In 1980, Michael Lewis-Beck published Applied Regression: An Introduction, a remarkably succinct, informative, and above all accessible introduction to regression analysis. I used it selectively as ancillary reasding in courses I taught in basic statistics, quantitative research methods, and multiple regression. Students loved it. Finally, a statistics text they could use to help them with material they didn't understand in class or that was not clearly explained in their other readings.
Unbeknownst to me, the same author later published another Sage paperback titled Data Analysis: An Introduction. I became aware of this 1995 text only recently, and, after reading it, I realized that the last fifteen of the twenty-three years I taught basic statisitcs before retiring in 2010 would have been much easier for me and much more useful for beginning students if I had learned of the book when it first appeared.
As with Lewis-Beck's Applied Regression text, Data Analysis is a tribute to the author's ability to write both succinctly and accessibly about counter-intuitive material that students commonly dread. No reason for dread here, however, because once again Lewis-Beck has written a text that is really useful for beginners who want to use it as supplementary reading for a course or -- wonder of wonders! -- a self-teaching tool. I had long been convinced that the latter objective, producing a reasonably thorough introduction to basic statistics that would serve the interests of those who wanted to teach themselves but who lacked a good deal of mathematical aptitude, was something that would never be written. Lewis-Beck, however, proved me wrong, and he did it in less than eighty pages.
If his text has limitations they are manifest in its peculiarly sociologically orientation, meaning that an obligatory introduction to elementary probability theory is not included, and one-way analysis of variance, a tool that is used and misused much too often in some behavioral science disciplines, is not introduced.
However, there is much more thorough coverage of regression analysis than is included in all but a few introductory texts. Also, simpler techniques, such as the ubiquitous t test and non-parametric measures of association, including those based on Chi-square, are given their due. Lewis-Beck, however, realizes statistics starts to get interesting when analysts begin applying multiple regression. By the time the reader has worked through the first forty pages of Data Analysis, he or she is ready to make a first foray into material that most instructors save for a second course. If Lewis-Beck's earlier book, Applied Regression, were read as a sequel to Data Analysis, a student could honestly and confidently claim to be pretty well on the way to being a reasonably capable social science statistician.
Textbook publishing is a notorious racket. But Lewis-Beck's inexpensive and masterfully written Data Analysis: An Introduction provides a welcome reprieve from the usual rip-off routine.