For an example, I will use a famous case of really bad sources and a vicious historical hatchet job, namely, King John I of England (I know this blog is about Dracula's. This example does have relevance to Dracula. Be patient!).
The best structure for an evaluation uses the classic "who, what, when, where, and how" criteria. Here's how to apply these questions to a primary source.
Who?
Who created it? A monk? An adversary? A hagiographer? An admirer? A satirist - hey, are you taking a sarcastic book seriously? (I've done that. It's very embarrassing.) The only contemporary chronicles for King John for whom we know the authors are Gervase of Canterbury and Ralph of Coggeshall. There are scanty anonymous annals from monks from Dunstable, Worcester, and Tewkesbury, and the abbes in Margam in Glamorgan and Waverly in Surrey.
When?
When did they write it? Are they a contemporary of the person or action? Did they write it later using their memories or rumor? John of Wendover wrote his stories after 1226. Matthew Paris began plagiarizing Wendover's stories starting in 1235. King John died in October 1216. John's son, Henry, did not take the throne until September 1217. Does something about those sources look funny to you, too?
What?
What did they create? A book, artwork, or an object? What kind of condition is the item in now? What purpose did it serve? Roger of Wendover very honestly stated that his book was not a history; he wanted an effective foil for sermons against various sins.
Where?
Note that there are regional and temporal differences in climate of opinion. Also, different geographical climates preserve things better than others. So we have wonderful artifacts from dry climates like Egypt, but few from moist climates like SubSaharan Africa. Remember, books are destroyed in a particular pattern as are scrolls and pottery.
How?
Did the creator use eyewitnesses? Other accounts? Is the account littered with unsubstantiated rumor? A good question to ask is how good is the rest of the source. Roger of Wendover's "chronicle" of King John is full of stories that no-one would take seriously. He writes of a washerwoman who broke the Sabbath to work and was exsanguinated by a small black pig for punishment. He writes another story of a loaf bread that because it had been baked on Sunday, ran with blood when it was to be sliced.
"There is one, (it is eighteen pages long) about a peasant named Thurkill from the village of Twinstead in Essex, who, in 1206, was led through the realms of Purgatory by St. Julian. As Wendover tells it, the story has many realistic touches, from the man's name and place where he lived to precise details about the torture chambers of the underworld: in one, for example, stand cauldrons of inky water so bitter that if a piece of wood is thrown in the bark instantly peels off. It is a grim and lively story; but is it true? Wendover certainly seems to think it as authentic as his stories about John; and it is difficult to see on what grounds historians should reject the former while accepting the latter." Warren. P. 11
Now, what if don't have a smoking gun" like the King John example? What if you just feel uncomfortable about using that source? Many historians are nervous about basing a text on controversial "primary sources". Historians will tell you to do another search to see if you can find any sources to supplement, replace, or confirm the references you are using from these sources. The citation will have more authority if you are able to back up your point with multiple sources.
The historiography for this section is based on W.L. Warren's book, King John, published in London by Eyre & Spottiswoode c. 1961. I highly recommend it.
