Rumors of Our Rarity are Greatly Exaggerated: Bad Statistics About Women in Science
The essay originally in this location was updated and published in the Journal of Humanistic Mathematics in July of 2011. It can be downloaded free of charge here. The abstract is below. Below that are sources of statistics about women in STEM. Last are updates about the garbled statistics discussed in the Journal of Humanistic Mathematics article.
Abstract / Synopsis
During the past few years, three bad statistics have been persistently used in discussions of women in science, technology, engineering, and mathematics (STEM). The first was questionable when it was published in 1983 and has since been widely used. The second came to prominence in 2006 – and now leads an international and perhaps eternal life on the Web. The third may have made its debut in 2007.* Its variants occur in popular and academic books and journals, including the 2011 Proceedings of the National Academy of Sciences.
This report presents case histories of the three bad statistics, suggests writing and editing practices which might reduce such occurrences, and provides primary sources of statistics on women in STEM.
Primary sources of statistics on women in STEM
American Institute of Physics (physics faculty demographics, enrollment from high school to graduate school, undergraduate and graduate degrees granted)
American Mathematical Society (mathematics faculty demographics, PhDs granted)
Computing Research Association (computer science faculty demographics, undergraduate and graduate degrees granted)
Conference Board of the Mathematical Sciences (mathematics faculty demographics in 2- and 4-year colleges; undergraduate mathematics enrollment and degrees granted. Preliminary findings from the 2010 survey show an increase in proportions of women. They are over 35% of tenure-eligible mathematics faculty at BA- and MA-granting institutions and over 25% of those at PhD-granting institutions.)
National Science Foundation (several different reports and data sources (e.g., Webcaspar) for STEM statistics. For example, statistics about high school course taking in mathematics and science and engineering are reported by gender and ethnicity in Science and Engineering Indicators 2008. In 2005, similar proportions of girls and boys took mathematics courses. Science and Engineering Indicators 2012 is here.)
2008 and 2009 discussion of trends in CBMS and NCES statistics for undergraduate mathematics majors note a decrease in female mathematics majors. At the same time, there has been an increase in female physical sciences majors, according to NCES data.
* Since the article was published, additional details about the history of the “garbled statistics” have come to my attention. It appears to have made its debut in 2005 rather than 2007.
The percentages given in Halpern et al. 2007 are given in a sidebar in the Washington Post January 30, 2005. These are correctly labeled “Percentage of women holding tenured and tenure-track faculty positions at the “top 50” university departments” (emphasis added) and their source is given.
Several months later, the same percentages were given in slide 2 of the Pinker–Spelke Edge debate on May 16, 2005. These are labeled “Underrepresentation of women among tenure-track faculty at elite universities” (emphasis added) and no source is given.