Chapter 5

The Role of Hyperlocal Media: Content Analysis Methodology and Findings

In order to examine the relationship between hyperlocal media and community engagement, we need first to understand the unique role that hyperlocal media serve compared with other media. To do so, this study employs a content analysis that compares a selection of hyperlocal media that cover Columbia Heights (exclusively or in combination with hyperlocal coverage of other neighborhoods) with WashingtonPost.com, Washington’s most-read online news source (Gloria and Hadge 2010). WashingtonPost.com provides a useful contrast for two reasons. First, from a theoretical standpoint, online hyperlocal media can be seen as filling the gaps left by shrinking coverage from local print newspapers; since Washington has few neighborhood newspapers remaining, the metropolitan daily Washington Post is left with much to cover if it is to be the city’s primary news source, a daunting task that would require hefty resources. Second, in 2010 WashingtonPost.com redesigned its homepage for local news as PostLocal.com and introduced efforts to encourage active participation from readers, such as “The Daily Gripe,” powered by the government transparency tool SeeClickFix (discussed in Chapter 3). Both of these points suggest The Washington Post website could play a significant role in providing news about the city neighborhood’s. However, this does not mean the Post has the manpower to cover the entire city at the same granular level of hyperlocal news outlets, nor does it have demand from its diverse, metro-area readership for news about all the city’s neighborhoods. This chapter proceeds with the methodology, findings, and discussion of the hypotheses tested by the content analysis and concludes with a discussion of observations specific to list-servs, Facebook, and blogs.

Methodology

If hyperlocal coverage is not a good fit for the Post, this opens the door to hyperlocal media to fill the information gaps that are not met by the demand of Post readers or the resources of its reporting staff. As the findings will show, the online news website medium is not the only one that is equipped to provide this kind of coverage. This chapter combines the results of a three-day content analysis with data gathered through the survey discussed in Chapter 6 to answer the following research question:

RQ1: How are hyperlocal digital media used in Columbia Heights?

The ensuing subsections will describe the selection of the sample for the content analysis and survey data, the coding scheme employed in the content analysis, and the results of an intercoder reliability analysis.

Sample

Part of the data used to answer this question were gathered through the survey administered online and offline to those who participate in Columbia Heights’ public life offline and use its hyperlocal media online.[1] Survey questions about respondents’ media use measured consumption and participation, as well as a range of possible uses and gratifications. In recognition of the frequent intersection between citizen journalism and hyperlocal media, questions asked whether respondents had ever e-mailed a list-serv, posted to a Facebook page, or commented on a blog or website that focused on Columbia Heights. Two questions (#21 and #22) asked respondents to choose all gratifications they sought from hyperlocal media, as well as their primary reason for using hyperlocal media. The first three options covered general uses and gratifications: surveillance, entertainment, and socialization. The fourth—“to connect with or meet others, online or offline”—was informed by conversations with my colleagues at Georgetown about how they used hyperlocal media upon first moving to Washington. The remaining items name specific subject areas about which respondents might seek additional information. Some of these — such as “local restaurants” — were also informed by my colleagues’ comments. Others were adapted from a previous study (Hadge 2010b) in which a content analysis revealed subjects such as weather, transportation, and events appeared regularly on hyperlocal media outlets in Seattle. The remaining subjects were based on those indicated as “information needs” in a community by the Knight Commission on the Information Needs of Communities in a Democracy (Mayer, Olson, et al. 2009).

The hyperlocal media sources included in the content analysis were chosen with several criteria in mind. First, the blogs included were chosen because they either cover news in Columbia Heights or provide information on getting involved in the community. Prince of Petworth, although technically not a hyperlocal blog focused on Columbia Heights, was included because about 30% of its readers are from Columbia Heights (D. Silverman, pers. comm.) and because the way that the blog covers news is consistent with other hyperlocal media: providing information of a level more granular than that provided by media with a metro-wide audience, and facilitating interaction with readers. Two news aggregator features from Outside.In and TBD.com were included in order to capture news from other outlets than those in this study, in case there were additional sources of coverage for the neighborhood. The list-servs included were those of which I had previous knowledge and was permitted to subscribe to (some list-servs require subscribers to be residents of the neighborhoods they cover). For the Metropolitan Police Department list-servs, only the most encompassing of three Third District list-servs was included, to avoid redundancy. Because many of the Twitter feeds and Facebook pages or groups associated with Columbia Heights are inactive or dormant, those included here were chosen for having some activity on the days included in the content analysis. The following figure lists the hyperlocal media that are included in this content analysis.[2]

Figure 5.1: Media outlets included in content analysis

Source Medium Description
WashingtonPost.com Website Associated with The Washington Post
The Heights Life Hyperlocal blog Solely covering Columbia Heights
New Columbia Heights Hyperlocal blog Solely covering Columbia Heights
North Columbia Heights Civic Association blog Hyperlocal blog Provides updates from the civic association
Prince of Petworth Metro-wide blog Provides hyperlocal news on Columbia Heights and other neighborhoods
TBD.com News aggregator Regional news website that includes an aggregation feature by ZIP code; filtered here to cover 20009
Outside.In News aggregator National news aggregator; filtered here to cover 20009 ZIP code
columbia_heights Yahoo! Group/list-serv Covers all of Columbia Heights
southcolumbiaheights Yahoo! Group/list-serv Covers South Columbia Heights
WardOneDC Yahoo! Group/list-serv Covers all of Ward 1, of which Columbia Heights is a part
MPD-3d Yahoo! Group/list-serv Covers the third district of the Metropolitan Police Department, of which Columbia Heights is a part
NCHCA Yahoo! Group/list-serv Associated with the North Columbia Heights Civic Association
NWCHCA Yahoo! Group/list-serv Associated with the Northwest Columbia Heights Community Association
@newcolumbiahts Twitter Associated with New Columbia Heights
@ColHeightsDay Twitter Associated with an annual late-summer neighborhood event, Columbia Heights Day
@PoPville Twitter Associated with Prince of Petworth
@theheightslife Twitter Associated with The Heights Life
@JimGraham_Ward1 Twitter City council member for Ward 1
Columbia Heights Dog Park (11th&Park) Facebook group Niche group for those who utilize a neighborhood dog park
North Columbia Heights Civic Association page Facebook page Associated with the civic association

