Saturday, November 15, 2014

Definition of rape catching up to the times

The FBI is changing the way it defines rape for the first time in decades (literally - the Uniform Crime Reports began in 1929 and the rape definition has not been changed since). Under any definition, rape statistics are notoriously unreliable, since we know there is significant under-reporting in the first place. But the new definition is a move in the right direction as it expands the actions that constitute rape (to include penetration with an object) and who can be raped (under the previous definition, only women could be raped). The one downside to changing the definition is that it will make comparisons across time impossible.

Wednesday, February 26, 2014

Modules for teaching with data

If you teach any course that involves data, you may find this new resource useful. The description from the site:, through a partnership with the Science Education Resource Center (SERC) has developed a library of pedagogic modules for educators across the curriculum. Each module features a particular pedagogic methodology including examples of how the method can be applied in a variety of subjects. SERC vets these modules with pedagogic experts; all pedagogic content is subject to a blind peer review process before it is made live.
A growing collection of classroom activities, submitted by faculty, is included within each pedagogic module. The result is an enhanced collection that allows users to seamlessly browse between pedagogic content and classroom activities. The modules can be used in their entirety or instructors can use the modules to generate ideas for their instruction.

The modules include: Teaching with Data, Developing Quantitative Reasoning, Quantitative Writing, Teaching Quantitative Reasoning with the News, and Using Socio-Scientific Issues-Based Instruction.

Sunday, February 16, 2014

Are there really so few strikes?

Have labor strikes "nearly vanished from the American economic landscape?" That's the question asked by sociologist Jake Rosenfeld in his new book What Unions No Longer Do. As we note in The Date Game (p. 202) the question is difficult to answer because of Reagan-era budget cutbacks to the U.S. Bureau of Labor Statistics. Since 1981 data on strikes are available only for labor stoppages involving more than 1,000 workers. Previously the Bureau spent funds gathering data on smaller strikes involving at least six workers for one eight-hour shift.

Rosenfeld notes the measured decline in large work stoppages from more than 200 per year in the 1950s, 1960s and 1970s to fewer than 50 during the 1990s and 2000s (and only nine in 2009).  But perhaps this apparent decline is a result of the fewer number of large workplaces, especially manufacturing plants in the Midwest. Historically most strikes took place in smaller workplaces, so it could be the case that measuring only large workplaces exaggerates the decline in strikes.

Rosenfeld filed a Freedom of Information Act request to obtain data from the Federal Mediation and Conciliation Service where most unions must file before any stoppage. These data show that indeed all work stoppages were several times the number involving more than 1,000 workers, peaking at 1,000 stoppages in the early 1980s. However, since then the data for all work stoppages show a steep falloff until 2002, the most recent date investigated by Rosenfeld. As a result he concludes the "declines in work stoppages are not due to the particularities of large strikes."

See chapter 4, "Strikes" in What Unions No Longer Do Harvard University Press, 2014.