I’d argue that it doesn’t affect science as much as is initially apparent. At first glance the literature is overwhelmingly skewed towards the reporting of positive data but negative data are indeed published. Ultimately if this negative-result is interesting it will be pursued to find out why, and will likely lead to a result explaining the effect. But then this is a positive results paper borne from a negative result. As a result of this, negative data are reported like so: x causes y, however did not affect the related pathway z. Furthermore, there have been some pretty high profile papers that published almost exclusively negative data.
…we are a demographic so dependent on technology for generating data but are so archaic in our approach to communicating that data.
I sense some of you are still not convinced by these arguments so I must confess I think there is a problem here, but I think it is part of wider issue in scientific publishing. How do we communicate results and ideas that are traditionally not published? Three particular categories come to mind: 1) negative results (I know, I know, I said this wasn’t an issue); 2) confirmations of published results (or lack thereof); 3) those little experiments you did to pad out your thesis but seeing as you are leaving and no one else works of that project, probably won’t be followed up.
Here lies the issue: we are a demographic so dependent on technology for generating data but are so archaic in our approach to communicating that data. The nature of communication within the science community is for the most part pretty low throughput. So, how do we change this? By altering the paradigm of scientific publishing. I do think there is some progress to this end, with initiatives such as F1000 trying to speed up the publishing and arXiv allowing physicists to get data out before formal publication. However, to really promote an all-inclusive approach to science we must get used to idea that we have out grown our 350 year old way of communicating.
Solution: Alter way in which we communicate science so it is quick and all inclusive.