Meet Peter Swanborn. Peter is considered to be quite the expert at methodologies of sociological research. Peter is also the author of the manual for research methodologies ('Basisboek sociaal onderzoek') on which I will have an oral exam on Thursday.
Boy, is the professor in for a treat.
I've been annoyed by Swanborn's book ever since he failed to differentiate between a theory, a theorem and a hypothesis in chapter 1 (he sets things right later on, but that's too late to my taste), and doesn't distinguish between a 'causality problem' and an 'explanation problem' (proving cause and effect can be done without offering any explanation whatsoever - in the end it turns out 'explanation' is just his phrase for 'causality').
It took me until the final chapter, however, to conclude that this book needs a dramatic overhaul.
It's not just that Swanborn can't be bothered to look up the name of famous statistician Maurice Kendall, who is adressed as Kandall all the way through, or that 'Kandall' is credited with having invented the statistic variable Q, which was the work of George Udny Yule. Kendall invented a method to calculate rank correlation coefficients, which is known as the variable Tau.
I can forgive that. After all, 'q' and 'tau' are only a few letters apart in the alphabet. Granted, they're two altogether different alphabets, one being the Latin version, the other one being the Greek, but hey, Rome wasn't burnt down on one day.
However, it's when he starts doing math when I find myself unable from trying to claw my own eyes out.
Here's one example, from page 275:
'(213/212) / (221/334) = (213/212) * (334/221), or 213 * (334/212) * 221 = 1,52'
What he means, of course, is:
'(213/212) / (221/334) = (213/212) * (334/221), or (213*334) / (212*221) = 1,52'
Mind you, this is the 'fourth, completely revised edition'. Obviously, the revision process at 'Boom' Publishing House ('Boom' means tree, which is a pity, because after reading this book, I feel they might benefit from a huge explosion) is managed by people who graduated summa cum laude in French reductionist mathematics, which teaches that calculations which yield correct results are examples of neodialectic western cultural imperialism, or something to that effect. Either that, or they're retards.
Anyone who can't spot the error in said calculation is kindly requested not to come near me, at least not tonight, unless you're Scarlett Johansson, or a clone thereof.
(Note: for clarity reasons I've added parentheses in the transcription of the calculation above. Swanborn doesn't bother with them, he just hits the space key a few more times. Also, I've substituted 'x' with '*'. Sure, any first year student should know you don't use 'x' for multiplying since it's also the most common variable in existence, but we're dealing with a sociologist here.)
Nitpickers can click 'Lees verder' for some additional information.
What he literally writes, for all you nitpickers out there, is:
"213/212 / 221/334 = 213/212 x 334/221, of 213 x 334/212 x 221 = 1.52".
I can see how the mistake was made. In the first two sections of the calculation, he uses spaces to make up for his lack of parentheses, and doesn't use them when he wants to indicate a fraction.
Then in the third section, all of a sudden the absence of spaces between 334, / and 212 don't indicate a fraction anymore, and the presence of spaces between 213, x and 334, and 212, x and 221, should not be construed as parentheses anymore either. Which is funny, because the book doesn't come with a 'Peter Swanborn Mindreading Hat'.
Swanborn is inconsistent, or lazy, or both.