I have (almost) committed a scientific fraud
Here is why scientists fake their results, and how to stop it.
Twenty percent of scientists confessed it: at least once in their life, they cheated in their scientific paper.
I have read this headline many times. Sometimes the percentage is 30%, some other time even 40%, it just depends on the methodology of the survey and the honesty of the scientists who replied to it. And every single time I thought “losers! I have never cheated! Every single bit of data in my papers is honest”.
Sure, I am not the best scientist in the world, but I still managed to make a career in science without faking my results. Why should one cheat? Sooner or later someone will find out that your findings are fake and you will lose everything!
While thinking that I was internally pumping up the ego of my moral integrity. I was judging them as thieves, dishonest, and unworthy of calling themselves scientists.
Until the day I realized they were just weak.
Everything started one day, when I was in my laboratory. The experiment I was working on for almost two years still did not work. I was desperate and I felt a voice inside telling me: come on, you can cheat a bit. Just modify this and that… nobody will ever know.
I was this close to cheating myself. I had never thought I could even do it, but at that very moment, I stopped judging other scientists who cheat.
Let me be clear: cheating on a scientific paper is wrong.
The scientific community relies on your results and they trust you. Publishing fake findings is disrespectful to the whole scientific community. If another scientist will design an experiment based on your false results he will probably fail, and that’s your fault.
I won’t ever condone those who cheat on their papers, and this must be clear. However, I have realized the mechanism leading a scientist to cheat is not just a lack of moral integrity. Only if we find the actual reason why scientists cheat we can fight this phenomenon. Spoiler alert: climbing a pedestal and judging others doesn’t help. At all.
There are few problems with scientific papers.
Cheating is extremely simple.
Firs of all, cheating is extremely simple. When a major scientific fraud is found I often read about investigators requesting the scientist accused of fraud to show “the original record” of the measured data. I typically laugh.
Nowadays all data is recorded digitally. I perform some measurements, the data arrives at the computer and nothing is preventing me to modify the data. It is as simple as opening a .txt file and change a 3 to a 5. How are you gonna find out that such text file does not contain the original measured data but my fake data? This applies, in fact, also to paper records. If you give me a free afternoon, a notebook, some pens, and a large amount of coffee I can write down the results on paper tons of results of imaginary experiments. How can you prove those experiments never happened? I can even write a fake date on the top of each page: does this make my “original record” any more reliable?
The sad reality is that producing fake data is very easy.
Sure, a few times they have busted scientists publishing false results, but that was only because they were naive enough to add some fake noise to their invented data… and they used the very same fake noise in all the figures! Unless you make such a mistake it won’t be so easy to prove you cheated. In the worst-case scenario, you can claim your experiment worked in the past but now it does not work anymore and you don’t know why (it does honestly happen sometimes).
Second: most of the time you get away with scientific fraud.
Which is also proved by the fact that 20% (or even more) of scientists admit they have cheated on their papers but don’t see 20% or more of scientists losing their job. If you publish a scientific breakthrough promising to revolutionize an entire sector then tons of scientists and companies will try to replicate your results and sooner or later they will find out you cheated. But if you promise something small most of the time only a few people will try to replicate your results. They will fail, in few cases, they will think it’s their own fault, in other cases, they will have a bad opinion about your scientific skills and avoid your talk and the next conference. Most likely, however, they will never rise a scandal and you get away with it.
As long as you don’t go too far with the fraud you will not suffer any consequence.
This leads scientists to an alluring temptation. Don’t claim any Nobel-prize level finding, just fake a little bit of your data. You will get some extra publications, some more citations, an earlier promotion to associate or full professor… and most of the time nobody will even realize you cheated.
The mistake everybody does
For many years I thought moral integrity was the key to fight scientific frauds. I have run several marathons in my life and cheating is a common problem even in amateur running. You cannot monitor 42.195 km of race, someone will cut it shorter to get under his personal best time and show it off to some friends. “What is the point to be proud of breaking your persona best time if you did it cheating? People won’t know it, but you do, you know you cheated… you won’t enjoy the achievement!”
I hear the very same type of argument also about scientific fraud: what is the point to get a paper published in a prestigious journal (and show it off to your colleagues) if the data you published is fake? Sure, your colleagues won’t know it, but you know, deep inside, you committed a fraud.
Unfortunately, I found out these arguments don’t work.
They don’t prevent you to cheat when you are in that state of mind leads you into the temptation to fake your data. I realized that when I was there, in those shoes.
My experience with cheating
As I mentioned earlier, I had been working on this idea since autumn 2019. It was of those moments when you think “hey, but what if I do this instead of that? We have always used 10% of this component and everybody claims it is the best but I am not sure about it. Maybe in some cases, 9% is better. Let’s try!”
It is a very typical situation in science: everybody keeps doing the same thing because they all think it is the best and they don’t even question it. Until a guy arrives, questions it, and finds out there is a better alternative.
That guy was me.
I enthusiastically started to dig into all scientific papers written when I was still at kindergarten, and after a couple of weeks of investigation, my response was clear: yes, all pieces matches, it should work. In theory.
What followed was a year and a half of a nightmare.
I changed laboratory, I moved to another country, a pandemic forced me and my associates to keep away from the lab. Making research became more difficult. Even ordering basic electronic components was a complicated when all shops were closed because of covid.
