Preface

Introduction to the Book

Welcome to Business Statistics Made Easy in SAS by Professor Gregory John Lee. This book is a breakthrough in business statistics learning, with a fresh new approach to explaining and teaching the exciting area of data analysis.
This book is designed as a user-friendly, practice-oriented text to teach businesspeople, students and others core statistics concepts and applications.
Business Statistics Made Easy in SAS steers away from complex mathematical-based explanations, and also avoids basing the explanations on traditional concepts such as distributions, probability theory and the like, which tend to lose the practice-oriented reader.
Instead of these traditional approaches, this book employs many features that have proved successful in a great number of MBA and other classes, some of which are completely innovative. These features include the following:
  1. Unique templates for understanding statistics: There are several chapters that present a process template overview of statistical thinking, in other words, they attempt to show the reader how the general process of statistical thinking works. These chapters are a completely fresh way to teach people how statistics works as a whole. Chapter 2 gives a process overview of statistics in general; Chapter 11, an overview of statistical analysis as a process; and Chapter 12, an overview of how to analyze a given statistic after it is generated. These chapters are mostly innovative, and enable readers to grasp the statistical method as a whole rather than learning this in the traditional way, in which a person is expected to learn individual techniques and concepts, and then piece them together to form the bigger concept of the statistical method.
  2. Extrapolation of statistics to business strategy, financial impact, and problem solving: I believe that one of the weaknesses of traditional statistics texts, from the point of view of the average business reader, is the lack of application beyond the statistics itself. I attempt to overcome this through two completely innovative chapters that take statistical outcomes and extrapolate findings to much broader implications, such as the profitability of business cases. This allows readers to answer the question “Why should I care?” that so often plagues statistics courses. I also take this approach extensively in practice questions.
  3. Binding case: At the beginning, in Chapter 1, I build a central, illustrative statistical case that most of the rest of the text builds on and uses throughout for illustration. This helps to focus the discussion and ground readers in a well-understood context.
  4. Practicality through SAS® Studio or SAS®9: The core idea behind the text is to focus on practical implementation through a specific statistics package, in this case SAS. The book has been written to work well for both the exciting new SAS Studio as well as SAS 9 (specifically I use SAS 9.4 or SAS 9.3). Extensive screenshots of each major step are provided to guide the user carefully through using the package, and the data and prewritten code used for every example are given with the textbook, which the reader can simply open and run to get the output. The reader can then usually change dataset and variable names to run the same programs on other data situations.
  5. Non-technical exposition: There is very little mathematical development, aside from some limited development in appendices for the interested reader. The text is developed verbally in a logical and clear way, often using metaphorical analogies that help the reader to connect the statistics concepts to life examples with which he or she is familiar. I avoid starting off with hard-to-understand terminology; instead I ease the reader into it from the perspective of what the terminology is really trying to achieve.
  6. Pictorial and metaphorical explanations: There is extensive use of unique pictorial and metaphorical explanations – not just diagrams but various different figures that help to explain the concepts. This proves exceptionally useful for teaching the less technical user, and adds completely new pedagogical features to many difficult sections, such as the one on the concept of power. I also include decision-making flowcharts for some techniques that have been hugely popular with readers.
  7. Case vignettes: I also have other illustrative cases of actual statistics applications in real life to help readers see the applicability of the techniques.
  8. Practice questions and datasets: I include a large number of practice questions and datasets. I also run online assessments that readers can take to self-assess or instructors might use to assess; these may be adjusted for the text.

The Book’s Use of SAS Programming

There are many several analysis and statistics programs available, each with substantial merit, including but certainly not limited to SAS, SPSS, STATA, EViews, R, STATISTICA, and NCSS. Different disciplines tend to develop their own favorites. This book specifically uses SAS because I believe it to have enormous and world-leading merit.

