Review of The Workflow of Data Analysis Using Stata精选.pdfVIP

The Stata Journal (2009) 9, Number 1, pp. 158–160 Review of The Workflow of Data Analysis Using Stata, by J. Scott Long Alan C. Acock Department of Human Development and Family Sciences Oregon State University Corvallis, OR alan.acock@ Abstract. This article reviews The Workflow of Data Analysis Using Stata, by J. Scott Long. Keywords: gn0044, data analysis, workflow, data management, wdaus 1 Introduction The Workflow of Data Analysis Using Stata (Long 2008) is a must read for every Stata user. The book defies a simple description. It is not a substitute for Stata’s own reference manuals on data management or programming, but most readers will keep Workflow handy as a reference book for when they cannot remember how to do something. It is a great supplemental book for statistics courses precisely because such courses rarely spend much time on the topics this book covers—topics that students need to know to be effective Stata users. Workflow is a reflection of what a senior scholar has learned about what happens as we move from raw data to a published report. Workflow presents many guidelines and few readers will agree with all of them. Many of the guidelines may seem obvious to experienced users, but this is only because they were learned the hard way. For example, on page 143 Long writes, “Never change a variable unless you give it a new name.” Those of us who have, on occasion, ignored this advice learned its importance when a pivotal variable was irreparably destroyed. I fear that some beginning Stata users who read such guidelines are likely to ignore them. Long minimizes this risk by showing the likely consequences of ignoring