About | HeinOnline Law Journal Library | HeinOnline Law Journal Library | HeinOnline
LOG IN
LOG IN

84 Soc. F. 1 (2005-2006)
Alphas and Asterisks: The Development of Statistical Significance Testing Standards in Sociology

handle is hein.journals/josf84 and id is 17 raw text is: Alphas and Asterisks: The Development of Statistical Significance Testing Standards in Sociology Erin Leahey, University ofArizona What is distinctive about sociology's history is the extent to which it both illustrates and challenges sociology itself. - N. J. Demerath, III (1994) Abstract In this paper, I trace the development of statistical significance testing standards in sociology by analyzing data from articles published in two prestigious sociology journals between 1935 and 2000. I focus on the role of two key elements in the diffusion literature, contagion and rationality, as well as the role of institutional factors. Ifind that statistical significance testing flourished in the 20th century. Contagion processes and the suitability of significance testing given a study's data characteristics encourage the diffusion of significance testing, whereas institutional factors such as department prestige and particular editorships help explain growing popularity of the .05 alpha level and use of the three-star system of symbolic codes (i.e., *p < = .05, **p < = .01, ***p <= .001). What alpha level do you use when testing for statistical significance? Many sociologists today would say .05. What symbol do you use to indicate this level of statistical significance? Many sociologists would say one asterisk. Although the choice of alpha level should technically depend on sample size, statistical power and sampling procedures, researchers routinely use the .05 percent level (signified by a single asterisk) as a benchmark despite variation in these conditions. In fact, one of the leading journals in the discipline has a policy that disallows the reporting of significance above the .05 percent level. How did statistical significance testing become normative? How did the .05 level and its symbolic code (a single asterisk) become dominant? What factors influenced the creation and propagation of these practices? Although some scholars point to the technical superiority of significance testing as the key factor affecting its diffusion, two factors undermine this argument. First, the technical superiority of statistical significance testing relative to other methods for testing hypotheses was, and still is, hotly debated (Carver 1978; Cowger 1984; Labovitz 1972; Morrison and Henkel 1970; Raftery 1995; Schmidt 1996). Second, statistical significance testing did not become dominant in all disciplines. Researchers with goals similar to those of sociologists rely on slightly different practices; for example, public health researchers typically employ (albeit 95 percent) confidence intervals, and psychologists rely more on effect size. If statistical significance testing was technically the most appropriate procedure and technical I would like to thank Barbara Entwisle for helping me formulate and improve the design and ideas expressed in this paper. Comments from Ken Bollen, Gail Henderson, Ted Mouw, the late Rachel Rosenfeld, Francois Nielsen, Sarah Soule, Ron Breiger, John Levi Martin, John Sonnett and Adrian Raftery were also instrumental in improving the manuscript. Correspondence should be directed to Erin Leahey, Department of Sociology, University of Arizona, P.O. Box 210027, Tucson, Arizona, 85721-0027. E-mail: leahev@arizona.edu. Social Forces, Volume 84. Number 1. September 2005 © The University of North Carolina Press

What Is HeinOnline?

HeinOnline is a subscription-based resource containing thousands of academic and legal journals from inception; complete coverage of government documents such as U.S. Statutes at Large, U.S. Code, Federal Register, Code of Federal Regulations, U.S. Reports, and much more. Documents are image-based, fully searchable PDFs with the authority of print combined with the accessibility of a user-friendly and powerful database. For more information, request a quote or trial for your organization below.



Short-term subscription options include 24 hours, 48 hours, or 1 week to HeinOnline.

Purchase Access

Contact us for annual subscription options:

Contact Us

Already a HeinOnline Subscriber?

Log In

profiles profiles most