Coding Scheme

The content analysis was conducted using data gathered from three days in February 2011: Wednesday, February 16; Monday, February 21; and Friday, February 25. These dates were chosen to capture activity at different times of the week, when the news cycle might progress at different rates. Several days’ space was also provided between collection dates to allow some time for the included sites to publish new content, as content is not necessarily updated on a daily basis. All posts or articles published on a given day were captured using the Mozilla Firefox browser add-on “Scrapbook,” which preserves a snapshot of the website for later reference, and Yahoo! Group/list-serv posts were collected via a daily automated e-mail digest received the morning after the collection date. For WashingtonPost.com, a search was conducted each day for the terms “Columbia Heights” and “Ward 1,” and only stories containing those search terms were coded, rather than coding the entire website. This was done to reflect the way in which users are anticipated to approach these sites: It was assumed that people looking for news on Columbia Heights on WashingtonPost.com would search for it, rather than comb through the whole site, while people looking for news on Columbia Heights on one of the hyperlocal media outlets would look at whatever content appeared there, since there would be less content through which to sort.

The content analysis tracked a number of characteristics, which were adapted to the structural differences of each medium. There were a total of 60 cases: 20 sources, each coded over three days. For each of these 60 cases, 20 variables were tracked. First, the number of posts was recorded, according to the standards for each medium—blog posts for blogs, headlines for aggregators, wall posts for Facebook, tweets for Twitter, and individual e-mail messages/posts to the Yahoo! Group for list-servs. Second, the number of comments (or replies) was calculated for each medium to compare the amount of two-way interaction between the medium and the reader. The designation of what constituted a “comment” was straightforward for blog posts and online news articles, with the exception that if a post or article republished from another site, the comments were not counted. For example, if TBD.com included a TBD-authored article in its list of aggregated stories, the comments would be counted. If the aggregator list on TBD included a story from WashingtonPost.com, the comments would not be counted. This was done to ensure that comments were counting readers’ engagement with the site on which they encountered the story (or content). On Facebook, direct replies to wall posts were designated as comments. On list-servs, replies to an e-mail message/post were designated as comments. On Twitter, a retweeted post counted as a reply if the retweeting poster included his or her own comments before the “RT” in the tweet, or if the original poster of a tweet mentioned another Twitter user (designated by an “@” symbol and the other user’s Twitter username).

Next, all of the stories in each case were tallied within the following three categories: original or non-original content, stories about Columbia Heights or not about Columbia Heights, and stories about one of 14 subject areas. For blogs, online news articles, Facebook posts, and e-mails, original content was counted as anything that was written specifically for the medium in question (including a press release sent by the organization from which it originated, because this is by nature meant to be distributed or copied to multiple sources), or any link or reproduced content that was accompanied by two sentences or more from the author who was reproducing it. The designation of two original sentences was meant to weed out cases in which an author or poster simply directs attention to the republished content with one sentence or sentence fragment. For Twitter, original content included any unique tweet, as well as retweets that were preceded by a comment from the person retweeting it. Content was counted as being about Columbia Heights if it discussed information specifically about any area, person, or organization that fell within the boundaries of the neighborhood used for this study.[3] The only exception made was for political news: If content was about Ward 1, in general, it was counted as being about Columbia Heights, because the information that affected Columbia Heights residents could not be separated from that which was relevant to residents of other neighborhoods in Ward 1.

Finally, all stories or posts were coded as falling into one of the following 14 subject areas: politics, crime, events, transportation/traffic, weather, business, sports, arts and entertainment, restaurants, jobs/employment, real estate, education, health and social services, and other. These subjects were chosen to align with those given in the study’s survey questions (Q21 and Q22). When a post covered multiple subjects, it was categorized according only to the predominant one. For example, a post about an ANC meeting might be construed as both an event and a political story; for this study, it was coded as a political story, because within the realm of political news, hyperlocal media are best suited to covering something within the narrow scope of an ANC.

Intercoder Reliability

The content analysis was also tested for intercoder reliability. In doing so, I explained the methodology given above to the second coder. The second coder was given access to all the data collected and instructed to choose at least 10 cases to code. I was present during the entire time the second coder worked on the sample and answered questions that arose as to how cases should be evaluated according to each variable by providing more detailed information, without indicating how I would code the case in question. The second coder coded a total of 14 cases. In two of the cases, there were significant discrepancies in the number of stories counted. These two cases included stories from days outside the scope of the content analysis, but the coder counted these as part of the total, instead of just counting the stories from the day in question. I removed these two cases from the sample before performing the reliability, so as to allow for a result that more accurately reflects the overall reliability of the coding scheme. Among many of the variables, there were too few cases to carry out a test of reliability. The results of the reliability tests appear in Methodological Appendix C (Table C.6).

