That is not entirely true but is very common, and I agree that negative results are very useful. In medicine in particular this has been the cause for a lot of debate recently and as such more people are beginning to publish negative results.
Its basically down to funding and publications. A publishing company wants to get lots of interesting science that will give it headlines or make it money. So they will often only publish things that do that for them. Publishing something that is wrong (even if they work out why it is) often doesn’t give big headlines so they wont publish it.
The second is funding. It is hard to apply for money to do negative testing or hard to get more money if you don’t get what you want. So most people carefully word their applications for money so they are working on something they know is likely to give good results so they are more likely to get the funding. Without funding we simply can’t do science.
Part of it is is also attitude. Some scientists don’t want to be told they are wrong or show they have failed to other people so they don’t publish the work. That i find really annoying as you can guarantee that there is someone else out there making the same mistake and wouldn’t have that problem if they published work saying “don’t bother doing it this way, it doesn’t work”
However having said all of this scientists are smart and can take a wrong result and try to look at in a way where it is actually a positive – “Ok our results for that are wrong but we have shown that there is variability between these things or a link between this and that”
So a lot of science is still ‘look we got positive results’ but they are actually wrong results that have been looked in a different way and found something useful from them.
I agree that many scientists don’t want to publish their negative results, as they aren’t ‘significant’, but knowing what went wrong is still very important!
Normally this is written in the results section of a paper. A paper is what scientists publish their results in. So when other scientists read this paper, they know what they did to improve their experiment, and get their results.
Really good question and one that has been suggested by many scientist. Having a journal of failed experiments or something along those lines. Someone will need to champion it and even if made free online scientist should be ready to use it. It could happen though because just like publishing results of clinical trials including the failed ones which many pharmaceutical companies did not want to do, a campiagn now has many signed up to publishing all clinical trials data. Check it out here http://www.alltrials.net/
Everyone has given such great answers already, I’m not sure I can add much more! Sometimes it’s disappointing to get the opposite results to what you were hoping, especially if you’ve come up with a good idea or theory about why it should happen, but we can still learn a lot from that!