Ben Goldacre

Even the BBC, usually so good, is terrible about this. And it's not like people haven't told them, they're just terribly intransigent about it.

http://bengoldacre.posterous.com/trying-to-get-the-bbc-to-link-to-journal-arti is Ben Goldacre writing in to the Beeb.

He later had a response ( http://bengoldacre.posterous.com/no-movement-on-the-bbcs-bizarre-links-policy ):

"Many thanks for your message, and interest in the site.
The issue of linking to specific research papers, rather than the journals in which they appear is something with which we have long wrestled.

Linking to the journal gives the reader the chance to look at the source of the story, understand something of the wider context and get an idea of whether the research is peer reviewed. It also gives a chance to read associated information in editorials, etc, which often may be of interest. It is also a way for us to credit the journal.

We accept that linking to the specific research paper itself is often a desirable thing to do, and it is certainly something we will try to do more of in future. However, we don't believe that it is something we can do on every occasion - for instance many papers are available on the web via subscription only, while others give only an Abstract summary. In these instances, the vast majority of our readers would not be able to read the full papers, without paying for access, even if we provided the relevant link.

In addition, we do have finite resources, and our view is that…

Our journalists routinely interview primary sources

That would be good journalistic practice, but that is not what Ben Goldacre is talking about in the blog post submitted here on HN. He is talking about news stories that purport to report on the result of a research study and what he is looking for is a citation (or a link, as appropriate to the journalistic medium) to the published research study. Not doing that doesn't pass the smell test for Goldacre.

http://bengoldacre.posterous.com/if-you-dont-link-to-primary-sources-you-are-d

Of course, even if the published study is cited, it still has to be examined carefully,

http://norvig.com/experiment-design.html

but there is no reason to hide the published study from the consumer of the journalistic report. You are referring to persons as "sources," the same way I did when I was a journalist, while Goldacre is referring to publications as sources.