Hypotheses

The hypotheses tested in the content analysis are predominantly descriptive. Four of the five hypotheses examined in this chapter concern coverage of several topics out of the list of subjects examined in the content analysis and questions 21 and 22 of the survey: politics, crime, events, and real estate. All of these were chosen because of the 14 subjects coded for, these four are the ones that, when particular to one neighborhood, will not likely be as relevant to residents of another neighborhood as other subjects might.

H1: Columbia Heights’ hyperlocal media provide information on neighborhood politics more often than WashingtonPost.com.

H0: Columbia Heights’ hyperlocal media provide information on neighborhood politics less often than WashingtonPost.com.

H2: Columbia Heights’ hyperlocal media provide information on neighborhood crime more often than WashingtonPost.com.

H0: Columbia Heights’ hyperlocal media provide information on neighborhood crime less often than WashingtonPost.com.

H3: Columbia Heights’ hyperlocal media provide information on neighborhood events more often than WashingtonPost.com.

H0: Columbia Heights’ hyperlocal media provide information on neighborhood events less often than WashingtonPost.com.

H4: Columbia Heights’ hyperlocal media provide information on neighborhood real estate more often than WashingtonPost.com.

H0: Columbia Heights’ hyperlocal media provide information on neighborhood real estate less often than WashingtonPost.com.

H5: Columbia Heights’ hyperlocal media include more online conversation than WashingtonPost.com.

H0: Columbia Heights’ hyperlocal media includes no more online conversation than WashingtonPost.com

Data Analysis and Findings

Does Coverage Meet the Demands of the Public?

To set the stage for the findings related to the above hypotheses, let us first compare the coverage of all subjects tracked in the study to the public’s demand for them. In order to know if hyperlocal media are meeting the needs of the public, we must first know what those needs are. By turning to the survey data, we can compare the most popular gratifications sought by survey respondents in hyperlocal media[4] with the frequency of their appearance in the content analysis. In Figure 5.2 I used these equations to calculate “demand” and “coverage”:

Demand = Number of “yes” replies to Q. 21 / Number of respondents

Coverage = Number of stories on a given topic in entire study / Total number of stories in study[5]

Figure 5.2: Comparing Coverage and Demand for Hyperlocal Media Gratifications[6]

Highest Coverage of Media Gratification (% allotted) Highest Demand for Media Gratification (% who indicated interest in this)
Comments (Mean: 2.69 per story; 50% of sources had comments/replies) Respondents who use hyperlocal media to connect with or meet others, online or offline (19.4%)
1. Restaurants (20.3%) 1. Local events (90.9%)
2. Local politics (19.3%) 2. Restaurants (87.5%)
3. Crime (10.9%) 3. Business (84.7%)
4. Local events (6.7%) 4. Crime (81%)
5. Real estate (5.7%) 5. Arts and entertainment (76.5%)
6. Business (5.2%) 6. Local politics (66.9%)
7. Arts and entertainment (4.2%) 7. Traffic/transportation (56.8%)
8. Education/schools (3.6%) 8. Real estate (53.7%)
8. Health and social services (3.6%) 9. Weather (30%)
8. Traffic/transportation (3.6%) 10. Education/schools (27.1%)
8. Weather (3.6%) 11. Health and social services (23.4%)
9. Jobs (2.1%) 12. Sports (12.4%)
10. Sports (0%) 13. Jobs (11%)

While coverage and demand do not match exactly, the priorities are roughly the same. In looking at the top six results in each column, the only discrepancies come in the areas of real estate and arts and entertainment. Coverage of real estate news exceeds the demand for it by three places, and coverage of arts and entertainment ranks two spots lower than the public’s demand for that information.

However, it became clear that not all hyperlocal media sources prioritize the same topics in their coverage or discussion. Instead, particular sources specialize in certain areas, and this was especially apparent because the data for this methodology are drawn from a sparse sample as a result of infrequent updating by some of the included sources. The following sections will therefore compare total coverage across hyperlocal media sources in the aggregate with that of WashingtonPost.com, with a few illustrative examples from each medium. Further discussion of the specialization of hyperlocal media can be found at the end of the chapter, and findings for each outlet included in the content analysis can be found in Methodological Appendix C (Tables C.1-C.4).

Neighborhood Politics

As seen in Table 5.2, coverage of local politics accounted for 19.3% of all stories coded in the study, the second highest topic covered, while survey respondents ranked it their sixth most frequent reason for using hyperlocal media. An analysis of the data gathered tested the following hypothesis:

H1: Columbia Heights’ hyperlocal media provide information on neighborhood politics more often than WashingtonPost.com.

As Table 5.1 shows, we can support H1 because, taken together, the sample of hyperlocal media outlets included greater than 12-times more political coverage on Columbia Heights and Ward 1 than did WashingtonPost.com.

Table 5.1: Percentage of News Coverage Devoted to Local Politics Throughout Study

Source # Political Stories % of Total Coverage
Mean # stories: 1.8
Total # stories: 37 20%
WashingtonPost.com 3 n/a[7]

Looking at individual hyperlocal media sources, the average number of political stories covered in the three-day study is less than that on WashingtonPost.com. However, three list-servs (columbia_heights, southcolumbiaheights, and WardOneDC) had more than three items on neighborhood politics, exceeding the political content than WashingtonPost.com. To be more specific, then, we can say that hyperlocal list-servs provide more information on neighborhood politics than the Post website. Included among this coverage was a notice about an upcoming special election (Boese 2011) sent to both the “columbia_heights” and “WardOneDC” list-servs and an agenda for an upcoming ANC 1B committee meeting sent to the “southcolumbiaheights” list-serv (Irwin 2011). In contrast to these announcements, the political information on WashingtonPost.com included a short news item on a former politician’s guilty plea to corruption charges; the item featured a comment on the outcome from Ward 1 Councilmember Jim Graham (DeBonis 2011). These examples illustrate the difference between political information as direct communication between the offices of elected officials and citizens (as in the list-servs’ announcements) and as information processed by a journalist intermediary.

