There are many reasons why a research paper could be retracted, although the one that tends to dominate public discussion is research misconduct.

Similarly, there are many different definitions of research misconduct, but the one that tends to draw attention is deliberate deception and data fraud. That can help to explain why, when Nature ran a news story in January 2013 about a new course that would attempt to rehabilitate misconduct offenders, many of the online comments below the story were negative (see Nature 493, 147; 2013).

One response was typical: “If a scientist wilfully pollutes the scientific record with damaging, self-serving, purposeful lies, it is too late. If they have received funding for that fraud, I am still mystified as to why the granting agency does not seek to recoup their costs in court. Now, someone suggests we further train them on something that is obvious to any scientist with half a brain on their shoulders?”

More than three years on, the architects of the rehab course offer a progress report, which appears as a Comment article this week. More than three dozen researchers have been through its doors, and, according to the authors, most leave as better scientists than when they arrived.

Attendees do not include the high-profile data fraudsters whose offences are so serious that they are fired. By definition, researchers on the course are scientists who have been caught out but whom institutions want to keep.

Most of them had seen their research privileges suspended — with offences ranging from plagiarism and poor oversight to falling foul of the rules and regulations on animal welfare and informed consent. Despite the ‘research misconduct’ label, instances of conscious wrong-doing were rare. As one participant said: “Prior to this situation, I tried to follow the spirit of the law. Now I try to follow the letter of the law.”

Two points stand out. First, the typical character and personality of these scientists, and their knowledge and attitudes, were no different from those of you and your colleagues. Misconduct, the authors say, can be down to circumstance: “we believe that most researchers may be susceptible”. And second, those circumstances are becoming more common.

The most common cause of an offence was a lack of attention, prompted, among other things, by being too busy and trying to juggle too many projects. Sound like anyone you know?