- A statement of what you are trying to prove
- any assumptions you’ve made about the records, the people, the time period, and the location
- the inconsistencies in the records and (with reasons) which version you believe to be correct
- the evidence you’ve used and how the evidence you’ve used makes your point
- if your conclusion “flies in the face” of a dearly held conclusion, consider including evidence analysis of why that conclusion is incorrect
- why you ignored or did not include certain pieces of information in your analysis
- the research process when it is key to the analysis (How you found a marriage record is not necessary. How you searched a census index to find “matches” for a three state area is.)
- complete citations for every statement taken from a document and for historical facts that are not common knowledge
- short summary of your argument along with a restatement of your original statement
The first two items on this list come from my mathematical background. In my worldview, there is no proof unless you state what you are trying to prove and clearly state your assumptions. You don't necessarily have to state them right at the beginning, but they should be stated. Twice for Casefile Clues, I've written up pre-1850 census families and both times I've included a statement of assumptions about the records. Those assumptions are also an important part of presentations I've given on the subject.
I'm not even certain this list is complete. However, if I think about all these items every time and try and make my point clearly, then I'm in pretty good shape. I also like to end any analysis with where to go next.
Because as we all know. One answer always brings more questions.