Neighborhood Crime

The next hypothesis turns to news and information about crime. The media sources surveyed in this study include some specific crime coverage: The MPD list-serv publishes Daily Crime Reports and Daily Arrest Reports, and PostLocal.com has a page devoted to crime.[8] While Columbia Heights is included in the third district of the Metropolitan Police Department, MPD-3d covers a broader area than Columbia Heights alone, so not all information on the list-serv was necessarily categorized as being about Columbia Heights. The following hypothesis was tested:

H2: Columbia Heights’ hyperlocal media provide information on neighborhood crime more often than WashingtonPost.com.

As with H1, a summary of coverage is presented in Table 5.2.

Table 5.2: Percentage of News Coverage Devoted to Crime Throughout Study

Source # Crime Stories % of Total Coverage
Mean # stories: 1.1
Total # stories: 21 11%
WashingtonPost.com 0

The data above allow us to reject the null hypothesis in support of the hypothesis that hyperlocal media sources provide information on crime more often than does WashingtonPost.com’s coverage of Columbia Heights. While WashingtonPost.com may have provided coverage of crime outside Columbia Heights during this study, hyperlocal sources provided more crime coverage specific to the neighborhood. This shows an interesting progression from Elizabeth McIntire’s observation (discussed in Chapter 3) that at the turn of the twenty-first century, Columbia Heights was often characterized negatively in the media, (E. McIntire, pers. comm.). In the 1990s in particular, several high-profile crime stories were set in Columbia Heights, as described in Chapter 2.

Instead, hyperlocal crime coverage in the content analysis incorporated ways of preventing crime. For example, a blog post on Prince of Petworth included an e-mail from a reader looking to get the word out about a man loitering in the neighborhood who was making her feel threatened. Prince of Petworth publicized the message by posting it and also recommended that the reader contact the police, despite her uncertainty that the man was actually doing something criminal (Silverman 2011b). An e-mail to the “MPD-3d” list-serv, on the other hand, was looking for the assistance of the public in identifying suspects in an armed robbery (Kishter 2011). In this case, the MPD took advantage of their opportunity to target a specific audience of those who are already interested in receiving crime-related information and presumably spend time in the area in which the crime was committed.[9] Bypassing mainstream media, both MPD and the concerned Columbia Heights resident in these examples are able to use hyperlocal sources to communicate directly with their target audiences.

Neighborhood Events

With the decline of the newspaper classified ad, many hyperlocal media have become a free way to publicize community events. While these are sometimes of interest to a broader audience than those who live in the neighborhood, there are many events for which the opposite is true that would not capture the interest of an audience outside a specifically local community. These can range from the annual Columbia Heights Day, an event that bypassed intermediaries in setting up its own Twitter feed, to reminders of ANC or civic association meetings or events posted on blogs and sent out to list-servs. With this in mind, I hypothesized the following:

H3: Columbia Heights’ hyperlocal media provide information on neighborhood events more often than WashingtonPost.com.

Table 5.3 summarizes the study’s findings.

Table 5.3: Percentage of News Coverage Devoted to Local Events Throughout Study

Source # Local Events Stories % of Total Coverage
Mean # stories: 0.7
Total # stories: 13 7%
WashingtonPost.com 0

Again, the data show support for the hypothesis that hyperlocal media provide more coverage of local events than does WashingtonPost.com. One illustrative example of the way in which hyperlocal media are expanding upon the role previously served only by local (physical) bulletin boards is a message sent about upcoming events at the Emergence Community Arts Collective to the “WardOneDC” list-serv (Robinson 2011). On the other hand, not all hyperlocal media are the best sources for event information. Outside.In linked to a round-up of events in The Washington Examiner that included one in Columbia Heights, but the story included in a February 16 screen-capture of Outside.In had originally been published February 1, making the information potentially outdated (“Around Town” 2011).

With a maximum of 2 items about events, hyperlocal media did not cover this subject extensively as one might expect, given that it was the number one use for which survey respondents seek out hyperlocal media. This can perhaps be attributed in part to the uneventful time of year during which the study was conducted (February). Nonetheless, the three days did show enough content that local events ranked fourth out of all subjects in terms of its coverage over the course of the study.

Neighborhood Real Estate

Real estate is another subject about which information could once be procured through the advertising circulars of print newspapers. How are hyperlocal media complementing this traditional approach? The fourth topic-oriented hypothesis tested the following:

H4: Columbia Heights’ hyperlocal media provide information on neighborhood real estate more often than WashingtonPost.com.

The table below shows the extent to which the media surveyed actually did cover real estate, which area included information on rentals, home ownership, and commercial real estate:

Table 5.4: Percentage of News Coverage Devoted to Real Estate Throughout Study

Source # Real Estate Stories % of Total Coverage
Mean # stories: 0.58
Total # stories: 11 6%
WashingtonPost.com 0

Because WashingtonPost.com did not cover real estate information in an editorial section (classified ads were not included in the study), there is support for the hypothesis that hyperlocal media provide more real estate news and information than The Post. By examining the cases in which information about real estate appeared, it became clear that hyperlocal media do not only include listings advertising available real estate; they also provide advice and information about the real estate market. For example, Prince of Petworth includes a regular feature titled “Good Deal or Not?” that features a real estate listing and looks for reader input on the price. To supplement these, the blog publishes occasional guest posts from a local realtor called “GDoN Revisited” that follows up on the fate of the property first listed on Prince of Petworth and offers analysis of why the sale turned out as it did.[10] This provides a unique opportunity to readers for not only acquiring real estate information but also gaining insight into both public opinion and a professional’s opinion on the real estate market.

