[Note: a long post of interest only to people who care about data analysis and bad statistics, and maybe about the distant stars influencing your life.] By now, we should all be able to list the many reasons that p-values (or null-hypothesis-significance-testing, NHST) are awful: that “statistical significance” has nothing to do with effect size … Continue reading How do I hate p-values? Let me count the ways…
In Part I, I wrote about how I started exploring the topic of machine learning, and we looked briefly described one of its main aims: automating the task of classifying objects based on their properties. Here, I’ll give an example of this in action, and also describe some general lessons I’ve drawn from this experience. … Continue reading Learning about (machine) learning — part II
Machine learning is everywhere these days, as we train computers to drive cars, play video games, and fold laundry. This intersects my lab’s research as well, which involves lots of computational image analysis (e.g.). Nearly everything my students and I write involves writing or applying particular algorithms to extract information from data. In the past … Continue reading Learning about (machine) learning — Part I
A long post, in which you’ll have to slog or scroll through several paragraphs to get to the real question: can we navigate using fallen sticks? These days we seem to be inundated with deeply flawed scientific papers, often featuring shaky conclusions boldly drawn from noisy data, results that can’t be replicated, or both. I … Continue reading On the replication crisis in science and the twigs in my backyard
As of yesterday, the graduate student union here at the University of Oregon is on strike*. (I walked past three separate picket lines on my way to get coffee.) I don’t have anything profound to write about labor issues, but I thought I should post something that’s graduate-student-related! Quite often, the topic of “time to … Continue reading Universality, Scaling, and Time-to-Ph.D.
Several sources have pointed me to this neat web site of spurious correlations, showing graphically how, for example, the age of Miss America correlates with the number of murders by steam, hot vapours and hot objects, or my favorite: Though spurious correlations can be dangerous (and hilarious), it’s often useful to look for correlations in … Continue reading Reading this post? You get an “A”!
Our Physics Department Colloquium this week is on a topic I’m fond of: the analysis of super-resolution microscopy images. This occurrence isn’t surprising, since I invited the speaker, Alex Small, with whom I co-wrote a recent review paper on the subject. The problem that superresolution microscopy confronts is that it’s hard to see tiny things. … Continue reading I should think of a title involving the words “Small” and “Microscopy”
I came across a short article at Science’s news site that notes that “Up to 1000 NIH Investigators Dropped Out Last Year” — i.e. the number of investigators funded by the NIH is presently dropping, a likely consequence of shrinking funding. The article includes this graph: What I find striking about the graph is the … Continue reading Culling the (science) herd?
This week’s bad graph* plots three-dimensional data with a squashed perspective, hiding any scatter of the points along the “out of plane” axis: The cognoscenti will recognize the plot as visualizing the outcome of principal component analysis (PCA), in this case applied to the microbial communities in mice fed dust from homes** with dogs (D) … Continue reading Bad Graph of the Week (in 3D!)
It is inherently challenging to use a model to make predictions beyond the range of data to which the model was fit — making predictions about the future, for example, based on the past and present. Or, as the Danish proverb more elegantly says, “It’s difficult to make predictions, especially about the future.” Still, there’s … Continue reading You can use any model you want, as long as it’s linear and has a positive slope