To get the free app, enter your mobile phone number.
What is a p-value anyway? 34 Stories to Help You Actually Understand Statistics 1st Edition
Use the Amazon App to scan ISBNs and compare prices.
Frequently Bought Together
Customers Who Bought This Item Also Bought
Top Customer Reviews
The little monograph by Andrew Vickers is absolutely a fun book to read - and how many times can it be said that a book about P-values is fun? My husband is often subjected to my very serious face and quietness as I read a statistics book. He often jokes that I look like I'm angry or in pain. But with this book, heard me laugh out loud, and do so often. Andrew Vicker's "What is a p-Value Anyway?" contains 34 short stories that wittily illustrate the correct way to think about and apply statistics in terms everyone can relate to and understand.
Vickers manages to put statistical concepts into a fun, understandable light that is entertaining and informative, no matter what your level of understanding and education in statistics. I highly recommend this book.
So did a friend of mine, living in Milan. Early in the morning he called me. His voice sounded, at the same time, angry and depressed. "How can I be so underpaid? I downloaded the incomes of people living in Milan and computed and re-computed the average all the night. I discovered that my salary is very, very low! The mean income of people in Milan is much higher!!" "Well, that's the Berlusconi's effect", I replied, "don't worry, take the median instead of the mean, and you'll recover your self-esteem... and stop your wife's complains".
But I wasn't persuasive enough...at the phone, without paper and pencil, I couldn't explain him the impact of outliers and the meaning of skewness, and he started looking for another job.
Had this happened now, I would simply send him a copy of "What is a p-value anyway", and a couple of pages at the beginning would explain him the concepts, using Bill Gates instead of Berlusconi. The concepts and ideas of statistics are indeed the focus of this book, from basic ones like median and mean, to more advanced ones like hypothesis testing, regression, logistic regression, survival analysis etc. These concepts are explained, without formulas, by means of funny stories, that you can tell to your friends at a party, or use with your students to raise their attention. Moreover, stories stay attached to our neurons, and we, or our student, don't forget them and their embedded concepts.Read more ›
Since then I’ve found Geoff Cumming’s book (Understanding the New Statistics). Spend the extra bucks and get it. He’s entertaining (great visuals) and will really teach you what you need to know about doing sound statistics, and no p-values.
If you care to bear with me, let’s talk about p-values.
If you’ve taken an intro stats course you were taught to do your stat, state a “null hypothesis” that your result is caused by chance, then find the p-value (computed by the software) to see if it was caused by “chance”. If p is > 0.05 (or 0.01) your result is “statistically not-significant”, and you lose. If the p-value is <0.05 your result is “statistically significant” and you win. Vickers knows this is a big problem. In chapter 13, he explains how null hypothesis testing (p-values) doesn’t really tell you what you want to know. (A “null hypothesis” is that there is no difference or nothing happened.) In chapter 14, he waffles on about it. In chapter 15, he shows how a LARGE p (>0.05) doesn’t actually tell you that your result is “insignificant”.Read more ›
In any case, here are the key ideas from the book which I found noteworthy:
(1) As measures of central tendency, both mean and median values have their use. Medians are often more useful when there are outliers, whereas means are often more useful for decision-making. Using both together will sometimes also make sense.
(2) Normal distributions describe many natural phenomena, as well as the additive effect of many random events (central limit theorem). However, many phenomena are limited to non-negative values and involve multiplicative effects, so the resulting distributions are non-normal and right-skewed; sometimes these distributions can be converted to normal by taking the log of the data.
(3) With normal distributions, about 50% of data falls within 2/3 of a standard deviation, 67% within one standard deviation, and 95% within two standard deviations (the typical confidence interval). When data are highly skewed, medians and percentiles (eg, quartiles) are often more useful than means and standard deviations.
(4) Study bias can come from many sources during study design, conduct, and analysis.Read more ›
Most Recent Customer Reviews
Excellent book. Andrew Vickers tells the basic statistical concepts clearly and humorously.Published 4 months ago by Ercüment Yerlikaya
Book disappoints. Whoever edited this book did a bad job. On page 67, there is a typo. The X axis says 100, 50, 0, 50, 100 when it really should be 10, 5, 0, 5, 10. Read morePublished 10 months ago by C. Welch
Tries to be funny, but is not half as clear in its explanations as it should be.Published 12 months ago by Roberto Perez-Franco
I bought this thinking I could use it to help students...it isn't written in a way that met my needs.Published 16 months ago by lms
Very well written for an excellent understanding of statistics.Published 20 months ago by Jeremy W. Gibbs
If you are learning, teaching, or using statistics, you should read this book. It describes what statistical jargon and tests mean and what they don't. Read morePublished 22 months ago by A reader
I never used it. Was a required text but kept it as it is a fun and useful resource when you need to use statistics. Read morePublished 23 months ago by MJR