Best reads of 2017

These are some of the best or most thought-provoking articles I’ve read this year. The categories and articles are organised alphabetically and I don’t necessarily agree with the ideas put forward.

Economics:

Labour’s Higher Education proposals will cost £8bn per year, although increase the deficit by more. Graduates who earn most in future would benefit most by Chris BelfieldJack Britton, and Laura van der Erve. A strong counter argument against free tuition for all university students.

NAFTA and other trade deals have not gutted American manufacturing — period by J. Brad DeLong. US foreign trade deals account for a small proportion of lost jobs and if America had followed the policies adopted by Germany their manufacturing industries would be in better shape.

On the effects of inequality on economic growth by Artir Kel. Critically analysing the claim that inequality stymies economic growth.

This Time Is Different: Eight Centuries of Financial Folly by Carmen Reinhart and Kenneth Rogoff. The most thorough analysis of the history of financial crises throughout the world and how the common theme is deregulation of the financial sector.

Securitization: Lessons Learned and the Road Ahead by Miguel Segoviano, Bradley JonesPeter Lindner, and Johannes Blankenheim. A detailed explanation as to the role competitive securitisation played in the Great Recession of 2008 and what can be done to have a healthy securitsation market.

Competition and Crisis in Mortgage Securitization by Michael Simkovic. A historical perspective on competitive securitisation and it’s role in various economic crashes.

Education:

The three generations of research on class size effects by Peter Blatchford. A summary of research on how class size affects attainment for students of different ages and abilities.

The school research lead and another nail in the coffin of Hattie’s Visible Learning by Oliver Caviglioi. Why you should be careful using aggregated effect sizes to determine the efficacy of teaching interventions.

Study of the Week: Hitting the Books, or Hitting the Bong? by Fredrik deBoer. A natural experiment shows how the availability of legal marijuana negatively impacts grades for university students.

why selection bias is the most powerful force in education by Fredrik deBoer. How a type of bias undermines a lot of conclusions drawn from educational research because the samples have not been randomly selected.

What does PISA 2015 tell us about deprivation and highly able children? by John Jerrim. Grammar schools don’t benefit poor but gifted students.

The myths of the digital native and the multitasker by Paul Kirschner and Pedro De Bruyckere. Dispelling two prevalent myths about students: young children have a natural affinity and ability for technology and some learners can multitask.

Bullies: Thugs or thinkers? by Jon Sutton. Is the typical conception of a bully as socially inept and emotionally stunted correct?

History:

Romance and Socialism in J.S. Mill by Helen Andrews. The personal history of John Stuart Mill and the impact his wife had on his thinking.

Guns, Germs, and Steel by Jared Diamond. An attempt at explaining the disparities in wealth and technology between countries by looking at their natural advantages due to geography and climate.

Response to: What are some of the main Anthropological criticisms of Jared Diamond’s Guns, Germs and Steel? by firedrops. A collection of criticisms of Guns, Germs, and Steel by Jared Diamond.

Miscellaneous:

How to stop being busy and become productive by Fabian Dablander. Some very useful suggestions on how to be more productive with your work.

What Do We Do with the Art of Monstrous Men? by Claire Dederer. How do we deal with creative geniuses who commit crimes?

Our Reaction to “Cat Person” Shows That We Are Failing as Readers by Larissa Pham. By presuming a piece of writing is based on an author’s life or trying to fit our world to literature we limit our ways of thinking.

The Way You Play Monopoly is Wrong by Alaric Stephen. Idiosyncratic house rules for Monopoly artificially extend the game and increase the influence of luck on the outcome.

Philosophy:

The African Enlightenment: The highest ideals of Locke, Hume and Kant were first proposed more than a century earlier by an Ethiopian in a cave by Dag Herbjørnsrud. The ideals of rationality and equality among humans were not first written down by white Europeans but by Ethiopian and Ghanaian thinkers.

The Absurd by Thomas Nagel. Why the often given arguments for life’s meaningless are inadequate and why we shouldn’t rail against this lack of meaning.

Postmodernism and “post-truth” by Paul Raven. A clear definition of what post-modernism actually is.

Philosophy of science and meta-science:

Understanding Psychology as a Science: An Introduction to Scientific and Statistical Inference by Zoltan Dienes. The best introduction to philosophy of science and the different statistical methods I’ve read.

Daryl Bem Proved ESP Is Real: Which means science is broken by Daniel Engber. A history of the group of studies that started the replication crisis in psychology.

What does research reproducibility mean? by Steven GoodmanDaniele Fanelli, and John Ioannidis. A clear set of definitions of often misunderstood terms.

Academic success is either a crapshoot or a scam by Ed Hagen. How science currently incentivises p-hacking due to the nature of studied phenomena and the way scientific publishing works.

Fuck Nuance by Kieran Healy. Why endless calls for nuance in understanding sociological phenomena dilute the power of theories developed to explain the world.

A Unified Framework to Quantify the Trustworthiness of Empirical Research by Etienne LeBel, Wolf VanpaemelRandy McCarthy, Brian Earp, and Malte Elson. An excellent suggestion for how to improve psychological science: by having a clearly defined rubric for replications and a centralised database to store replication attempts.

