I find myself dissatisfied with the state of the art of document review. At work we have formal review processes that involve checking boxes, pressing buttons on server applications, receiving change requests, etc; but what none of these procedures enforce is adequate reviewing of the actual document. The only time I’ve seen documents reviewed in a significantly better way than here is when I submitted my final project at University.

Amusingly, http://www.stsc.hill.af.mil/consulting/sw_testing/defect/ddr.html has a course on effective document review, in which they will “develope” (sic) effective document rules and checklists. That web page wasn’t very effectively reviewed, then.

I think checklists are dangerous. For someone with expertise and experience, a checklist is a useful way of helping prevent oversights. For someone with less of either, it provides a strong encouragement to perform a superficial covering of the points in the checklist. If a checklist says “check tank” the novice may say to self: “yep, there’s a tank there” whereas what is really meant it “check that there is enough fuel for the journey considering the journey’s length, the traffic and the terrain and whether it will be possible to stop and find more fuel on the way”. In the context of document review, this might mean that the reviewer ends up checking the presence and formatting of paragraphs without thinking about what the target audience will conclude from the content.

I would suggest a better way to review a document is to imagine you are going to explain the subject matter to someone, and try to use the word “because” a lot. This forces you to take a deeper and broader view than if you are simply skimming for factual implausibilities.

“Because” is a really important word, it’s the answer to the question “why?”. If there is no “because” to match a “why” then there is no purpose for doing something and time is being wasted.