Research Your Writeup!
Good writing demands research. Even if you write pure fiction, research helps.
I've been making a living off my research for many years. I'm a lawyer. My specialty is doing research and writing for other lawyers. Everything I write gets torn apart by someone with an axe to grind. I have a lot of practice doing bullet-proof research. (Stop me before I mix more metaphors!)
Know your standards
You have to know the criterion for "truth" in your field. In science, something is "true" if it can be demonstrated empirically by a repeatable, verifiable experiment. Thus, if someone is reporting a scientific breakthrough, the first question is whether it has been independently verified. If not, it's just a research announcement.
Similarly, good journalism requires corroboration of sources. If you have only one source, it's at best an "allegation".
Know your sources:
In law, we have a very strict hierarchy of sources. Depending on where you live, some laws are relevant, some are not. Law is constantly changing, so you have to have up-to-date sources. Law books I used in law school, thirteen years ago, are now almost entirely useless.
When it comes to other fields, the sources may not be so easy to rank, but you can often rate them. The New York Times is probably more reliable than The Weekly World News. Tax advice from the IRS is notoriously unreliable. Tax advice from your CPA, on the other hand, is likely to be close to 100% correct. Why? You can sue a CPA.
Sometimes unreliable or biased sources are useful as a starting point. You can find out a lot of useful medical information from drug company advertisements, for example. Just don't accept it as the last word. Which leads me to the process of research:
Follow the cite trail or corroborate
Some sources cite other sources. In formal legal writing, for example, every sentence must be followed with a citation supporting the assertion. Academic papers, likewise, have lots of footnotes. To do complete research in these fields, you must check all the relevant cites, then check relevant citations in those sources, and so forth. Eventually, if you have narrowed your research topic sufficiently, you start going in circles. New sources start citing sources you have already read. If that doesn't happen, either (1) your research topic is too broad, or (2) your research is not complete. With sources that cite sources, the trick is framing your research question narrowly enough.
Some material doesn't cite sources. For example, print journalism may "attribute" or refer a quote to a "source", but not give a cite. You get "the OMB reports that ..." not "See OMB Report, at page 5". Without cites you can't follow a cite trial. What you do instead is corroborate: you get as many sources on the same facts as possible and see what matches.
The trick with corroboration is knowing the difference between corroborating and merely duplicating the same story over and over. Matching facts stated in different ways from different viewpoints is ideal. It is not necessarily a good thing if all your sources are saying the exact same thing in the exact same words. For example, you can frequently get articles from a dozen newspapers all saying exactly the same thing, because they all copied and pasted the same wire service article. This is not corroboration, this is duplication. True corroboration requires primary sources with a useful or credible viewpoint: an eyewitness, a document, etc.
Either method --looking up cites or getting corroboration-- gives you multiple sources. Using multiple sources forces your writing into your voice and your point of view, which is what we want here.
NB: Once you *have* sources you should cite them: E2 FAQ: How to cite your sources. This has summary of the MLA cite forms for several different kinds of sources.
Our brother editor Cletus the Foetus observes, pedantically albeit correctly, that "'Cite' is a verb -- the noun is 'citation.'" I've been using "cite" as a noun since my law school days. It doesn't grate on my ears, but I work for lawyers, socialize with lawyers ... heck I'm even married to a lawyer. Perhaps you should avoid this cant if you are not surrounded by lawyers.
True Confessions of a Content Editor
I'm not naming names anymore. You know who you are.
After whining to bones about how I was too busy and it was too much work I updated the Iraq debate metanode collection by creating War on Iraq 2003. I then firmlinked it to War on Iraq 2002 and vice versa, which firmlinks were promptly stripped out by bones (firmlinks, with few exceptions, are supposed to point from an empty nodeshell to the node with content.
Proving once again that editors can be n00bies, too.