and cumulative prospect theory. Aggregating expected payoffs and th...

What is the physical significance of entropy? What is the physical origin of irreversibility? Do entropy and irreversibility exist only for complex and macroscopic systems? For everyday laboratory physics, the mathematical formalism of Statistical Mechanics (canonical and grand-canonical, Boltzmann, Bose-Einstein and Fermi-Dirac distributions) allows a successful description of the thermodynami...

Sibley and Ahlquist compared the single-copy nuclear DNA sequences of the hominoid primates using DNA-DNA hybridization. From this data set they estimated a phylogeny that clusters man and chimpanzees using a distance Wagner procedure. However, no assessment of statistical confidence in this estimated phylogeny was made, despite the fact that their data set contains internal inconsistencies con...

This study documents reporting errors in a sample of over 250,000 p-values reported in eight major psychology journals from 1985 until 2013, using the new R package "statcheck." statcheck retrieved null-hypothesis significance testing (NHST) results from over half of the articles from this period. In line with earlier research, we found that half of all published psychology papers that use NHST...