Conversation

As the above real estate example demonstrates, publishing information or news on particular subjects is not the sole concern of this study. What sets hyperlocal digital media apart from other platforms is their ability to facilitate conversation. Especially in an urban environment, where people might not know their neighbors, this is a useful feature for supporting engagement with an online community that overlaps with the offline geographic community to which one also belongs. It also coincides with Rutigliano’s (2007) definition of community blogs as participatory media that involve some form of public input or assistance in media production, discussed in Chapter 4. With this in mind, I hypothesized the following:

H5: Columbia Heights’ hyperlocal media include more online conversation than WashingtonPost.com.

In this case, conversation was measured by the number of comments or replies recorded for each case over the study. While one could consider all e-mails to list-servs or Facebook wall posts, for example, to be conversation because of the social nature of their media, this study looked for evidence of two-way conversation—messages or content that attracted responses. The results are tabulated below.

Table 5.5: Conversation Through Comments and Replies

Source # Comments/Replies Total # Stories/Posts During Study[11]
Total: 509 189
WashingtonPost.com 7 3

Overall, the data support the hypothesis: In total, there were more replies or comments on hyperlocal media than on WashingtonPost.com. However, this finding does come with a caveat, because Prince of Petworth generated far more conversation than any of the other sources, which all had tallies numbering in the single digits. This can be explained in part by two points. First, comments were counted at the end of the business day on which the original content was published, so these data do not account for replies added on subsequent days, and conversation on hyperlocal media can sometimes take an extended period of time to unfold. Second, there is a vast difference in the web traffic of these media outlets, with WashingtonPost.com and Prince of Petworth generating the highest amount of traffic. Nevertheless, Prince of Petworth still generated significantly more conversation than WashingtonPost.com, and this is without a doubt due to the unique mission of that blog to facilitate conversation among its readers. Such interactivity is part of the nature of hyperlocal media in a way that it is not in the case of mainstream online media, and that observation should not go underappreciated.

Race, Class, and Gentrification

In conducting the content analysis and background research for this study, it became clear that there are certain subjects not tested in the above hypotheses that have often incited debate and discussion via hyperlocal media. Gentrification is a shared concern of many of the long-term residents of Columbia Heights, as well as more recent arrivals to the neighborhood. Andrew Wiseman of New Columbia Heights cited the influence of gentrification and its attendant difficulties of rising real estate costs as his most serious concern about the neighborhood’s future (A. Wiseman, pers. comm.). As mentioned earlier, the Pleasant Plains Civic Association emphasized that issue at its February 2011 meeting. PPCA President Darren Jones agreed that one shared concern of the Columbia Heights community is that “home ownership is becoming unaffordable” (D. Jones, pers. comm.). Often implicit in such discussions but left unsaid are issues related to class and race. Because these dynamics have been at play at various points in the history of Columbia Heights, a closer look at their presence in hyperlocal media content is warranted.

The history of the “columbia_heights” list-serv laid out in Chapter 2 noted that debates about gentrification with racial undertones were a topic of heated discussion as early as 1999, but that the membership of the list-serv had its share of white gentrifiers, as well. Similar patterns are emerging in other neighborhoods today that have more recently begun to experience gentrification. In an article in the Washington City Paper, a resident of LeDroit Park—a neighborhood just east of Columbia Heights that has become home to more white residents at the start of the 2010s—observed about the community’s list-serv, “The list became a reliable source of information about crime in the neighborhood, yet it seemed to be actively used mainly by white residents—though perhaps there were some black lurkers like me” (Hilton 2011). This comes at a time when 2010 census data showed the black population in D.C. falling dramatically to a very slight majority at 50.7%, attributed to 15 years of gentrification preceding the census (Morello and Keating 2011).

Unsurprisingly, race has remained a topic of discussion in Columbia Heights’ hyperlocal media during these years, too, though not always overtly so. A post (Silverman 2011d) on Prince of Petworth in March 2011 featured a message from readers about the status of a restaurant on Georgia Avenue NW in Pleasant Plains, a predominantly black neighborhood.[12] In the reader’s inquiry, she mentioned an occasion when she went to the restaurant with her boyfriend and saw the annotation “White Couple” on their tab (Silverman 2011d). The reader did not seem to express an offended tone when relating this reference to her race, but the commenters launched a 20-message-long thread debating the racism behind this action. The sender of the original e-mail to Prince of Petworth also commented on the post, saying:

I did not want to instigate a race debate. I merely wanted to know what happened to (what looked like) a viable business that my boyfriend and I liked to patronize. Waitstaff was always welcoming to us and we want to support business on GA Ave (black/white/whatever). (Silverman 2011d)

While gentrification has brought new commercial development to 14th Street NW in Columbia Heights as of 2011, parts of Georgia Avenue NW show less revitalization. In wanting to support business on Georgia Avenue, the reader is in fact showing an effort to support local businesses even as the chain stores associated with Columbia Heights’ gentrification continue to open on other streets. In this case, then, what began as a discussion on gentrification became one about race. There are no signs that these debates will disappear, and it is useful to keep in mind the demographic shifts that they belie when thinking about the questions of community engagement and hyperlocal media use explored in this study, as they are often relevant even when they are not explicit.

