So, there's a burgeoning cog sci-lite literature... Danny Khaneman, Rolf Dobelli etc. A lot of these books are essentially (well-written, entertaining)So, there's a burgeoning cog sci-lite literature... Danny Khaneman, Rolf Dobelli etc. A lot of these books are essentially (well-written, entertaining) catalogues of cognitive biases. This book comes at the problem from the other angle - less 'why do we think wrong?' and more 'how can we think better?'. Parrish does so by riffing of Charlie Munger's idea that good thinking requires a collection of time-proven mental models which draw from across different disciplines. Compared to some other work out there this book is light on for scientific references, and it trades heavily in historical examples. So, for example, there is no real evidence that reading the book would make you a better thinker (whatever that is). For me, these concerns are minor and I really enjoyed the book. Favourite chapters: The Map is not The Territory: "The map of reality is not reality. Even the best maps are imperfect. That’s because they are reductions of what they represent. If a map were to represent the territory with perfect fidelity, it would no longer be a reduction and thus would no longer be useful to us. A map can also be a snapshot of a point in time, representing something that no longer exists. This is important to keep in mind as we think through problems and make better decisions." Draws together a lot of ideas with some cool examples.
Hanlon's razor: "...we should not attribute to malice that which is more easily explained by stupidity. In a complex world, using this model helps us avoid paranoia and ideology. By not generally assuming that bad results are the fault of a bad actor, we look for options instead of missing opportunities. This model reminds us that people do make mistakes. It demands that we ask if there is another reasonable explanation for the events that have occurred. The explanation most likely to be right is the one that contains the least amount of intent." Final chapter in the book - stitched together a lot of other ideas really well....more
OK. So, this is a beast which is neither fish nor fowl. It’s somewhere between popular science and a coursebook.
Many model thinking is so hot right noOK. So, this is a beast which is neither fish nor fowl. It’s somewhere between popular science and a coursebook.
Many model thinking is so hot right now (see Munger, Parish etc etc.). This book kinda fits into that genre.
On the other hand, it is also a much more serious treatment of how to apply analytic models, and it’s almost a textbook for his course: https://www.coursera.org/learn/model-...
That said, the book doesn’t get into the nitty gritty of how to use each approach. That’s not necessarily a problem – learning how to use an analytic technique ‘for realz’ takes time.
What I liked: I’m a researcher/ data analyst. My stats skills are pretty solid, but are very much grounded in a psychology background (almost everything I do belongs to the family of regression/ structural equation models). What this did is blow things wide open and offer a lot of other possible views.
The book uses a variety of examples to make it’s point.
He also makes the argument of multiple models as offering complementary views and tools (a sort of pluralist approach).
My favourite parts: The illustrations of using multiple models to offer complementary explanations of the same phenomenon, e.g. the Global Financial Crisis, the Cuban missile crisis, and rising social inequality.
Minor objections: Page is clearly a strong advocate for combining models (many model thinking). The book could maybe do with a little more discussion on the practicalities of combining models, and also potential errors/ drawbacks.
Given Page talks about causality a bit, I think it could have done with its own chapter. Selfishly, I’d love to see his take on Judea Pearl’s work, and how he compares and contrasts it to other approaches.
TBH, my biggest objection is that some of the images and equations came through with very poor image quality on the kindle. That problem is not unique to this book, but it’s still a bit annoying.
All that said, it was good enough for me to enrol in the Coursera version....more
This was cool – it gave a nice flavour of the characters involved, the philosophy associated with Bayes, and many of the uses. There was very little teThis was cool – it gave a nice flavour of the characters involved, the philosophy associated with Bayes, and many of the uses. There was very little technical detail, which I found a bit frustrating at times. But this was also the intent behind the book; it was a pop science overview and gave a very big picture without bogging down in specifics. ...more
The book is on the note-taking method of German academic and lawyer, Niklas Luhmann. Luhmann had an incredBest (only) book on note-taking I ever read.
The book is on the note-taking method of German academic and lawyer, Niklas Luhmann. Luhmann had an incredibly productive career, which he attributed to his remarkable process of keeping notes in a slip-box. Basically this means jotting ideas on small cards of paper, and putting them in a shoe box. Without re-describing the whole process, you focus on building links between cards and having a dialogue with the slip-box.
Subsequent to reading this it seems there is a small community of German social science nerds out there who are REALLY into this stuff.
What I liked — the book was an unexpectedly exciting treatise on how we think. Ahrens makes the case quite convincingly that we don’t record our thoughts in writing; writing is thinking. He gives the examples of physics or maths – without the symbolic language and notes the thought process and discoveries simply could not have occurred.
He bolsters his argument with a broad range evidence and examples. For a book on note-taking it was a surprisingly entertaining bit of pop cognitive science.
