• Question: How much do you need to check your work to make sure it is right?

    Asked by ciara to Meeks, Pete, Stephen, Steve, Tom on 22 Jun 2010 in Categories: .
    • Photo: Tom Hartley

      Tom Hartley answered on 22 Jun 2010:

      Good question. Checking and double checking is one of the most important things in my work, and I think this holds for all sciences. When you get an exciting result its very tempting to believe it and rush ahead with the next thing. Everyone knows that you should check if you think you might have made a mistake, the trick is to keep checking even when you think you’ve done everything right. Over the years I am gradually learning to check, check and check again, and now I am teaching my own students the same thing. The best way to learn about checking is to write computer programs (as I did for my PhD). They are very unforgiving and go wrong in all sorts of ways if they aren’t very carefully checked.

    • Photo: Marieke Navin

      Marieke Navin answered on 22 Jun 2010:

      Hey Ciara – I actually check my work loads as i’m always scared it’ll be wrong!! this means repeating experiments and analysis and everything taking twice as long!

    • Photo: Steve Roser

      Steve Roser answered on 22 Jun 2010:

      The aim of most scientists is to put their work out into the public domain – generally in papers in scientific journals. You have to be really careful if you want to put your results out there – there are lots of people who might check them, and trash you if you are wrong

    • Photo: Stephen Curry

      Stephen Curry answered on 22 Jun 2010:

      Hi Ciara

      That’s a very insightful question. The answer is: a lot!

      When we do an experiment, we need to plan carefully and make sure that we do a fair test – so that we are not deluding ourselves about what the experiment means.

      Even when you take good care, it is sometimes possible to convince yourself that your experiment is right, even when it is wrong.

      So when we write up the results to be published in a scientific journal, the paper is read by 2 or 3 other scientists who check it quite carefully. They may spot mistakes – or gaps – and ask for changes.

      Once the changes are made, the paper is then published. But even then the checking is not finished because everyone in the world (practically) can read the paper and make up their minds about it. Sometimes, the results will be tested by other scientists who may find errors.

      When they publish their paper, you may find that your results have been contradicted.

      By all these checks, science slowly but surely gets closer to the truth of what is really happening in the world!