R for Marketing Research and Analytics (Use R!) 2015th Edition

There are many things to commend about this book.First and foremost, this is the first major and successful attempt to present analytic techniques to marketing researchers from a modern perspective. It replaces the standard multivariate technique books used by marketing researchers (starting with Paul Green in the 1970s and ending with the currently in print Hair, Tatham et al.). There are some excellent contemporary books on analytics relevant to marketing researchers such as "An Introduction to Statistical Learning" by James, Witten, Hastie and Tibshirani, and "Applied Predictive Modeling" by Kuhn and Johnson. However, they are not directly designed to address marketing research issues. This book is. The advantage to this book being specific is that it can address the problems specific to marketing researchers rather than dealing with such issues tangentially.Second, Chapman and Feit do not deal with marketing research problems from an academic perspective with artificial scenarios. Their examples are of the type a marketing researcher would deal with on a day-to-day basis. When I first saw that the authors use generated data as opposed to real-life data to illustrate the techniques, I had misgivings. It is easy enough to create contrived datasets to solve imagined problems; in real life, datasets are not always that cooperative. The authors have not fallen into this trap and they have generated datasets skillfully to illustrate the points they are trying to make.Third, as a modern take on traditional bivariate and multivariate techniques, Chapman and Feit present Bayesian methods, which are becoming increasingly popular. I believe Bayesian methods (especially with the advent of R) will soon be part of mainstream data analysis in marketing research. The book includes sections on many relatively newer (in any case, less frequently used) techniques such as random forest and naïve Bayes.Fourth, in several places Chapman and Feit explore the implications and extensions of basic techniques, which I have not found in other comparable texts. As an example, while discussing factor analysis, they discuss how to use factor analysis to create perceptual maps.. Such extensions are seldom discussed explicitly in other texts dealing with factor analysis.Fifth, the book is comprehensive. It covers all aspects of analysis a beginning or intermediate marketing researcher or analyst is likely to encounter. Although initially I wondered if it was necessary to devote a third of the book to basic statistics and R, it does provide a good foundation for data manipulation.Sixth, the writing is clear. This is not a technical book and it is not meant to be. This makes the book widely accessible to marketing researchers with different proficiencies in mathematics. I also liked the fact that Chapman and Feit point out the limitations of traditional techniques like confidence intervals.Finally, the authors do a good job of teaching the R language and graphics to beginners. The book is not unique in that respect because many other books do an equally good job when it comes to teaching R and graphics.Some standard techniques (neither numerous, nor serious) are missing from this book. A case in point is linear discriminant analysis. While logistic regression (which is included in the book) can be seen as an alternative to LDA, there are several instances where LDA is a better alternative. Other missing topics include correspondence analysis and maxdiff. But it is the authors’ prerogative to choose what goes into their book and Chapman and Feit’s coverage is comprehensive enough for most purposes.While the authors do indirectly bring up validation issues and deal with them, they do not treat validation as a systematic and explicit part of using any technique. They devote less than a single page to the widespread problem of overfitting and touch upon bootstrapping only minimally while discussing PLS/SEM. I am not sure if they discuss bias-variance tradeoff and cross-validation seriously at all. I believe, as we move into the era of big data, samples drawn from an unknown population, do-it-yourself research and the like, validation issues become critical and they should be a part of any analyst’s thinking. Most users of the techniques know much more about “R-squared” and “number of hits” etc., than about the perils of overfitting, about model bias or about the reproducibility of the results. For many decades we had no alternative. Programming was complicated, datasets were small and computer time was expensive. Now we don’t have any of these limitations and I believe validating results should not be optional or an afterthought but an integral part of data analysis.Despite the title, which emphasizes R, the book is more about data analysis and analytics. "Data Analysis and Analytics for Marketing Research With R" would have been a more appropriate title for this book. The book has a lot to teach about analysis whether you are interested in R or not.While I wish the book had dealt more systematically with validation issues, what it does it does well. Beginning and intermediate researchers who need to analyze data will be hard put to find a better source than this book; learning R in the process is a big bonus. I highly recommend this book to beginning and intermediate researchers seriously interested in data analysis and analytic techniques.

If you’re a marketing student or a marketing/analytics professional and you want to learn R, then this book is perfect for you because it covers R specifically for marketing research & analytics. While many how to code books may be difficult to read or comprehend, this particular book is engaging and comprehensible.The book is easy to comprehend because it uses detailed graphics and clearly designed blocks of code to help you learn R. You will learn hundreds of commands and their specific applications. You’re not just reading code from the book and typing it out, instead the book helps you understand every line of code you write by teaching you the functions/outputs of every command. This style of teaching will help R become more intuitive to you as you progress through the book.The modern take on R makes this text highly engaging because it uses examples of modern day research techniques like• Analyzing data from social media platforms• Analyzing customer survey responsesThe book will teach you how to perform data analysis with R as well. You’ll run dozens of statistical models that will help you derive information from large datasets. Not only that, but you will visualize your data by creating interactive graphs such as• Quantile plots• Scatter-plot matrices• Correlation plotsLearning R will take time for you if you’re completely new to coding. This is why the first three chapters will teach you the fundamentals of R. It may seem tedious to you, however it’s essential to learn these basics in order to understand the following chapters. Fortunately, the first few chapters are short and engaging, so you will breeze through it in no time and perform the codes in later chapters.So if you’re interested in R for marketing research & analytics, then I highly recommend reading this book. It’s straightforward, engaging, comprehensible, and you’ll distinguish yourself from your peers in the job market. Investing in this book is a great investment in your future careers as marketing researchers & analysts.

If you're the kind of R learner who wants to experiment and play around with datasets so as to get a feel on how the output of your analysis varies when you tinker with the input, then you'll be in for a treat with this book. I truly enjoy the authors' approach of teaching the reader how to simulate his/her own dataset. IMO, the simulation process addresses two major things:1. It helps you understand better the concepts of R when you build a data frame/ dataset from scratch - from simple vectors to a workable data frame using practical and essential functions along the way2. It enhances your appreciation of the value of the analytics tool you're using because you get to understand how the inputs were set-up and how it affects the end resultI'm not a marketing researcher so I'm basically using the book as an overall-business-analytics-reference material.

This is a great book as an introductory text to market research. The writing is clear and there're dozen of nice examples. The only problem is that the book fails to address to unequal sampling. In marketing, we tend to create a complex survey design. There is no mention at all. Otherwise it would have been 5 stars.