Customer Reviews

Making Sense of Statistics: A Conceptual Overview

5 star:		(1)
4 star:		(3)
3 star:		(1)
2 star:		(0)
1 star:		(1)

 

 

The critical reviews of this book all note its shortcomings, but without understanding its purpose. This is not a stand-alone statistics text. If you try to use it as such, all the criticisms of other reviewers are applicable.

This is a supplemental book, designed to be used alongside a statistics text. That is why it has no equations--those are already found in a stats text. This book is a useful addition to stats texts for students who find the formulas overwhelming, and for those cases where the texts do not explain the concepts intuitively. The latter is precisely what this book provides--intuitive explanations for the mathematical concepts behind statistics.

It is not a perfect book, which is why I can't give it a full five stars. But it is quite useful for its particular purpose, and the critics either have inadvertantly not recognized that purpose or, being math snobs with no patience for the less mathematically adept, they have intentionally (but wrongly) disdained that purpose.

In summary, this book is useful for those who need to take statistics but are not wholly comfortable with math.

While this book does conform to the title segment "A Conceptual Overview", I am not sure that it satisfies an existing market niche. I have been teaching basic statistics at the college level for decades and am used to equations and formulas in a statistics book. Therefore, I will admit that my background may be causing some bias in my outlook.
There is only one formula in this book, the expression for standard deviation and it is in an appendix. And unfortunately, it is not well presented; the sigma notation is used without being well explained. The coverage is thorough; all of the topics that I cover in my basic statistics class are in this book. But covered in such a superficial manner that I find it difficult to believe that the reader will really be learning anything about what statistics really is. I don't see how it is possible to really learn what a mean is without seeing it in formula form.
The only people that I could possibly recommend this book to would be those who have no idea at all what statistics is and what it is used for. Other than that, there seems no other place where it can be used.

This was a book I needed for Business school. It's much cheaper here new than a used one was at school.

expensive for a paperback but I guess it is in high demand. Book was in very good shape. Shipping was average in speed.

This book is primarily designed for psychology and other social sciences students, so that they can understand the statistical information in reports, etc., that they may read. It is based on two premises: (1) that a little knowledge is a good thing, and therefore the smaller the knowledge, the better the results; (2) that the best and more significant learning takes place without any 'doing' component; simply reading the text is enough.

Unfortunately, the book calms the unknowning reader into a false sense of security. A very common phrase in the book is "this is beyond the scope of this book," and this is used to dismiss even simple extensions and all calculation. The examples used are trivial and simplified, because anything more might require a calculator and some understanding. The reader is led through some interesting areas of descriptive and inferential statistics, but unless they go somewhere else, they won't learn anything of any significant value. Some terms are used that in proper statistics are now either outdated or given alternative names, but nothing seems to back them up. For example, Gosset and the t test are mentioned, but the common statistical appelation of "Student's t test" is not. ANOVA is introduced, but not how to get the results, nor (really) how to interpret them.

The second premise is perhaps the worst. Unless a student actually crunches some numbers, they will never appreciate how these things work, and in particular the risk of errors by the user. Too much statistical analysis today seems to depend on stuffing all the data into a package, such as SPSS, and looking at the print-out, and only applying some thinking at that point, rather than at the experiemental design, data collection and data entry stages. This book supports this approach in the reader.

If people today are scared of numbers, especially quantities of numbers, surely giving them the tools to deal with those numbers and make sense of them through understanding them is the preferred solution, not the analgesia of non-thinking that characterizes this book. If they were in the engineering or physical sciences, tools and understanding would be the solution adopted. Why is this not the case in the social sciences? Is EVERYONE in these fields both ignorant and terrified of statistics, and numbers in general?

The entire design and philosophy of this book is back-to-front and ultimately counter-productive to the education of people who can actually make sense of statistics. It smooths the pillow of the dying intellect and promises a better life in the next world. It breeds a group of people who drive statistics packages without understanding statistics, while feeling that, somehow, they actually do. It makes the statistical impoverishment of researchers in the social sciences more certain, rather than less.

Because there is no significant level of mathematics involved, readers must learn terminology, rather than what's happening with the data. Instead of an understanding of the fundamentals, the reader learns to identify different statistics and their symbols, different sampling methods by name, etc., but there are no logical connections made, no sense of what statistics is really about in this field. To add to the student's difficulties with a terminology/definitions-based learning strategy, there is no index in this book!

The ideal book for meeting the need for basic statistical understanding is Darrel Huff's "How to Lie with Statistics", which is basic, comprehensive, sceptical and almost devoid of calculation. It actually serves to increase statistical literacy at a basic level, and does not try to extend itself beyond dealing with basic descriptive and inferential statistical problems.

While Huff's book is great for a sceptical analysis at a general level, such as a technical report or paper's conclusions, unfortunately it is of little use in trying to analyze the details of the statistics spewed out in those reports and papers in the social sciences. Pryczak's book attempts to deal with this, but fails miserably on all counts. Readers have no overview, despite the title, of the bigger issues or the need for scepticism and common sense. The explanations are too comfortable and comforting to suggest scepticism of statistical analysis. Readers have no means of accessing the meaning of the statistics, because there is no attempt to deal with real data. Readers never have to DO anything, beyond read the text and look at trivial, doctored examples. The book presents no help whatsoever to a student in a course where even basic statistics have to be derived from some data, e.g., using Excel and its Data Analysis add-in package.

While Huff spends some time on a rigorous interpretation of 'average,' Pyrczak cannot even manage this without waffling, and fails to explain how to calculate a mean, median or mode ("it will be all right, don't be frightened, the horrid statistics won't hurt you, be calm, trust the software package or the other person to get it right, it will be all right"). A formula for the standard deviation is tucked in an appendix. Pyrczak gets into a tangle over standard deviations and standard errors, not being clear about either, really. And from here things go downhill, but the comfortable, comforting tone remains ("In know, statistics is SO hard, but this will allow you to forget all that and feel OK again. You won't be challenged by it if you don't want to be (and who would ever WANT to try to actually understand statistics?).").

In summary, this book is a waste of ANY student's time and money. It is unfortunate that it is a required text in some courses, but that fact probably explains more about the statistical weaknesses in the social sciences than anything else (r^2 > 0.9). This book is actually counter-productive to education in statistics.

Recommendations: (1) If it's a required text in a course you are taking, I pity you. Get a proper statistics text so that you can actually learn something useful.
(2) If you are thinking about using this as a text in a course you are offering, actually look at the book. If you think it's good and useful, get someone else to run your course--your course will not help anyone's statistical understanding (including yours); in fact it will do quite the opposite, without anyone realizing it.
(3) If you are simply interested in learning about statistics, don't even consider this book.
(4) Amazon needs negative star ratings, as this would be about -3 (90% CI of 0.1).

The book arrived two days early. It was in great condition for a used book. I would not mind buying more of my school books from Amazon.