A Quick Introduction to SAS

SAS is not just one program. Instead it is a family of programs that mostly draw on a powerful central program. In this book we will use SAS Studio or the latest SAS 9 releases, which are based on keyword-entry programming. Two of SAS’s other programs (SAS® Enterprise Guide® and JMP®) are point-and-click interfaces. However, SAS has many other programs as well, ranging from specialist time series analysis to big corporate data analytics to matrix programming to geographic information systems.

Why SAS?

Here are several reasons for using and choosing SAS.
  1. SAS is the most powerful statistics package: There is literally no statistical or analytical technique that SAS cannot accomplish. First, SAS already has programs for most of the analytical techniques that students and practitioners can possibly need. Even if SAS lacks something, it can technically be programmed in directly by the user using modules such as Interactive Matrix Language (IML), although admittedly this takes a lot of technical ability. (I do it all the time.)
  2. Specifically, compared to SAS, other packages lack too much of what we need: I like other packages too, and respect the popularity, tradition and offerings of products named earlier. However, these packages lack too many things that business students, professors and practitioners need. Business is widely multi-disciplinary, ranging from the financial and economic-type disciplines to more social science disciplines such as industrial psychology or marketing. Business requires wide-ranging statistical techniques. In addition, more mathematical types of techniques like linear programming should be available, along with data analysis, if disciplines like operations and logistics are to be served well. SAS has it all, or the ability to make it, whereas I know of no other package that integrates such complete data manipulation (including major functionality for accessing databases and manipulating datasets), so many statistical techniques (including cross-sectional techniques, full time series modules, structural equation modeling, classification models and many others), and such powerful mathematical techniques (such as matrix programming, operational research, etc.). I should also say that other programs inevitably have certain things that are more conveniently provided than in SAS. For instance, SPSS has bootstrapping at the click of a button. However, the overall loss from using other packages is simply too high.
  3. SAS caters for all levels and types of users: There is a fallacy that SAS is difficult to use, because it stems from its programming roots. However, in reality, anyone can use the point-and-click interfaces such as SAS Enterprise Guide and JMP, which approximate the SPSS-type environment. The more serious users can try their hands at programming at various levels of complexity which, after all, is just a few keywords. In this book, I provide code files that eliminate the need for readers to learn programming from scratch anyway.
  4. SAS is deeply embedded in the organizational world: On balance, SAS is the package of choice for serious corporate analysts, as well as those in other organizations such as government. There are various reasons for this: aside from the above, SAS is unparalleled in being able to draw on and work with database technologies; it has serious scalability; it is very stable; and it offers specialized business analytics solutions. If you are using this book as a teaching tool, part of your concern should be the employability of your graduates, in which case you want to align your teaching with the practices of external organizations. Other packages have powerful organizational penetration too, but SAS is simply more widespread and geared up for the big data era.
  5. SAS skills are seriously marketable: A related note is that SAS skills add marketability to a CV. There is a thriving market for SAS analysts and, while I do not see all students or practitioners as being specialists, marketability nonetheless accrues on a level I have not seen with other packages.
  6. SAS has great support: SAS has fabulous support structures including:
    1. Massive reading resources: SAS has a large official knowledge library, both online and in print, supplemented by multitudinous unofficial online papers. This means that readers who need help can easily find a variety of examples, perspectives and the like. The SAS publishing arm provides by far the best practical set of books and manuals, eclipsing any other provider.
    2. Great helpfiles: SAS has what I think are the best helpfiles in many cases, including many worked examples for every technique.
    3. Serious e-learning: SAS has major online e-learning streams. This is a unique resource that allows professors to co-teach the more technical aspects of certain courses, and get external examination.
    4. Great teaching resources: SAS provides fantastic teaching resources (e.g. slides, manuals, case studies, etc.).
    5. Online and other communities: SAS hosts and has fantastic online communities that can help solve problems as well as frequent high-profile conferences and events.
I note as an aside that it is lack of such resources that hampers many of the providers. The stark example here is the freeware market: these programs are free and popular with academic statisticians as a result, but there still is lamentably poor support around their use. SAS itself has now released a major freeware product with SAS® University Edition, but backed by its formidable support. Readers need all the support we can muster.

