Merriam Webster reports that in 2005, “integrity” received more hits than any other word in their online dictionary. It’s not clear how many more hits scientific integrity can take: An MIT researcher is fired for fabricating a dozen papers. A pharmaceutical company omits data from key publications about side effects. A South Korean stem cell researcher admits to a stunned nation that, “blinded by work and a drive for achievement,” he submitted a “fake it before you make it” article to Science. It appears that research misconduct has taken its place among the epidemics that scientists need to worry about.

An aphorism attributed to Mark Twain holds that there are three kinds of lies: lies, damn lies, and statistics. At first the public points to a bad apple who paints mice or switches out slides, and fumes if the researcher conspires to hide it. But it takes a village to do big science: authors, collaborators, students, sponsors, regulators; different languages, different countries, disparate goals. A lone scientist can offer mea culpa, but fraud on the scale of South Korea’s almost always involves collusion and conspiracy, hidden in the complexity of the research. It is a nightmare for scientific journals, but more than anything it terrifies the public.

Science depends on public support. Too often, Hollywood sends the message to a fickle public that scientists-cum-fundraisers cannot be trusted to make what Malcolm Gladwell calls “blink” judgments about recruitment of egg donors, or to review thousands of pages of data. In the wake of the Hwang scandal in Korea, stem cell researchers who had been worshipped as heroes scraping for support become “rogues blinded by ambition.” Those waging jihad against stem cell research talk about tangled webs, and they ask how far it is from an exaggeration of the results to an exaggeration of the benefits of embryonic cell research.

The solution to what the public – incorrectly – perceives as an epidemic of scientific misconduct is not obvious. Public relations is not the answer. Scientists who have spoken publicly about the Hwang matter have only made things worse. There are dozens of commissions on research ethics and programs to provide certification in it. There are conferences, journals, and agencies. Companies post ethics codes in hallways. Government pressure for compliance depends on funding, and in the United States, there’s scant funding for stem cell research.

The great hope lies in teaching new generations of scientists, yet there’s no evidence that it has an effect on the rate of misconduct. Scare tactics about the fate of Woo-suk Hwang will not transform those who enter graduate school ready to fabricate results, particularly when students report that their mentors could care less about the ethics course.

Nothing will prevent Dr. Jekyll from becoming Mr. Hyde, but mentors and oversight can help vulnerable newbies to science eschew bad habits. Researchers learn what is important by watching the boss; perhaps the boss should learn to teach the integrity course using lessons from the lab. Funding and institutional review-board approval should depend less on consent forms and more on ethics training and strategy. University compliance should emphasize remediation for those who play with matches rather than punishment for burning down the house.

Ultimately we don’t have any clue what works, but it’s a safe guess that the institutions that innovate in research ethics and study outcomes will be the ones that prevent misconduct. If we’re going to throw buckets of money at frontier science, we’d better throw a little bit more at finding the best ways to help new scientists do it responsibly.

Glenn McGee is the director of the Alden March Bioethics Institute at Albany Medical College, where he holds the John A. Balint Endowed Chair in Medical Ethics.