Models Are Stupid, and We Need More of Them by Paul Smaldino. Models are flawed but essential and a science suffers by not developing more clearly defined models.

Vast literatures as mud moats by Noah Smith. When someone says you need to “go and read the literature” in order to understand the topic, they need to give two papers that sum up the arguments so the phrase can’t be used as an obfuscating debate tool.

be your own a**hole by Simine Vazire. Why you should be sceptical of multiple p-values between 0.02-0.05 and critical of your work in general.

Politics:

Passing the Baton by Perry Anderson. How the first celebrity president helped pave the way for the current president through neoliberal capitalism and creating a cult of personality around him, to the detriment of his party and policies.

England is suffering from an internal brain drain – and it’s centuries old by Anthony Breach. More intelligent people have been migrating from the north of England to the south for centuries and this movement is only going to grow.

Germany’s Working Poor by Olivier Cyran. How deregulation of the labour market led to many German workers falling into poverty.

Why We Fight Wars by Matthew Evangelista. How neoliberal economic policies can create the political circumstances for violence to rise during times of upheavel.

Forget empire — Britain wants less of the world, not more by Janah Ganesh. The article that helped dispel my belief that many Brits voted Leave in the Referendum because they wanted to cling to some ill-defined concept of an empire.

You Are the Product by John Lancaster. How Facebook takes your data and sells it, whilst building a more comprehensive picture of you than any dystopian government can dream of.

Stop Pretending You’re Not Rich by Richard Reeves. America is not a meritocracy.

How the Self-Esteem Craze Took Over America And why the hype was irresistible by Jesse Singal. The history of America’s obsession with self-help books and pop psychology, the shoddy science behind it, and the consequences of it.

What the Opioid Crisis Can Teach Us about the War on Drugs by Robert Verbruggen. The (admittedly weak) evidence suggests decriminalisation of hard drugs like opioids will lead to more overdoses and addictions.

Statistics:

Introduction to the concept of likelihood and its applications by Alexander Etz. A clear explanation of likelihood that details why Bayesian statistics isn’t affected by the sampling method.

The failure of null hypothesis significance testing when studying incremental changes, and what to do about it by Andrew Gelman. Recommendations for publication procedures and study design to improve the accuracy of one’s inferences.

The “What does not kill my statistical significance makes it stronger” fallacy by Andrew Gelman. Why believing the significant result you found using a noisy test means the effect must be strong is false.

Some natural solutions to the p-value communication problem— and why they won’t work by Andrew Gelman and John Carlin. A short explanation of why various alternatives to p-values won’t work and how model expansion and Bayesian inference are the solution.

The Null Ritual: What You Always Wanted to Know About Significance Testing but Were Afraid to Ask by Gerd Gigerenzer, Stefan Krauss, and Oliver Vitouch. A history of Null Hypothesis Significance Testing and how it’s a Frankenstein’s Monster of different ideas.

Statistical tests, P values, confidence intervals, and power: a guide to misinterpretations by Sander Greenland, Stephen Senn, Kenneth Rothman, John Carlin, Charles Poole, Steven Goodman, and Douglas Altman. The best article on what p-values, confidence intervals, and statistical power actually are that dispels many of the misconceptions surrounding these terms.

Equivalence Tests: A Practical Primer for t-Tests, Correlations, and Meta-Analyses by Daniel Lakens. Equivalence tests allow researchers to state whether a sufficiently large effect size isn’t present (based on the smallest effect size they can detect with their statistical power), so null results are more informative.

How p-values solve 50% of the problems with p-values by Daniel Lakens. How equivalence tests solve many of the problems that arise due to incorrect interpretations of p-values.

Too True to be Bad: When Sets of Studies With Significant and Non-significant Findings Are Probably True by Daniel Lakens and Alexander Etz. Why an array of significant and non-significant results are more persuasive (and more likely to be true) than a series of only significant p-values.

Abandon Statistical Significance by Blakeley McShane, David Gal, Andrew Gelman, Christian Robert, and Jennifer Tackett. An explanation of the problems with both the 0.05 and 0.005 significance thresholds and a call to dethrone p-values from its privileged position to one piece of evidence among many.

Why most of psychology is statistically unfalsifiable by Richard Morey and Daniel Lakens. A response to the Reproducibility Project which details why low power in the original study is such a problem for the replication, as well as many misconceptions about statistical power.

The prevalence of statistical reporting errors in psychology (1985–2013) by Michèle Nuijten, Chris Hartgerink, Marcel van Assen, Sacha Epskamp, & Jelte M. Wicherts. The paper that introduced the world to statcheck.

Boosting power with better experiments by Sam Schwazrkopf. Why you don’t need to necessarily increase your N to improve power.

The selection-distortion effect: How selection changes correlations in surprising ways by Sanjay Srivastava. The blog post that helped me finally understand Berkson’s Paradox (also called “conditioning on a collider”).

Statistically Controlling for Confounding Constructs Is Harder than You Think by Jacob Westfall and Tal Yarkoni. The title says it all.

Starter tips on sharing data and analysis scripts by Katherine Wood. The resource I used to help prepare my first open data, code, and materials project.

Write a Comment

Your email address will not be published. Required fields are marked *