Bests & Worsts Reviews from Amazon

The author has a recent Ph.D. from Western Michigan University and runs her own consulting firm. This book on data presentation, her second, is focused on elementary displays of data using MS Office. It joins an increasingly crowded market and does not stand up well given some excellent alternatives. If you have access to a copy and skim through, you may find examples of unfamiliar graph types that could be useful. That's the most likely positive use of this book I can imagine. In particular, I agree that what the author calls 'dot plots' (Cleveland dot charts is another name) are often very helpful and should be used more frequently. Similarly, 'slope graphs', an old idea under Tufte's new name, can be very good for showing paired changes as labeled line segments. The book mentions 11 technical reviewers, all from universities. I have to wonder whether they read the book, or perhaps more likely, how they feel about being named if their advice was ignored. Technically, this book is unreliable, and shows much evidence of the author's inexperience and prejudices. Evergreen is negative about scatter plots and lays that on repeatedly: they are a 'complicated visual' (p.161), can be 'confusing to interpret' (p.165), and are suited for '[m]ore sophisticated groups' who have taken an 'advanced statistics course' (p.172). I come from a different educational system, evidently, in which scatter plots appear early in children's education and are a staple of every introductory statistics course. The author's main scatter plot example shows % of people of color in various areas of New York City (y axis) versus number of military recruits per 100,000 residents (x axis). She repeatedly brushes off the standard point that the axes would be better reversed (a point also poorly handled in her first book). This isn't just a cosmetic choice: which variable is outcome (response) is crucial to thinking about such problems. Evergreen also fails to comment that the pattern of scatter is strongly nonlinear, so the straight line fitted is absurd for that reason, and indeed on other grounds too. Evidently, the author needed simpler examples that she could handle confidently. Similarly, Evergreen dislikes histograms which are 'not the sexiest' (p.140) and 'feel clunky to me' (p.145). These aren't serious comments, for all that she does continue with examples. The author is evidently no programmer (that's a comment, not a criticism) and focuses on how to get there step by step in MS Office (meaning MS Excel, mostly) using your mouse. Those sceptical of the merit of doing graphics at all within Excel will find their opinions confirmed by the dodges and fudges needed even to do some very simple things. There are some unsystematic comments on using macros instead, which are mostly a distraction. There is little mention of the many internet resources available from more expert Excel users. There is some token R code on the author's website, which looks like a waste of effort given the mass of outstanding code already available. There are other technical mistakes that should have been caught by reviewers. There is reference to debate on whether bubble charts should be sized by their diameter, radius or circumference (p.17), but there is no such debate. Any issue is over using areas or lengths, as any length basis is equivalent to any other. 'Upper confidence interval' (p.25) is a slip for 'upper confidence limit'. Cluster analysis is not defined (even vaguely) by combining quantitative and qualitative data (p.183), although that is one of its applications. Evergreen broaches the difficulty of paraphrasing technical terms for non-technical audiences, but some of her own suggestions fall far short of acceptable. 'Statistically, there's a chance the actual score falls in this range' is offered (p.26) as an explanation of confidence interval: that explanation is not just vague, it's completely vacuous. As other reviews note, no colors are used here except blue and grey. A side-effect of that, which is not quite inevitable, is that many graphs here depend heavily on minor differences between shades of those colors. It is hard to like this book if, as I did, you find its written style slangy and immature. Sample vocabulary includes amazing, awesome, baby, cool, cute, freak, geek (often), heaps, hush-hush, incredibly, kid, kinda, loads, nerd (often), ninja (oh so often), pow, rad, rock star (ditto), sorta, stellar, Super Academic, super bad, super cool, super critical, super famous, super hero, super honest, super interested, super weird, um, whew, wimpy, wow, yay, yeah, yep, yikes, and yowza. (Also, if it's news: using MS Excel for presentation graphics doesn't make you a geek.) Combined with that style come careless editing and proof-reading. There is repeated confusion between it's and its and bizarre typos on the graphs themselves, such as camradarie, kindney, ful (for 'flu') and individulized. There is also confusion between i.e. and e.g., affect and effect, axis and axes, rectangle and oblong, and disbursed and dispersed. Names of key figures in the field (Jorge Cameos, Mike Bostok) are mangled. No editor should have let through `the bottom of Indiana' and `the top of Kentucky' (p.229): the words being reached for might be southern and northern, or something more precise. In total, writing and editing were poorly done. I evidently don't know how blame is to be shared between the author and the publisher. I do find Sage's technical books a very mixed bunch. Better books? Berinato: Good Charts: The HBR Guide to Making Smarter, More Persuasive Data Visualizations. https://www.amazon.com/Good-Charts-Smarter-Persuasive-Visualizations/dp/1633690709 Cairo: The Truthful Art: Data, Charts, and Maps for Communication. http://www.amazon.com/Truthful-Art-Data-Charts-Communication/dp/0321934075 Camões: Data at Work: Best Practices for Creating Effective Charts and Information Graphics in Microsoft Excel. http://www.amazon.com/Data-Work-practices-effective-information/dp/0134268636 Knaflic: Storytelling with Data: A Data Visualization Guide for Business Professionals. http://www.amazon.com/Storytelling-Data-Visualization-Business-Professionals/dp/1119002257 Robbins: Creating More Effective Graphs. http://www.amazon.com/Creating-Effective-Graphs-Naomi-Robbins/dp/047127402X Camões is best of these if you need an Excel-based book. Robbins has the greatest depth of analysis.