Bad science papers can have lasting effects. Consider the 1998 paper in the journal The Lancet that linked autism to the MMR vaccine for measles, mumps and rubella. That paper was fully retracted in 2010 upon evidence that senior author Andrew Wakefield had manipulated data and breached several proper ethical codes of conduct.
Nevertheless the erroneous paper continues to undermine public confidence in vaccines. After the Lancet article, MMR vaccination rates dipped sharply and haven't fully rebounded. This decline in the MMR vaccine has been tied to a rise in measles cases resulting in permanent injury and death.
Each year hundreds of peer-reviewed scientific articles are retracted. Most involve no blatant malfeasance; the authors themselves often detect errors and retract the paper. Some retractions, however, as documented on the blog Retraction Watch, entail plagiarism, false authorship or cooked data.
No journal is safe from retractions, from the mighty "single-word-title" journals such as Nature, Science and Cell, to the myriad minor, esoteric ones.
Yet as astronomer Carl Sagan once said, "Extraordinary claims require extraordinary evidence." Below are five science results retracted in 2011, pulled permanently off the books in part for falling far short of meeting the Sagan standard.
#5: Los Angeles marijuana dispensaries lead to drop in crime.
Keep smoking. The RAND Corporation retracted its own report in October after realizing its sloppy data collection.
Crime data compiled from neighborhoods with these highly contentious medical marijuana dispensaries supposedly revealed slightly lower crime rates. The authors attributed this decline not to marijuana itself but rather the presence of security cameras and guards in and around the dispensaries, having a positive effect on the neighborhood. [The History of 8 Hallucinogens]
The L.A. city attorney's office was incensed with the report, having argued the opposite — that the dispensaries breed crime. The city's lawyers soon found critical flaws in RAND's data collection, largely stemming from RAND's reliance on data from CrimeReports.com, which did not include data from the L.A. Police Department. RAND blamed itself for the error, not CrimeReports.com, which had made no claims of having a complete set of data, and, in fact, didn't even know about the study.
#4 -- Butterfly meets worm, falls in love, and has caterpillars.
The Proceedings of the National Academy of Sciences (PNAS) published a fantastic claim in 2009 by zoologist Donald Williamson, which was delightfully reported in the science news media. Williamson claimed that ancestors of modern butterflies mistakenly fertilized their eggs with sperm from velvet worms. The result was the necessity for the caterpillar stage of the butterfly life cycle.
The PNAS paper got a few laughs among evolutionary scientists, but it hasn't yet been retracted. Williamson's follow-up 2011 paper in the journal Symbiosis, however, has been retracted.
Researchers Michael Hart and Richard Grosberg at the University of Texas, Austin, systematically refuted all of Williamson's claims in the pages of PNAS by the end of 2009. They based their arguments entirely on well-known concepts of both basic evolution and the genetics of modern worms and butterflies. When Symbiosis published its butterfly-meets-worm article in January 2011, Hart raised questions with the editor. As of November the paper is no longer available.