I picked up some cool questions to ask myself while reading, and I like his emphasis on outlining.
When I take notes, I tend to copy massive chunks of quotes. Ahrens has (somewhat) beaten this out of me, and I’m trying much harder to put things in my own words.
Where I’m struggling — the software package he recommended is less than intuitive. It could do with some more worked examples. Unfortunately the YouTubes associated with this particular software package are in German. There are other resources out there, so I’m now in the grind of working out how this will marry up with my daily work flow as a quantitative social scientist.
If you are a science or stats geek, or frustrated with the replication crisis in across various disciplines, or even a philosophy/ cognitive science bIf you are a science or stats geek, or frustrated with the replication crisis in across various disciplines, or even a philosophy/ cognitive science boffin, this book is highly recommended.
Judea Pearl is a heavy heavy hitter. He was a big deal in Computing and Artificial Intelligence (at the forefront of Bayesian networks, which are central to mobile phone signal technology), before he made the leap to questions of causal inference.
The knock on Pearl has been his writing – it’s so hard to get through, I suspect his work didn’t get serious leverage until better communicators (Greenland, Hernan, Robins etc.) came to the party. I attempted his monolithic ‘Causality: Models, Reasoning and Inference’ but was defeated by it.
This book is the fix. Someone needs to buy Dana Mackenzie a shiny new car, or at least a carton of beer, for his work in helping Pearl get his ideas across in an accessible fashion.
Some of the main complaints from other reviewers are that either: 1) it’s too technical; or 2) it’s not how-to enough. So, yes IT IS A BIT TECHNICAL. And yes, THIS IS NOT A HOW-TO.
1) IT IS A BIT TECHNICAL. I think whether or not it is too technical depends on where you are at – I take this stuff seriously, and I expected a bit of pain, so I think he simplified just enough (I describe my background below, if it helps you orient yourself to the review). The book is not at all mathematical, although it is brutally logical in spots.
2) THIS IS NOT A HOW-TO. It’s a big picture overview (I suggest a few how-to’s below).
The third major criticism of Pearl is about where he sits into other approaches to causality, particularly those in economics (and to a lesser extent machine learning) – I don’t know enough about this yet but I’ll add an appendix to my review as I build sufficient background.
The book is about what Pearl calls the Causal Revolution. It’s about scientists (especially in social science and epidemiology) taking the question of when you can (and cannot) infer causality seriously. The book gives an excellent review of the evolution of ideas in statistical and science about causality, and lays down a serious challenge to the mantras ‘no causality in observational studies’ and ‘correlation does not imply causality’. At the very least, Pearl helps make explicit when these mantras make sense — Pearl makes extensive use of the debate over smoking and lung cancer to illustrate his point. As Hernan and Robins point out in their book, how many of us need evidence from an RCT to confidently deduce that putting your hand on a hot stove causes burning?
(As an analyst/ research fellow working in social science I’m often struck by how we make a statement in our limitations along the lines of ‘this study is observational, we cannot infer causality’ and then make an implicitly causal recommendation like ‘support mothers with mental health issues’, ‘don’t smoke’, ‘eat less fatty food’ or ‘school attendance is good for your grades’ (incidentally, these are all likely sound recommendations and it’s really our tradition of denying causality as a matter of course that’s the issue.))
Pearl is the originator of the Directed Acyclic Graph (the DAG), that is causal graphs, and a formal logic of causality. He is a relentless evangelist for these ideas. He has converted me to his religion (Judea-ism?) but it’s important to recognise that he offers a particular perspective on the issue of causal inference. There are other views out there (particularly in economics) that differ on some issues with Pearl (if/ when I come up with a good summary of their issues with Pearl I will add it as an appendix to the review).
Pearl does a good potted history of statistics, science and causal inference, with a lot of love for Sewell-Wright and his guinea pigs. He devotes a chapter to an overview of the Bayes rule and it’s applications. Including the Monty Hall problem, which unfortunately still confuses me (I don’t blame Pearl for this).
The book itself makes extensive use of causal diagrams to help build the reader’s intuition. This covers off a more systematic approach to selecting which covariates an analyst should (and shouldn’t) adjust for, and the language of common causes and common effects. He also gives an accessible review of Simpson’s and Berkson’s paradoxes.
Using causal diagrams offers an accessible tool for communicating instrumental variable and Mendelian randomisation analyses.
Pearl thinks about causal inference in mind-bogglingly abstract terms. The weakness is that (until now), it’s been left to others to help communicate his ideas. The strength is the sheer power and imaginativeness of his approach. Pearl offers up several extensions of his work that I was less aware of from the work of others. In particular, he is a strong advocate of the front-door adjustment method (basically the piece-wise synthesis of causal models from separate studies).