A Note on Learning Statistics through SAS Programming

There are two predominant ways to run SAS:
  • You can use various point-and-click windows to perform tasks in SAS. This is relatively simple to use, and favored by most non-technical people. If you were using the point-and-click options, you could, for instance, open and use SAS Enterprise Guide instead of programming in SAS.
  • SAS – notably in Base SAS® - often uses programming code to input keywords that tell it what dataset to analyze, which variables to analyze, and what statistical analysis to do on these variables.
While point-and-click options are easier at first and more popular, there are several reasons why I base this book on the programming code of SAS:
  • Point-and-click is very limiting: All point-and-click programs, from SAS Enterprise Guide to SPSS, are very limited in what they have been programmed to do, with a large amount of the statistical universe either left out altogether or constrained to simple versions. While this may seem alright to a beginner for whom the starter topics are well covered in point-and-click programs, as soon as you or your organizational colleagues start to get more advanced, the limited versions become constricting and problematic. The simple fact is that users with expanded needs often then have to start accessing the programming back-ends of such programs anyway! While this may change in the future, starting your journey in SAS with easy-to-use programming code that has been written for you is the best way to make sure that later you are ready to go further on your own.
  • Programming is efficient: The programming code input method is very efficient and advantageous. It is far quicker than using point-and-click and takes far less memory, and also you can save programming code for later use. Finally, point-and-click takes a lot of time to go through if you are in a classroom teaching situation, whereas opening and running a programming code file is quick.
  • Saving and re-using programming code: You can save the keywords you used in a programming code file and re-use them time and time again (see for instance the programming code files in the “SAS Code” folder). Generally, once you have the keywords you like to use, the only thing you have to do is change the names of the variables.
This book mostly uses programming code. Because of the advantages of programming code, it is this type of input method I will mostly use and teach in this book. You will not have to learn what keywords to use, I will give you programming code files (see the “Textbook SAS Materials” folder), and each time we run an analysis you will be directed to open and run a pre-existing file.
If these reasons for using the programming-based SAS do not seem as attractive as using point-and-click, then consider looking at SAS Enterprise Guide as an alternative option.

Additional Book Resources

This book comes with substantial additional resources available through the book’s website at https://support.sas.com/publishing/authors/lee.html including:
  1. Microsoft PowerPoint slides: Well-formulated and animated slides are provided for each chapter.
  2. Textbook datasets and code files and programs: The textbook comes with pre-prepared datasets and program files. Each chapter is accompanied with prewritten code files which the reader can simply open and run to get the output. The reader can then usually change dataset and variable names to run the same programs on other data situations. In addition, there are two type of programs provided:
    1. Bespoke programs and macros written for the book: Some of the code files provided are an exciting set of programs written exclusively by myself to help readers get directly to the types of outputs you really want. For instance, in the regression chapters I provide macros in which you have to put only the names of the dataset and variables and a few other details to get advanced, formatted regression outputs (including nifty additions like p-value stars), robust regression including comparisons to the normal model, and bootstrapped regressions including again comparisons to the normal model.
    2. Usual SAS programs in which the reader can simply manipulate the dataset and variable names.
  3. Extensive exercise and exam questions: There is a large bank of extensive and mixed-format questions for most chapters, including multiple choice, theory questions, longer and shorter questions based on SAS outputs, and longer or shorter questions based on datasets and instructions for SAS analysis.
  4. Additional online chapters: There are several extra chapters that are not printed with the book but that are available online with the book materials, including advanced topics on regression (hierarchical regression, mediation, moderation, and non-linear regression).
..................Content has been hidden....................

You can't read the all page of ebook, please click here login for view all page.
Reset
3.16.212.217