Analysis of the Medium

In addition to the characteristics presented above about coverage of individual topics, the content analysis also yielded a number of observations about medium-specific attributes. Each medium will be discussed further below with the exception of Twitter, which is the focus of in-depth analysis in Chapter 7. Song (2009) observed that online communities can have highly permeable boundaries separating them from offline communities, and although divided according to medium, the following discussion observes that the boundaries between media are not always discrete.

List-servs

The Columbia Heights list-servs studied here show evidence of three primary uses: routine communication, time- and place-sensitive communication, and discussion or debate of a timely issue. The findings presented in Chapter 5 showed sparse content for the Columbia Heights list-servs included in this study, but these three uses each set their own pace. The first, routine communication, maintains some consistency in the list-serv’s activity. Every month, ANCs or civic or community associations send announcements about their upcoming meetings or circulate meeting minutes on some of the neighborhood list-servs. Similarly, ANC 1B Commissioner Sedrick Muhammad sends daily digests of local “News Headlines” to the “southcolumbiaheights” list-serv. In these messages, he compiles a selection of headlines about metro news and sometimes neighborhood news, drawn heavily from the websites of daily newspapers such as The Washington Post and The Washington Examiner.

The second broad category is time- and place-sensitive communication. For example, a few weeks after the time period covered by this study’s content analysis, on March 21, 2011, a seventeen-message thread unfolded over the space of two days about a recurring problem of certain individuals leaving trash on the street in front of a retailer in South Columbia Heights (Sarper 2011). The message was addressed to Commissioner Muhammad, and the Commissioner responded immediately to say he was going to look into the situation in person. After doing so, he reported his findings back to the list-serv, less than two hours after the original message was sent. Although the original poster could have e-mailed Muhammad directly, by sending his message to the list he opened up the opportunity for others to weigh in on the matter, and in the course of this thread, several other individuals did send more general complaints and comments about the frequency with which this occurs at that particular location. What could have been a routine communication to a public official instead was opened up to a broader discourse among concerned residents. However, it should be noted that after the initial action required by the Commissioner, the conversations unfolded over two days. This gives some indication as to why, in the content analysis, counts of replies and comments for each one-day period included were low. By taking advantage of a highly specific communication medium, members of the Yahoo! Group can bring certain public concerns to light among a targeted audience.

The third use of the hyperlocal list-serv is the one that really stirs the medium to attention: as a space for debate around a specific, timely neighborhood issue. In Chapter 4, I argued in favor of expanding Lawrence and Bennett’s (2000) theory of media effects, which recognizes engagement among publics that respond to “big” stories; instead, I suggested the public responds to stories that align with their particular interests and things that affect them personally. An example of this was a 28-message thread on the columbia_heights list-serv March 14 and 15, 2011, concerning the opening of a new, 24-hour restaurant on 11th Street NW in Columbia Heights. While 11th Street NW has some commercial buildings, it is still a more residential part of the neighborhood than Columbia Heights’ main commercial corridor along 14th Street NW, and many residents of that immediate vicinity voiced their opinions in this discussion. While the initial poster only raised concerns about a potential parking overflow and sought more information about the forthcoming restaurant, some respondents had stronger feelings for and against the restaurant opening. Respondent “whjmela” raises concerns about the process through which a new restaurant should be approved:

The project should be taken seriously, and not just as a popularity contest.  This will have positive and negative impacts especially if it goes 24/7 as others/bars on 11th St. may follow.  This area is still primarily residential.  There should at least be a focused neighborhood/ANC meeting not just on this project but the overall impact of how 11th St is developing.  If residents who are most affected should be consulted.  In terms of  parking, some folk drive to the dog park, so yes parking will be impacted.  Outside of the potential 24 hours, I don’t see why the project should not happen, but it should not be rubber stamped. (“Wendy A.” et al. 2011)

Although a previous poster had mentioned the business owner’s presentation at an ANC meeting the previous week, poster “whjmela” (a frequent participant on the list-serv) either was not aware of that meeting or was asking for further action. This raises the question of how well formal announcements sent to list-servs (such as those for ANC meeting agendas) convey information to that particular audience. However, it is clear that some were aware of the restaurant’s presentation at the ANC meeting mentioned, as poster John David Molesky commented, “I’ve never seen so many first-timers at an ANC meeting patiently sitting through two hours of administrative boredom just to show their support for a 5 minute presentation” (“Wendy A.” et al. 2011). Of course, the content analysis and survey results showed restaurants are a popular topic in terms of both the number of people who seek out this information on hyperlocal media and the amount of coverage restaurants receive there. This large turn-out to which Molesky referred illustrates community engagement on an issue of interest spanned multiple online and offline spaces.

Facebook

While all content was sparse in this study, Facebook groups and pages on Columbia Heights were noticeably stagnant. I counted at least 14 Facebook pages and groups associated with Columbia Heights, D.C. Many of these had no posts on their “walls,” and while some had many members, others had only a handful. Several replicated the same purpose—that of a general interest page for fans of the neighborhood. Others were as specific as a group for mothers in the neighborhood or broad enough to include the neighboring regions of Adams Morgan and Mount Pleasant.[13]

There were a few exceptions. As mentioned earlier, the North Columbia Heights Civic Association uses its Facebook page to publicize new blog posts on its website, a use similar to that of the Twitter users discussed above. In addition, the Columbia Heights Community Marketplace also updated its Facebook page with several posts over the course of the months in which this study took place. Another exception was the Facebook group “Columbia Heights Dog Park (11th&Park).” This was the second Facebook group created around a neighborhood dog park; the first was used to establish the dog park. Between February 15 and March 25, 2011, the group sent five (private) Facebook messages to group members about subjects such as lost gloves, lost dogs, the group’s shared interest in dogs, and a bocce league that one member was starting. There were also 9 wall posts or comments during that period, although only one fell on a day included in the content analysis. One of these posts was from Jeff Zeeman, president of the North Columbia Heights Civic Association, publicizing the community service day the Association was planning and inviting dog park users to take part, in an example of Zeeman’s efforts to reach across online media sources (“Columbia Heights Dog Park [11th&Park]” 2011). In addition to discussing a shared interest of dogs that already unites them in an offline space, members of this Facebook group also use the medium as a means of finding people to participate in other community events.

