Quote:
Originally Posted by dannat
i went to lecture at latrobe the other day by a prof who has collected data both here ans o/seas and reffered to another study that says of all scietific research whether published in nature/science/whatever abut 50% of papers have errors in them, wheter it be form misconduct..or stat alalysis (most common)..it seems most scientists love to use facts figures but only use them to show there own point of view
|
I'd suggest that 100% of papers have an error in them, it's just that some are more significant than others. However if there is a study that suggests that scientific fraud is commonplace then I'd suggest there is a significant error in that paper!
Of course as astronomers we should be aware of a major bit of scientific fraud and misconduct - the famous work by Hubble. Firstly he used the data of Slipher without any acknowledgement (= misconduct). Secondly, apparently the data set used in the first paper was not large enough to draw the statistical conclusions he drew. To be sure he was correct and he soon afterwards had enough data to prove that, but he did publish (and was accepted) too soon.