DOI stands for Digital Object Identifier, and in short is a permanent ‘address’ (like a URL) which will always point at a particular document. In particular DOI’s are used to make it easy to find academic papers, and saves one having to find the journal’s webpage, navigate to its archives, pick up the volume in question, and finally find the paper. More importantly unlike a URL you know that the DOI won’t change, and will still work when the journal decide to re-organise their system. For example, the DOI of Crick ans Watson’s DNA paper is doi:10.1038/171737a0.
As a scientist I have become used to reading heavily referenced papers and reviews. If someone says something they back it up, either by demonstrating it themselves, or otherwise pointing out where someone else has shown this. This will usually result in a huge list of references which, especially in review articles, can stretch on for an entire page or more. Typically a reference will provide the authors’ names, the year, the title of the paper (occasionally), the journal the paper appeared in (usually abbreviated), the volume, the issue (occasionally) and the page number. The DOI is also appearing in the bibliography with increasing regularity, and it certainly makes tracking down references much easier. But my issue isn’t with the academic journals so much as the newspapers.
Mainstream media coverage of science stories is often shockingly inaccurate and simplified to the point of inaccuracy. While a certain amount of this is understandable, your average science paper being outside the ken of most folks (and in some fields most scientists), it does leave me itching for more details. And here is the biggest problem… newspapers don’t reference. If you’re lucky you may get the journal, and then its a matter of clicking back through the past couple of issues, matching titles and abstracts with the story in the paper. However in many cases even this information is skipped, leaving me in the dark. (Of course, on some occasions the story in the paper reflects an as yet unpublished paper. Shockingly in some cases they haven’t even passed through peer review.)
Now I understand that in your average newspaper space is at a premium, they don’t want to spend column inches devoted to meticulously citing the paper they are reporting (Although frankly there is no excuse on the web). However a DOI is short, and provides all the information needed. It would allow for rapid finding of the paper in question, allowing people who are interested to find out the nitty-gritty. Furthermore, if a person happens across the article in a library archive in fifteen years time the DOI will still work.