The reason for the overwhelming Facebook silence might be that the majority of the pages have no specific purpose. As the examples above show, these neighborhood Facebook groups and pages appear to be most useful when employed in support of offline activities or shared interests. Like Twitter, Facebook is utilized in Columbia Heights mostly as a space that overlaps with other media or forms of community engagement.

Blogs

The hyperlocal blogs in this study combine multiple functions of the other media discussed above. Prince of Petworth, New Columbia Heights, and The Heights Life all incorporate aspects of discussion, publicity, a space for socializing, and supplementation of other media. What sets them apart is that unlike the news aggregators on TBD and Outside.In and the unstructured lists of subscribers to the list-servs, these blogs all benefit from the editorial hand of one or two bloggers who source information, set the agenda, guide public opinion, and, to a certain extent, moderate conversation. These characteristics allow blogs to stir up conversation, facilitate information-sharing, and cover certain news differently from other media.

The findings in this chapter showed a tighter relationship in the “Follower” index between respondents who follow varied media than in the “Contributors” index, which combined different means of participating in hyperlocal media. However, it was the act of contributing to—not merely following—media that was correlated with offline community engagement, so it is important to understand the weaker relationship among variables in the “Contributors” index. By looking at the elements that make up the “Contributors” index, we can see that this is because the behavior that most survey respondents practice is commenting on blogs: 46% of respondents had posted a comment on one of the hyperlocal blogs or news sites named in the survey. By contrast, only 4% had posted a comment to the “wall” of the one of the Facebook pages named and 17% had e-mailed one of the list-servs. In the survey’s question on uses and gratifications of hyperlocal media (Q. 21), only 14% of respondents use hyperlocal media to connect with others, but 44% use hyperlocal media as a source of information to give them something to talk about with others. This suggests that even those who do not socialize online through regular conversation still use hyperlocal blogs as a means of anticipatory socialization to prepare them for offline social encounters.

Bloggers recognize this role in furthering discussion. The site in this study with the most commenting by far was Prince of Petworth, as the content analysis in Chapter 6 revealed. Prince of Petworth blogger Dan Silverman (often referred to by the abbreviation for his blog, PoP) explained that he views enabling conversation as his primary role on the blog:

I connect people. I help people feel proud of where they live. I help people deal with situations that are unique to their neighborhods. But it’s not necessarily me who’s giving advice, who’s helping these people. It’s in my role as a facilitator that I’m able to help people. … I sort of establish this trust between people. (D. Silverman, pers. comm.)

Silverman is the only one to focus on building social capital through online engagement so explicitly, but it is the one attribute that most sets his blog apart from others in the city. His blog includes two weekly “Random Reader Rant and/or Revel” posts, open threads Silverman starts each Monday and Friday in which commenters can discuss anything they like.  On one open thread during the content analysis, commenters shared a link to a humorous article; thanked PoP for some of the pictures recently posted; discussed playing live musical gigs in the neighborhood; and gave feedback on a local restaurant, among other topics both related and unrelated to Washington (Silverman 2011c).

Coverage of certain events is also handled differently on hyperlocal blogs. Rather than become merely an online virtual bulletin board, for example, hyperlocal blogs are often in a position to highlight certain events in advance. On Prince of Petworth, events are sometimes posted exactly as they were sent to Silverman by a reader, such as an e-mail sent during the content analysis time frame by a reader whose organization was sponsoring a charitable happy hour that evening (Silverman 2011a). New Columbia Heights sometimes spotlights events already published elsewhere. For example, a post during the period of the content analysis linked to a DCist blog post about National Pancake Day at IHOP. Because there is an IHOP restaurant in Columbia Heights, New Columbia Heights published a brief post pointing readers to the previous announcement of the event. Later, New Columbia Heights published another post that included more information about the event from a press release that tied the free pancake promotion on National Pancake Day to a partnership between the restaurant and a charity (Wiseman 2011a, 2011b). Certainly, these are not significant events at the city-level, but for residents who happen to live near that restaurant, it is something that might be of interest. On other occasions, these blogs will include event recommendations that are not time-sensitive: for example, a post on New Columbia Heights that offered a suggested route for hiking in nearby Rock Creek Park (Wiseman 2011c).

Additionally, issues of crime and public safety can turn hyperlocal blogs into a virtual neighborhood watch. This is one way in which contributors to hyperlocal media can take advantage of both the blog medium’s ability to highlight and emphasize important information and the influence of an individual blogger in the community in spreading their message. Although crime news is circulated via the MPD-3d list-serv, not everyone will seek out crime information from such a specific source, but 81% of survey respondents indicated they did turn to hyperlocal media for information on crime, as Chapter 6 found. Therefore, posts from bloggers on crime information have the ability to spotlight news that is out of the ordinary—news that could risk being overlooked when bundled with the daily crime reports in a digest of messages sent to the MPD-3d list-serv. An example from the Prince of Petworth shows that readers recognize this role of blogs, and that for some, hyperlocal blogs may be a primary source of hyperlocal news. A reader e-mailed Prince of Petworth a message about a “peeping Tom” in Columbia Heights that began and concluded as follows:

Dear PoP,

I’m not sure who else to report this to (other than the cops) so I thought I should tell you so that the wider neighborhood would know. …

I wanted to report it so that if it has happened to anyone else we can know it’s a trend and also to keep people aware/solicit advice about security for basement apartments. (Silverman 2011e).

