User:Runfellow/Thoughts on peer reviews
The purpose of this essay is to elaborate on some of my views regarding peer reviews, mostly for the sake of anyone who might be undergoing one of mine.[note 1] I enjoy the process of peer reviewing because it allows editors to suggest changes rather than impose their views. The "peer review process", as defined in this issue of The Signpost and on WP:PR, is deliberately vague, and I'm fine with that. People will conduct peer reviews in their own way, and there should be plenty of leeway as to which style would work best for a particular article. Reviewers should find a way to merge their personal style with what has been requested.
What a peer review should be
- Thorough – A reviewer should at least read the entire article, and the comments should cover the article from beginning to end. Yes, I nitpick, but I also include broad statements about the general tone, structure, or style of the article. Nitpicking allows the reviewer and the requesting editor to spot any small-yet-persistent issues.
- Organized – Bullet points are the most common format for comments, since they allow an editor to address each issue specifically. They're not the only format, though. Insightful prose allows an editor to get a more in-depth perspective. My preferred style is to mix in some longer text with shorter, easier fixes.
- Constructive – Comments should be focused on improving the article, not just critiquing it. Sometimes these criticisms come off as cold or mean-spirited. I promise that's not what I intend!
- Specific – A good review will be specific as much as possible and will include suggestions on how to fix just about every issue. If the article requires extensive work, the review should point the requesting editor in the right direction towards improving it (i.e. a WikiProject or WP:GOCE).
- Engaging – As Ruhrfisch stated in a 2008 interview, "the most effective thing is dialog – peer review should be a conversation about improving an article." Practically speaking, though, that dialog is often elusive. We would all like to have multiple contributors submitting comments from various perspectives, but the fact is that those kinds of reviews aren't common. More often than not, one person provides most of the comments, and one person responds to them.
What a peer review should not be
- Comprehensive – Expecting one stranger unrelated to the article to go through every minute detail in the peer review process is asking too much. The difference between "thorough" and "comprehensive" is that a review should cover all of the major points, but not necessarily all of the minor ones.
- Sparse – All too often, reviewers insert a few random comments, just enough to say it's "done". Consequently, these reviews sit unfinished until they're archived. That leaves the submitter with very little feedback, which can negatively affect their GA or FA review.
- Spastic – A review shouldn't expect other editors to follow along as the reviewer casually stumbles across an issue. If the reviewer isn't moving through the article section by section, he or she should make it clear what system is being used.
- Specialized – More often than not, peer reviews will come from someone who is not intimately familiar with the subject. That's a good thing. It's very easy to get caught in the process of arguing over specific details and lose track of the more important parts of Wikipedia: accessibility and clarity.
- A featured article/good article review – Peer reviews are not objective, nor are they based on any set criteria. They do not function as bulwarks against an article's progress to a higher status. Additionally, since peer reviewers are no more important than any other editor, I do not approve of the common refrain in GA/FA reviews, "you haven't addressed all of the issues mentioned in the peer review." No one is obligated to accept the suggestions of any peer reviewers.
My peer reviews
My reviews often include issues with the lead section, prose, and the Manual of Style. In my comments, I tend to focus on:
- Clarity – Can I understand what you're saying, not just what you're trying to say? People often write in a way that they think impresses rather than clarifies. Unfortunately, it often does neither. Editors often mistake Wikipedia's summary style guideline by combining sentences rather than summarizing important information. "X said 1. Y said 1. Z said 1." can be summarized as "X, Y, and Z said 1", but it should not written as "X said 1, with Y also having said that, and Z, having concurred with X and Y, speaking a similar note by stating 1."
- Language – Are you using phrases that would make sense to an English learner? Are you using a specific lexicon where plain English would suffice? We often insert aphorisms, personification, and technical terminology into an article where it shouldn't be.
- Inadvertent NPOV issues – Most people on Wikipedia avoid NPOV issues. That's great. But there's a difference between overt vandalism and the subtle NPOV issues I see so often. For example, if an editor writes "Team X finally scored on the second half" in an article about a particular sports contest, the word "finally" injects the author's opinion because it implies that it took a long time for the team to score. I also tend to look for words like "really" and "actually", and a very common and sneaky one, "just". As in "Person X had just N widgets." If it is a small number, people will form their own opinions. You could add "just" to almost any figure to make it look worse.
- Structure – Many articles' structures are predetermined, but that doesn't mean it can't be improved. The process of editing the article one piece at a time often creates structural issues. Bold structural changes are daunting and frustrating, but often necessary.
List of reviews
Featured or Good Article status does not necessarily mean I had anything to do with getting it to that status, although for some articles that may be the case.
- If you're one of those folks and you're happy with your review, perhaps you'll consider something nice as a reward?