This Article Full Text (PDF) All Versions of this Article: 1094428108321153v1 12/4/687    most recent References Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Alert me to new issues of the journal Add to Saved Citations Download to citation manager Request Permissions Request Reprints Add to My Marked Citations Citing Articles Citing Articles via Scopus Google Scholar Articles by Bedeian, A. G. Articles by Streiner, D. L. Social Bookmarking                         What's this?

Decimal Dust, Significant Digits, and the Search for Stars

Louisiana State University, abede{at}lsu.edu

Michael C. Sturman

Cornell University

David L. Streiner

University of Toronto

The practice of rounding statistical results to two decimal places is one of a large number of heuristics followed in the social sciences. In evaluating this heuristic, the authors conducted simulations to investigate the precision of simple correlations. They considered a true correlation of .15 and ran simulations in which the sample sizes were 60, 100, 200, 500, 1,000, 10,000, and 100,000. They then looked at the digits in the correlations’ first, second, and third decimal places to determine their reproducibility. They conclude that when n < 500, the habit of reporting a result to two decimal places seems unwarranted, and it never makes sense to report the third digit after the decimal place unless one has a sample size larger than 100,000. Similar results were found with rhos of .30, .50, and .70. The results offer an important qualification to what is otherwise a misleading practice.

Key Words: estimating and rounding numbers • significant digits • statistical precision • confidence intervals

This version was published on October 1, 2009

Organizational Research Methods, Vol. 12, No. 4, 687-694 (2009)
DOI: 10.1177/1094428108321153


CiteULike    Complore    Connotea    Del.icio.us    Digg    Reddit    Technorati    Twitter    What's this?