Another ‘innovation’ from the book is Pearl’s way of thinking about the problem of ‘transportability’ (I’d always called it generalisation) – how do we apply results from one context, population or setting to another? Again, Pearl uses the causal diagram to communicate his ideas.
As behoves his background in AI and Cognitive Science, the book is also rich with speculation about intelligence and consciousness (human and artificial). I found all this entertaining and thought-provoking. He contrasts his approach to the Big Data approach, but also proposes a marriage between the approaches.
I’ll give the caveat – I didn’t come to this book cold. I’ve worked in research for about 8 years with a bachelor’s degree in psychology and a Masters in Applied Stats. Over the last year I’ve been to several short courses on this issue, and I’ve done a lot of reading on the topic. Despite that, I’d recommend this as a good place to start (possibly in conjunction with working through all the examples and quizzes on DAGitty.net, and Miguel Hernan’s excellent HarvardX course).
It is not a “how to” for applying causal models to a specific analysis. Depending on the reader’s specific needs and interests, there are a few good resources out there but I quite liked Bill Shipley’s Cause and Correlation in Biology. The DAGitty web package is also excellent. Tyler VanderWeele’s book. OR, cross over to the dark side and look into econometrics.
******
Critiques of Pearl (work-in-progress)
The criticism of Pearl — He’s partisan. In presenting his history of causality, he emphasises his own contribution, and de-emphasises the contribution of others.
This is true. This doesn’t diminish his approach or the utility of his methods, but if you want balance, you’ll need to shop around for other points of view.
A few other reviews note that causality in social sciences is maybe not the new thing that Pearl argues it is. It’s an interesting area – economists have been on the causality bandwagon for years – but the other social and behavioural sciences are replete with examples of poor choice of control variables reflecting a lack of causal reasoning (and the explicit denial of it).
Pearl reviews examples where we would accept causality has been proven without his formal causal logic (e.g. smoking and lung cancer, John Snow’s cholera studies). What he communicates well is that, without a clear causal logic, it quickly got messy.
From where I’m sitting the causal diagrams offer a tool of communicating with other branches of the social sciences, and also offer a useful means for interrogating the assumptions behind economic causal models.
Don Rubin (the king of missing data and potential outcomes), disagrees on the value of causal diagrams. I’ve only seen the argument well articulated from the Pearl camp, and I’m not sure if this is purely a matter of ‘who owns causality’ or if there are worthwhile lessons in here for practicing scientists/ analysts (to be continued...).
There is also a back and forth series of letters in the International Journal of Epidemiology between Pearl on the one hand and Nancy Krieger/ George Davey-Smith on the other hand. The latter camp are arguing for a more pluralistic approach, and point to some instances where using DAGs and DAGitty has produced some implausible models. This intrigues me. I’ll add some notes on this later.
Similar review to book 1. Introduces some more memorable characters and the characters from book 1 evolve a bit (sometimes a bit inexplicably). Still mSimilar review to book 1. Introduces some more memorable characters and the characters from book 1 evolve a bit (sometimes a bit inexplicably). Still manages to have you invested, even in characters you should find unlikable. Without any spoilers, I like how magic is realised in this universe....more
Excellent (although I bogged down in the last few chapters). Does a good job going between the causal model, and statistical thinking. If there is a betExcellent (although I bogged down in the last few chapters). Does a good job going between the causal model, and statistical thinking. If there is a better resource for thinking about causal modelling, someone let me know. ...more
Ok, so the only reason I bought the book is because I installed the PROCESS macro, and when you go to the hep doc it just tells you to buy the book. TOk, so the only reason I bought the book is because I installed the PROCESS macro, and when you go to the hep doc it just tells you to buy the book. This seemed a bit mean to me. That said, book is well written, with some nice insights into how the author thinks about statistics, and it's a good introduction to mediation-moderation....more
OK to review this book, I first need to define a genre. Then I’m going to put this book into an adjacent genre, and define a grand unified theory for bOK to review this book, I first need to define a genre. Then I’m going to put this book into an adjacent genre, and define a grand unified theory for books about learning (my GUTBAL). Then I’ll do a quick review.
1) New genre – learning porn.
This is watching someone else do it (learning, you animals). The more exotic, an unrealistic the better.
Tim Ferris at least sells his stuff as “meta-learning” (teaching you how to learn) but I think it’s really just watching him do stuff.
2) A grand unified theory for books about learning (my GUTBAL). OK, there’s actually a spectrum of books about learning. On one end of the spectrum we have learning porn. Watching other people do it. On the other end of the spectrum, we have basic learning materials (textbooks, manuals, instructional). Doing it yourself. Somewhere in the middle we have books which are a bit of both – a bit about learning something specific, a bit about learning in general, and maybe some popular science thrown in.
I’d maybe put Moonwalking with Einstein in this category, and this book is definitely in that middle zone.
