Category: Ikke kategoriseret

New guide to reaction time analysis

Tweet Share Email Share Share FollowTL;DR: I just published “Reaction Time Distributions – An Interactive Overview“. There is a big literature on the analysis of Reaction Time data. Everybody can see that reaction times are not normally distributed but there is little consensus about how they are distributed. This has resulted in many advanced mathematical arguments why one or the other distribution is better. Others propose obscure rules-of-thumb ways of deleting or transforming data so that it looks normally distributed. Meanwhile, the average researcher is left bewildered, often resorting to some variant of the normal distribution which they know and …

New tutorial on optimal decisions using Utility Theory

Tweet Share Email Share Share FollowTL;DR, here is the tutorial! There is a wave towards using data more and more in decision making across all levels of business and society. However, people often use data quite informally: look at an Excel sheet or a graph, then make a decision based on your impression. This often works well, but it can be fragile because of our many deep-skin biases as well as a general poor ability to reason about quantities and complex interactions. Decision theory to the rescue! By adding a few axioms to the basic axioms of probility theory, we can …

Do We All Have “Impaired” Awareness of Our Abilities?

Tweet Share Email Share Share FollowThis is my poster for Neuroscience Day 2019. It is quite provocative, and there are nuances to this story: This is likely a sort of Simpson’s Paradox in reverse, where there is little sensitivity to objective performance within groups (patients vs. healthy), but some sensitivity between groups. I do not dispute that subjective reports reflect real subjective experiences. As such, measures on Quality of Life, emotional distress, etc. are not to be disregarded. But care should be taken to generalize from, e.g., reports of emotional distress to impacts on real abilities. I do have a very nice dataset coming …

Scanner radiation caused 1% of flight-related cancers.

Tweet Share Email Share Share FollowAfter extended public anxiety about cancer risks associated with back-scatter scanners, EU and U.S. banned them in 2012 and 2013 respectively. But how many people actually developed cancer from these scans before they were banned? I have yet to find articles that estimate the world-wide mortality using consensus numbers. Most just state that the risk is “negligible” or “truly trivial”. That vague language is not comforting to a pedantic like me, so let’s look at the actual numbers. See the end of this post for a full list of sources and informative infographics. Risk per …

History and future of R formula syntax

Tweet Share Email Share Share FollowThe R formula syntax is wonderfully condensed yet instructive. Python has basically given up coming up with its own syntax and now just use the `patsy` module to use R syntax in Python. However, this particular syntax has no name. During twitter interactions the last few days, people have suggested “symbolic model notation”, “abridged model notation”, “Wilkinson notation”, and a few others. I think none of them did a good job of delineating this exact short notation, so I looked into the historical origins and posted this Twitter thread (click to read it all): Tweet …