October 30, 2015 report
Study shows economic studies done in lab offer similar results to those done in the field
(Phys.org)—A pair of economics researchers has found evidence that suggests that economic studies done in the lab offer results that are comparable to those done in the field. In their paper published in the journal Science, Daniel Herbst and Alexandre Mas with Princeton University, describe how they compared study results that used one or the other of the two methods and why they believe what they found likely translates to other studies as well. Gary Charness and Ernst Fehr with the Universities of California and Zurich respectively, offer a Perspective piece on the work done by the research pair in the same journal issue and offer some ideas on why the results from such studies were found to be similar.
There has been some debate in the economics field regarding the validity of economic experiments done in the lab due to the constrained environment in which they are conducted. Until now however, as the research pair point out, very few if any studies have been done to find out if such concerns are warranted—to find out, they designed and conducted a study on earlier studies.
The study consisted of analyzing the results of studies done on what is known as the spillover effect—where productivity in a group goes up if just one member of a group increases his or her productivity. The research pair found data from 34 such studies, 11 done in the lab and 23 in the field over the past fifteen years, and compared the results. They found only very small differences, suggesting that the work done in the lab was every bit as valid as the work done in the field.
Charness and Fehr suggest that there is a reasonable explanation for the results—it is because those conducting work in the lab add aspects to their studies to counter the effects of the lab environment. Incentives can be added, they note, or volunteers can be used that are not university students, or by offsetting the impact of behavior changes that result from being observed by offering comparisons to workers that do their work in the real world under observation. For this reason, they suggest the results found by Herbst and Mas likely apply under other scenarios as well, which might be enough to cause some scientists to change their views on lab studies.
More information: D. Herbst et al. Peer effects on worker output in the laboratory generalize to the field, Science (2015). DOI: 10.1126/science.aac9555
Journal information: Science
© 2015 Phys.org