(OK, not the greatest GUTBAL ever)
3) The review itself. Good book. It’s only a little bit about maths and science, and a lot more about learning, problem-solving, different modes of attention, distributed practice, and dealing with procrastination.
I thought the cognitive psychology-lite was very good, and it was mixed with practical tips from some very good educators.
If you gave this to someone in last year of high school, or first few years of uni AND they implemented the suggestions, you’d be doing them a huge favour....more
The book is a sort of mash up of epistemology, statistics, and general musings. I thought early on, it was going to be more tightly focussed on truth The book is a sort of mash up of epistemology, statistics, and general musings. I thought early on, it was going to be more tightly focussed on truth versus truthiness, then I thought it was going to be on the Don Rubin causal model, then I thought it was a manifesto for modern data science. It was a bit looser than this, but an enjoyable read nonetheless.
The author has a particular interest in education, and this informs many (but not all) the examples.
If there were themes tying this together, other than an epistemological bent to his statistics, it is the idea of always starting with an idealised experiment, then working back to reality, and a focus on clear thinking and communication.
I even had that deceptive sense of understanding the 'counter factual' while I was reading the book, although sadly that passed.
This isn’t bland. The author’s voice comes through clearly. He has a great knack for quotes and turns of phrase that support his interests.
“Dear God, make my enemies ridiculous” Voltaire. “I will listen to any hypothesis, but on one condition – that you show me a method by which it can be tested.” von Hofman
Its another one that gets a worthy spot on the bookshelf, and would do with a bit more note-taking. It's not so technical that the interested generalist would necessarily struggle through.
BTW truth in the book’s case is belief backed up by evidence. Truthiness is when you believe something based purely on the feels. He has a bit of fun with this....more
Started reading this when I joined Institute for Child Health Research. It's not overly technical, but it keeps you focused on the basics - ask the riStarted reading this when I joined Institute for Child Health Research. It's not overly technical, but it keeps you focused on the basics - ask the right questions and use lots it pictures. Excellent training wheels....more
The basic idea, that statistics is an argument, where you need to logically make your case and defend it, is excellent.
I agree 100%, and I think statThe basic idea, that statistics is an argument, where you need to logically make your case and defend it, is excellent.
I agree 100%, and I think statistics needs to be taught in the context of scientific process and epistemology.
Ableson is entertaining enough, and his guidelines (Magnitude, Articulation, Generalisation, Interest, and Credibility of argument), plus his 8 laws are excellent (esp. '1. Chance is Lumpy').
Somehow, the execution didn't do it for me, but I can't put my finger on what it is....more
Nice overall read. The authors are good story tellers, and the best bits are excellent, e.g. real estate agents, Sumos, school teachers, and abortion.Nice overall read. The authors are good story tellers, and the best bits are excellent, e.g. real estate agents, Sumos, school teachers, and abortion. Good story about why cars driven off the lot lose value. The book sometimes comes across as a bit of a breezy magazine-style analysis “Levitt is an economist not like other economists, he’s got long hair, doesn’t like maths, and likes to ask interesting questions.” (cue food channel voiceover). In some cases the analyses are a bit weak or light-on, e.g. the position around incentives sometimes seems a bit tautological – someone did this, therefore they must have an incentive… I’m not sure that the ‘incentives’ argument is even falsifiable. In the case of crack foot soldier and taking risk, ignores the influence of culture and assumes it’s all about maximising ‘ROI’ – that people risk their lives for a big payoff. I think some respect for anthropology or sociology could have helped them out. There are a few other critiques, but if you take it for what it is – 50% education, 50% entertainment – and you consider it really kicked off a boom in this sort of popular analysis, it needs to be given it’s dues. ...more
Written by academic economist/deputy shadow treasurer/ general hotshot, Andrew Leigh. Tries to build a series of arguments around the idea that politicWritten by academic economist/deputy shadow treasurer/ general hotshot, Andrew Leigh. Tries to build a series of arguments around the idea that politics is more like poker than chess. Cleverly defines luck as something out of your hands (thus avoiding going down a philosophical wormhole on randomness), and has quite a few convincing examples to prove that ‘luck’ is a factor. Leigh’s lefty/ labour point of view comes through pretty strongly – life can be unfair (like a lazy Norwegian will live better than the average hardworking Nepalese) and based on that we should be egalitarian and look after the less fortunate. I agree with that, strongly, but I think a bit more time trying to talk about non-luck factors would have led to a more balanced discussion. I think where his politics shone through a little too brightly for me, was in every alternative scenario he poses the world would be better if only Labour got in. Not all the material ties closely to his thesis but the material on hyperpartisanship and prediction was compelling, at least to me. Despite my quibbles, I enjoyed it so much I ended up forming a bookclub just to talk about it some more....more