Also, the pandemic made me impossible to meet my loved ones. I have a family spread over thousands of kilometers in three countries and two continents. My significant other could not reach me for months and governments did not care at all about our situation. I did not know if our relationship could survive (don’t worry, it did).
My mental state was a mess.
In all of this, the state agency sponsoring my research project (and the salaries of the members of my team, mine included) wrote me just to say they were not happy about the publications I achieved.
I was struggling. In theory, my idea was working, but in the lab, everything was more complicated. Results did not arrive as I expected. I was nervous and scared of losing my project. The world around me was a mess and nobody appeared to care about my family issues with the pandemic. Right at the moment when I most needed support, the agency had the brilliant idea to scold me because I had not published enough papers yet.
Fuck off
When one more experiment failed I simply said that: fuck off. I will fake the data and that’s it. They don’t care about me, why should I care about the project?
Fortunately, I did not cheat. The following day I made one more experiment and suddenly everything started to work. I finally understood what was wrong in my experiments (a stupid, very stupid thing), and once I fixed it the flow of good data streamed to my computer. I almost wanted to cry. First of all, because the idea I had almost two years ago was right. Secondly, because for a moment I was this close to cheat for the first time in my scientific career and lose my integrity, but eventually I did not. My reputation is safe.
This experience, however, taught me an important lesson.
People often condemn those scientists who cheat in their results. I have done it too in the past, and I have been very hard on them. I have given a moral judgment about them.
While I still think cheating is wrong, now I am no so strict anymore.
After being in their shoes, after being this close to giving in to the temptation to cheat, I have learned people often don’t do it because they are evil, but rather because they are too weak.
What is the best way to avoid scientific frauds? How can we keep scientists away from the temptation to cheat?
A call to moral integrity is useless.
I was there, in the lab, struggling. The government was messing up my life and they did not care about my family situation despite multiple calls to let international partners reunite. Half of the faculty staff was at home enjoying for months their salary doing nothing. The job in the lab was just impossible to carry on because of crazy regulations which only made everything more complicated. I felt the entire world was not fair to me: why did I have to be fair to the world?
You can call someone to moral integrity only if you are fair to him in the first place.
If a scientist feels to be the victim of injustice he will think cheating is not that bad after all. It’s just a way to compensate for the justice he is suffering.
In a normal situation, he would feel bad about cheating because he knows it is wrong. However, when he can identify as a victim of injustice then he has the perfect excuse to pardon himself. He still knows, deep inside his soul, that he is wrong, but in the trial running inside him he can play this card of the extenuating circumstances. He feels there is somehow a balance between the bad he received and the bad he did.
In this case, there is no call for moral integrity which could work. That scientist has already acquitted himself.
The first rule for fighting scientific fraud is the following: make the world around scientists fair.
Sure, you cannot change the entire world, but at least the working environment should be as fair as possible. If you pay your scientists fairly, if you have fair rules which reward more those who more work and punish lazy colleagues, if you just even acknowledge their efforts and motivate them, then they won’t have any excuse to pardon themselves in case of cheating. They won’t be able to say “ok, I did something wrong in manipulating this data, but come on… also this and that is wrong…”
Second: limit the pressure to achieve results.
One of the reasons I was about to fall into cheating in my paper was the pressure I had from the agency which sponsors my grant. They were demanding high quality publications soon. If they just pressed me a bit more and required publications just a bit earlier probably I would have cheated. Thank God I had some extra time and I could finally find out that my idea really worked, but not always scientists have this time.
The unbearable pressure to achieving soon results and publications can push scientists to take the shortcut of cheating just to make their boss, the rector, or the ministry happy.
Don’t take me wrong, I am an advocate of evaluating the quality of research. I am absolutely fine with asking scientists tangible results of their work, but sometimes the pressure is just too much. If you start a project today maybe the results will arrive in one year or maybe in three years. Investing money in a research project is a risk, it might fail.
It is way better for a project to fail than obtaining a scientific fraud because scientists cannot bear the pressure to achieve publications.
Third: don’t underestimate the care of scientists’ mental health.
Yes, sure, the scientific method is rigorous, you just need to follow it. Fine. Unless the reality is that we, scientists, are also humans and we also have the same weaknesses as all humans. This pandemic was a good stress test also for us. If your mind is not healthy you want to perform well as scientists and this includes also the temptation to cheat in your scientific results just to make your superiors happy. This is not different than the typical situation of an insecure teenager, who has strict parents who demand high results and start lying to them about his achievements just to make them happy and (temporarily) move the problem out of the view.
This type of situation is not solved by being even stricter parents but rather working on the insecurity of the child who doesn’t feel comfortable enough to talk about himself and his failures to the parents because he is afraid of his judgment.
The same applies to science.
Scientists could fall into a spiral of lies about his achievements simply because of his internal fears.
We must provide scientists with support for their mental health. Which does not only mean therapists, but also a friendly environment. The more a scientist is alone, the more he can fall into bad conditions about his mental health.
Prevent vs. suppress
These are just three points I suggest to prevent — rather than suppress — scientific frauds, and they all start from a different approach: before judging try to understand what has to lead the scientists to cheat.
Will there be scientists cheating simply because they are evil, they like to cheat and they aim for a promotion before other honest scientists. Sure, and I do not pardon them. However, there is a large majority of scientists who cheat for these other reasons. By helping them we can reduce the scientific frauds, and in turn help all of us.
