Question: So you rely on data from other scientists? How can you be sure about the data of other scientists, to be not faked?

Asked by finnpetersen to Anna, Daniel, Denise, Jesse, Rehemat on 19 Jun 2014.

0

  1. Photo: Rehemat BhatiaRehemat Bhatia
    answered on 19 Jun 2014:

    Hi finnpetersen,

    Good question, and one that researchers have to think about all the time too! When dealing with published data you have to be cautious and question everything.

    If data is published in a scientific journal its peer reviewed. This means when a paper is submitted by researchers, it is then passed to a panel of scientific editors, who then scrutinise it for common mistakes or things that people could question. Data must also be reproducible; this means thats the experiments the scientists do have to be able to be done again. And they should outline all methods that were done in their paper. All data, where possible, should have error bars too. This is crucial in determing data accuracy. An error bar basically shows the highest and lowest value that a data point could be. The smaller the error bar the more accurate the data.

    If a scientist fakes their data, and is found out, they would basically ruin their reputation. I don’t think many people would want to do that!

    If you do make a mistake in your journal article and its slipped through the peer review system, you are allowed to write an erratum saying why it was wrong and provide a correct version.

    0

  2. Photo: Jesse DavenportJesse Davenport
    answered on 19 Jun 2014:

    Hello finnpetersen! That is an excellent question. I think Rehemat has answered it nicely. A big controversy that happened recently was the creation of a number of fake papers by a computer program that basically generated nonesense and was subsequently published (http://retractionwatch.com/2014/02/24/springer-ieee-withdrawing-more-than-120-nonsense-papers/). A scientist in France caught this and the papers were retracted, but it illustrates very well the fact that sometimes things slip through the cracks and publishers, journals and scientists can’t always catch everything. However, it is up to scientists to be truthful and ethical in their research and publications and to respect the scientific process.

    0

  3. Photo: Anna RabittiAnna Rabitti
    answered on 19 Jun 2014:

    Hi finnpetersen,
    Rehemat and Jesse gave a nice overview of what is the reality at the moment. I will tell you how I personally proceed, that is..write to that person directly! If you want to use data collected from someone else, I think you should always know how they actually collected the data, if they “clean” them, the methodology, and so on.
    Now we are very lucky, because through websites and journals you can always have the email address or the phone number of the person that is responsible for those data you want to use. So you write to him/her, or phone him/her, ask 1. if you can use the data (not everybody will give their data around very easily.. they cost a lot of money and effort) and then ask how they were collected, processed etc.
    In this way I think if there are “weird”, or fake data, you will then find out for yourself!

    0

  4. Photo: Daniel Lao DavilaDaniel Lao Davila
    answered on 19 Jun 2014:

    The best discoveries occur when scientists question the assumptions of other scientists. So you are right, we need to be sure before using data from other scientists. One way to be sure, is by the process of peer review when submitting papers for publications. Scientists check to see whether the methods used are reasonable. If there is doubt after publication, the experiment could be tried by other scientists to see if they have the same results. Now, some studies have been studied a lot by many scientists so other scientists would assume that they are correct and build upon the results of other scientists. That is why scientific reviews and corroboration of results are very important.

    0

  5. Photo: Denise De GaetanoDenise De Gaetano
    answered on 22 Jun 2014:

    Dear Finnpetersen,

    As the other scientists have mentioned, most published scientific data is peer reviewed. Most theories and data is used so much over and over again, that if anything is false, then this would be found by some scientist at some point.

    In fact, most data which is obtained from other scientists, is systematically analysed before being implemented into any new research.

    Apart from being easily found, it would not be fund doing science, if data is falsified. How would we know how the earth really works if we could not use real data?

    Hope this helps! 🙂

    0

Comments