A great many empirical researchers in the social sciences take computational factors for granted: For the social scientist, software is a tool, not an end in itself. Although there is an extensive literature on statistical computing in statistics, applied mathematics, and embedded within various natural science fields, there is currently no such guide tailored directly to the needs of the social sciences. There is also an abundance of package-specific literature, and a small amount of work at the basic, introductory level. What is lacking is a text that gives social scientists modern tools, tricks, and advice, yet remains accessible through explanation and example.
The overall purpose of this work is to address what we see as a serious deficiency in statistical work in the social and behavioral sciences, broadly defined. Quantitative researchers in these fields rely upon statistical and mathematical computation as much as any of their colleagues in the natural sciences, yet there is less appreciation for the problems and issues in numerical computation. This book seeks to rectify this discrepancy by providing a rich set of monographs on important aspects of social science statistical computing that will guide empirical social scientists past the traps and mines of modern statistical computing.
"It has the potential to be enormously influential across the social sciences, but at present only fairly sophisticated quantitative modelers will be able to appreciate it. It should be required reading for everyone who performs statistical computing at an advanced level. Chapters 1 and 2 should be required reading for anyone who intends to publish the results of a statistical analysis using a commercial software package."
Journal of the American Statistical Association 2005. Vol. 100(470) pp pp. 707-8 (subscription required)
"...illuminating and necessary reading assignment for all students of applied statistics."
"This book provides a detailed rendering of numerical issues that any applied statistical researcher should be aware of when analyzing and interpreting results from computer-derived statistical analysis of a dataset. Although the book's title may narrow its appeal primarily to social scientists, large sections of it would be beneficial to any data analyst. It is timely and interesting, and on the whole provides a good balance of theory, applications, and computation."
Technometrics 2005. Vol. 47(2) pp pp. 241-2 (subscription required)
"This comprehensive research and guidebook by Altman, Gill, and McDonald offers to social scientists modern tools and tricks previously lacking in other works."..."Highly recommended."
Choice, June 2004, Vol. 41 No. 10)
Usually, the authors of these kinds of books pay insufficient heed to the simple principle that the reader needs to understand the overall purpose of the work being discussed, in order to best assimilate the details. This book avoids that problem; starting with the introduction, it emphasizes the importance of correct statistical work in the social and behavioral sciences. ...This is not an easy book, but, despite the complexity of the approach, it is a very important one for researchers, social scientists, and, of course, graduate and post-graduate students in various disciplines who understand that they must have strong skills in statistics and mathematics. Students in particular will benefit from the development of the formulas on almost every page in the second section of the book....This will be an important book for anyone interested in computational statistics; such readers must understand that there is always a problem with any result of a statistical computer program, and most of these problems are not known to the program's users.
ACM Computing Reviews Review #: CR129850 (subscription required) , July 26, 2004.
They offer readers a practical guidebook to the numerical methods underlying computerized statistical calculations specific to these fields. The text introduces readers to the basic principles of numerical computation, outlines the optimization process, and provides specific tools to assess the sensitivity of the subsequent results to problems with these data or model.
SciTech Book News Mar2004, Vol. 28 Issue 1, p14
A compact guide to the voluminous literature on optimisation, numerical analysis, and computational statistics. This is no small achievement.
Statistical Software Newsletter in Computational Statistics and Data Analysis, forthcoming
This is a very interesting book in an area that hasn't gotten much attention in the social sciences, but can expect to have more than just a niche audience ... Numerical Issues in Statistical Computing for the Social Scientist is the right book for any social scientist who has stumbled upon error messages ... and who is not just interested in some rough guidance on what to watch out for, but rather wants to understand the source of these problems ...
Journal of Statistical Software 2005. Vol 12, Book Review 5
Altman, Gill and McDonald have succeeded in providing two useful pieces of information to researchers involved in statistical analysis and interpretation of data: 1) a good understanding of the potential pitfalls involved with the implementation of methodology computationally, and 2) good advice on dealing with the problems that can arise.
Statistical Methods in Medical Research 14: 319-324, 2005. (subscription required)
Clarity of the presentation is excellent... This book is useful for a graduate level course on statistical computing. Applied statisticians and computer scientists will like this book and find it very useful.
Journal of Statistical Computation and Simulation November2005, 75(11):935
| Copyright © 1995-2009 | Micah Altman |