Dillon, Neo, and Seely's "Civil Keystrokes" (Ohio Communications Journal, 56, March 2018) is an excellent example of a research article summarizing a primary research study; it follows the established structure and includes all the standard disclosures that make a research article worthy of serious consideration and debate. Their study comprised a quantitative content analysis aimed at identifying relationships between anonymity and the civility and politeness of comments in the online comments sections found after the articles in daily newspaper websites. The researchers chose four newspapers based in Ohio given that state's political importance during presidential election years. For comparison purposes, they chose two newspapers which allowed commenters to be anonymous and two newspapers where the commenter was logged in using their personal Facebook account, thereby sharing their identity. The total content pool (sample) comprised 582 comments by 210 unique users across the four newspapers.

While I don't agree with the boldness of the authors' opening sentence under Discussion ("Our current study found civility and politeness, and possibly rationality, are hard to come by when users are anonymous to others"), the study did indeed conclude, with statistical significance (p < 0.001), that anonymity in general produces MORE comments and those comments are also LESS civil and LESS polite than comments from user-identified accounts. This may not be surprising to most of us, but the authors take pains to point out that this study fills gaps in the literature: for example, it's the first such study to measure comments on the exact same news article (the AP story was syndicated to all four papers) during the same period of time; it's the first study to clearly define civility and politeness as separate metrics (earlier research had conflated the concepts); etc. Personally, I found this last point of particular interest: Papacharissi's (2004) view of civility as a form of "collective politeness, as such people denigrating entire social categories of people is a form of incivility," was a novel way of explaining the difference. One positive finding, in my opinion, is that commenters of both types are more likely to assign stereotypes (a form of incivility, according to the authors) to "non-present others" than to their interlocutors (discussants) in the forum, which means, I suppose, that we humans are more likely to deindividuate (or dehumanize) an amorphous group of other people than we are the real individual person we're interacting with. By the way, one thing I appreciate about this study is the authors' contextualizing, particularly in the literature review, around the importance of open expression, specifically the potential value of incivility-breeding passion when discussing politics, for the health of democracy. I found that interestingly counter-intuitive.

Of course, the authors also recognized the inherent limitations in their methods, including the fact that just because a user has a Facebook account, that doesn't mean they're a real person (as opposed to a bot) or who they say they are. Given that the study involved latent content analysis (in which humans are required to read, evaluate, and categorize comments) rather than manifest content analysis (in which simply the presence of a word or image determines a comment's categorization), they also recognized that their coding of comments by fallible human beings leaves room for bias, although, to their great credit, they conducted a pretest to ensure coders were being consistent (this is referred to as interrater reliability). I also must bring up the timing of the study: the 2012 election seems like a century ago; the 2016 election and Trump's presidency brought about (or at least coincided with) dramatically increased divisiveness, intensified rhetoric, and plummeting trust in the news media, so it's natural to wonder how the results of this study would be different if conducted a mere decade later in 2023.

After reading the article in its entirety, please respond toat least twoof the following questions, each answer citing an excerpt from the study, and reply to your classmates throughout the week:

1. The authors mention several established theories throughout the literature review (for example, the SIDE model). Why was this necessary? What purpose does this serve? In your response, cite at least one example/excerpt from the study.
2. Aside from those mentioned by the authors, what other research limitations can you think of that would make one question the reliability of the results? Provide details. In your response, cite at least one example/excerpt from the study.
3. Most people would naturally assume that anonymity breeds less civility online. So, why was research like this necessary? How might the study have been redesigned, if at all, to provide more useful insights? In your response, cite at least one example/excerpt from the study.