Book Review: "When"
2/5 stars
Daniel Pink
"The number of books yet to be written by nailing together weak psychological papers is only bound by the imagination"

*******
I'm just not sold on this book, for a lot of reasons.

PROBLEM 1: 

If these were the best examples that he could find out of hundreds of papers, was the effect really that big?

First example is the graph on page 171, showing a change in divorce filings. This uses an old trick (that is a textbook example in books debunking statistical errors) of "trying to make something both bigger and smaller at the same time." The range from the highest to the lowest is about 0.3 filings, or 10%. But it's presented in such a ways to make it look *much*  bigger.

Second example is on page 120, and it has absolutely *no* units on the ordinate axis. I don't think I've ever taken a statistics class where the instructor would accept a graph with no units. 

Third example is on page 10, and it is the same thing that's the first example, except that the entire range of the graph is 6%.

Fourth example is on page 148, and this time the graph actually *does* have units. And the units are "per 500 runners." So 27 out of 500 (5.4%) runners will run their first marathon at 29, but then 14 out of 500 (2.8%)  will at 28. So, the distance between the highest and lowest scores was an eye-watering 2.6%. In Statistics classes, that is known as "statistically significant, but not practically significant."

And that's even if we are willing to assume that these many papers have external/construct validity; remember that in that pseudoscience known as "Psychology," things vanish without a Trace once people stop believing in them. (The author himself debunked the notion of "midlife crisis," and I'd have to say that that was the single most useful thing that was brought forward in the entire book.)

PROBLEM 2:

The author talks about the benefit to starting schools later in the day, but this is yet another example of academics being completely out of touch with reality. (And that's about as rare as hydrogen.) The bad news is that when you have to get up and go to work everyday and pay a mortgage, the work schedule is not arranged for your convenience.

If somebody paid attention to this book and implemented his suggestions in school, it would just be ONE MORE thing that you had to unlearn as you left school and went into the real world. (You know, that place where bills have to be paid without a lot of discussion?)

PROBLEM 3:

Some of this is empirically false--and that's even worse than weak statistical effects or low external validity.

1. (p.199) "Operating in synch expands our openness to outsiders and makes us more likely to engage in prosocial behavior. In other words, coordinating makes us better people - - and being better people makes us better coordinators" 

I'm sorry, but I happen to live around some people that have an, um, very interesting interpretation of what it means to be a human being. (That would be Haredim.)

To a man, these people are all dressed in the same (penguin) clothes and read from the same books and live basically interchangeable lives. But, to say of them "open to outsiders" and "being good people" is probably about as accurate as describing R Kelly as an example of "moral rectitude" and "conservatism."

2. Lots of black Americans try to sync with Arab Muslims by changing their names to Arabic names and wearing silly desert garments. Given that I have seen exactly zero such interracial couplings, I don't think that it has been successful.

3. (p.120). Hanukkah candles. He seems to think that a significant fraction of Jews are two trifling to even light all 44. It's definitely a comparatively low energy Jewish holiday, into light candles over eight nights is really as easy as breathing.

PROBLEM 4:

A lot of stuff that the author tells us is something that anybody who is of a certain age (let's say over 30) should already know.

1. Of course if you want to take advantage of something, that it is a matter of being in the right place at the right time.

2. Of course if you have a full schedule, then you will organize your time of necessity.

3. Of course if you operate a business and you do it all day, then you will become successful at what you do as a matter of course. And if you don't, then you just won't survive. He wrote a chapter about the Indian delivery service, but that's pretty much what you would expect in a country where wages are low and cheap labor is abundant. (If they're so good at that, why can't they seem to figure out anything else? Like "synching" enough people to attain 99% literacy. 

Is there really information here?

*******
This author has 325 references in a 200 Page book, but the graphs that he showed us actually showed pretty weak effects.

Really, the most use of this book is as by going through and picking out a lot of the most glaring statistical errors. 

At least it makes you think that you learn something from other books that you've read about statistics.


Second order thoughts:

1. I guess if you read enough trivial psychological papers, you can staple together any arbitrary number into a book about almost anything. That has certainly been the case with books such as "Invisible Gorilla"/"Predictably Irrational"/etc.

2. This book sold quite a few copies. "Fifty Shades of Grey" also sold 15.2 million copies, so there's that.

Verdict:

Not recommended. I'd recommend three other books in preference to this one:

1. "Thinking, Fast and Slow" by Daniel Kahneman (the great majority of this book is derivative of that, to be frank).

2. "Inumeracy," by John Allen Paulos and it is a discussion of statistical errors such as this one.

3. "How to Make The World Add Up" by Tim Harford, which is similar to the second but more recent and more popular.