The second half of the same blog post included another e-mail from a reader about a man in Columbia Heights who has tried to follow her home on multiple occasions. This reader concluded her message, “Just thought that this may be worth posting to the blog as a kind of head’s up, etc.” (ibid.). Whether Prince of Petworth receives these sorts of messages from readers who do not use other hyperlocal media or those who feel the blog is the best way to share the information with the neighborhood, they constitute an online alternative to the neighborhood block watches and public forums espoused by some of the community associations discussed above. This kind of communication has the added potential of reaching people who do not live in the neighborhood but do pass through it.

Despite the advantages of the blog medium, there is also a role for overlap among media. Discussions that migrate from list-servs to blogs, and vice versa, are a useful demonstration of this. As mentioned earlier, it is easy to overlook information sent out over the list-servs. Even Dan Silverman from Prince of Petworth admitted that he often skips reading some of the list-servs to which he subscribes (D. Silverman, pers. comm.). Furthermore, for those new to the list-serv or the offline community, it can be difficult to discern potential biases presented by those who contribute. For example, a new follower may not know that one particular individual has a reputation for incendiary posts or that another was involved in a past controversy related to the subject about which he wrote. Neighborhood politics are nuanced, and in online communication, a person’s tone can often be ambiguous. All of these factors can create a need for an editorial voice to provide an overview of the nature of the conversation unfolding.

When hyperlocal bloggers cross the fluid boundaries of their medium to highlight information from the list-serv, it can signify the importance of a topic, or it can exemplify a kind of political coverage not captured through the content analysis in this study: the power struggles that determine neighborhood politics. One example of this can be seen in a post from New Columbia Heights that republished a graphic sent to the “columbia_heights” list-serv by William Jordan, a former ANC 1A commissioner and frequent list-serv contributor (a.k.a. “whjmela,” mentioned in an above example; Wiseman 2011d). The graphic was Jordan’s explanation of connections between Councilman Jim Graham and a local developer building in Columbia Heights. The point of the New Columbia Heights post was partly to call attention to the controversy surrounding the developer’s recent actions, as well as to clarify Jordan’s claims. Blogger Andrew Wiseman e-mailed Jordan for clarification and posted his reply (which promised to clarify the points of confusion at a later date), and the commenters on the post responded largely by expressing skepticism towards Jordan’s claim. One post signed “W Jordan” was presumably from Jordan himself and defends the purpose behind his diagram. This example demonstrates the inevitable crossover between different hyperlocal media in Columbia Heights as well as the ability of the blogger to frame communications from other media within a particular context.

As the findings above show, there are great differences in scale among the hyperlocal media sources counted in this study, but each one contributes in its own way to the broader media environment. Specialization allows certain media to excel in certain roles: list-servs and blogs as sites of information-sharing and discussion, for example, and Facebook and Twitter as supplementary channels to other media or spaces for communicating with very specific audiences. Each of the hypotheses found support for the claims that hyperlocal media cover more information on neighborhood politics, crime, events, and real estate than WashingtonPost.com, in addition to featuring more frequent conversation among users of hyperlocal media. The low number of total stories (n=192) published over the course of the three-day content analysis emphasizes that hyperlocal media are not yet a replacement for other sources, but they do possess unique strengths, nonetheless. In the next chapter, we will see whether the impact of hyperlocal media use on community engagement shows evidence that quality, not quantity, is what really matters in this context.


[1] The survey methodology is discussed in detail in Chapter 6.

[2] URLs for each site can be found in Appendix A.

[3] These boundaries are 16th Street NW to the west, Spring Road NW to the north, Georgia Avenue NW to the east, and Florida Avenue NW/Barry Place NW to the south. If content referred to something outside these boundaries as occurring within Columbia Heights, I did not count it as a story about Columbia Heights for the sake of uniformity.

[4] See Methodological Appendix A for survey questions.

[5] For all sources tracked in the content analysis, total n=192.

[6] WashingtonPost.com data are included in the calculations for “coverage” because survey respondents were given that site as a source to name among their news sources before they indicated their hyperlocal media gratifications. Percentages do not sum to 100 for either column because coverage included an “other” category, and survey respondents were allowed to name multiple uses and gratifications indicating the demand calculation.

[7] Only stories about Columbia Heights were coded for WashingtonPost.com, whereas other sites’ tallies of the total number of stories also included those not on Columbia Heights. Therefore, a percentage of total coverage for the WashingtonPost.com sample would not be comparable with that listed for the other cases. To provide some context for comparison, though, in a content analysis conducted during the research for Gloria and Hadge (2010), The Washington Post print edition included an average of 141 stories per day.

[8] See http://www.washingtonpost.com/local/crime, accessed 21 March 2011.

[9] A positive correlation between living in Columbia Heights and following its hyperlocal media is supported in Chapter 6.

[10] See all entries tagged in these categories at http://www.princeofpetworth.com/category/good-deal-or-not/ and http://www.princeofpetworth.com/category/gdon-revisited/, accessed 2 April 2011.

[11] For Twitter and list-serv sources, the comments/replies are also counted in the total number of posts because of the coding scheme designated above in this chapter.

[12] See data for tract 35 in Appendix B, Table B.1.

[13] See methodological appendix for complete list.

Be Sociable, Share!