Chapter 6

Community Engagement and Hyperlocal Media Use: Survey Methodology and Findings

As earlier chapters have shown, hyperlocal media are still evolving. Regardless of the specific medium used, online hyperlocal communication, though not an entirely nascent concept, continues to be marked by experimentation. Communities are finding ways to adapt hyperlocal media to meet their needs, and hyperlocal media are presenting new possibilities for communicating information to different groups of people. Just as this study takes a broad view in defining community engagement, so too does it employ a comprehensive understanding of what it means to engage with media, as well as the potential for intersection between online and offline modes of participation. However, not everyone uses online media, nor can those media fill all of a community’s information and communication needs. Thus, there is still a role for offline communication spaces and events. This chapter will begin with a discussion of the roles civic and community associations and Advisory Neighborhood Commissions (ANCs) play in Columbia Heights, before turning to the survey methodology employed in this study and an analysis of its findings.

The Role of Community Meetings

Civic and community association and ANC meetings represent only a fraction of the ways Columbia Heights residents get involved in their community. However, they share many things in common with the goals of hyperlocal media. Different types of community meetings determine their own unique role in terms of the scope of the subject matter they cover and the audience they attract. This diversity of roles helps to explain why community engagement in the survey data was not strictly correlated with many of the usual demographic characteristics, as the findings in this chapter will show. Instead, Columbia Heights residents can find ways to get involved that are tailored to their own interests.

This finding should be understood in the context of the history of community and civic associations in Washington, D.C. As Chapter 2 mentioned, the first citizens’ association in Columbia Heights, founded in 1904, was for white residents only; nearly two decades later, the Pleasant Plains Civic Association was founded for the African-American population that settled in that part of Columbia Heights. Even at the meeting I attended, the Pleasant Plains Civic Association maintained a majority of African-American members. There are still parts of Columbia Heights that have racial majorities, so it is difficult to untangle the influence of race from that of the sub-neighborhood itself. It is likely that both of these factors play a part in the concerns that garner attention from the various civic and community associations. While this study has so far emphasized the unique roles of hyperlocal media, community and civic associations also embody their own roles within the context of Columbia Heights and its engaged public, each varying somewhat based on the characteristics of its leader and the concerns of the blocks to which it caters.

Pleasant Plains Civic Association

Observations of meetings of several community and civic associations show distinct differences in the role played by each one in its immediate neighborhood. At a meeting of the Pleasant Plains Civic Association (PPCA) in February 2011, there was a clear emphasis on preserving a way of life that was long-held in that area, which stretches between 11th Street NW and Georgia Avenue NW in the eastern part of Columbia Heights. Three of the 11 presentations on the meeting agenda addressed plans for the development of affordable housing in the neighborhood, the economic benefits of home ownership over renting, and the effects of neighborhood gentrification. While this information catered to those who were long-time residents of the neighborhood, a presentation from a local lawyer on estate-planning services also addressed the advanced age of many of the attendees. Additionally, in recognition of Black History Month, a staff member from Ward 1 Councilmember Jim Graham’s office gave a rendition of Martin Luther King, Jr.’s “I Have a Dream” speech while the audience sang “We Shall Overcome” in the background, also a tradition for the Civic Association.

This approach, blending shared culture with more tangible needs of community, informs the PPCA President’s perspective on the Association’s role. While President Darren Jones describes participants in the Civic Association as an “active minority” in the community, he also said that the Civic Association exists for “getting information to the neighborhood that doesn’t get through through normal channels, and then I also see us as advocating for the people who live here” (D. Jones, pers. comm.). The current president has been involved in the Civic Association since 1983 or 1984 and has held leadership positions since the early 1990s. Thus, it is not surprising that this would be a neighborhood where the concerns of long-term citizens are a priority among the engaged public. Although Jones uses list-servs to get out information about upcoming meetings and disseminate meeting minutes in the hopes of attracting new attendees, he acknowledges  that they “get some of the information out, but there are so many who aren’t online” (D. Jones, pers. comm.). He perceives this digital divide as a consequence of advanced age or lack of interest in going online, not the financial cost of Internet access.[1] Consequently, he still distributes fliers door-to-door to 60 to 70 recent attendees or asks meeting attendees to bring fliers back to their neighbors.

Northwest Columbia Heights Community Association

While the potentially negative effects of gentrification are a specific concern for the PPCA, the Northwest Columbia Heights Community Association (NWCHCA) primarily focuses its attention on crime prevention and other public safety efforts. The president of the NWCHCA, Cecilia Jones, is also chair of the ANC 1A Public Safety Committee (ANC 1A 2011). On February 15, 2011, I attended the NWCHCA’s Criminal Justice Community Public Forum, which consisted of a panel discussion among neighbors involved in neighborhood block watches as well as public officials and professionals working in the criminal justice system, followed by small group discussions. In contrast to the PPCA meetings, which emphasized informing the community, the public forum encouraged discussion with an eye towards getting involved in crime prevention outside the forum, as well (e.g., through a neighborhood block watch program).

The dialogue during the meeting illustrated the role of face-to-face communication in fostering community engagement. Tom Cochran, a panelist and member of the Newton Street Block Watch, noted, “The best way to connect with your neighbors is in person.” Fellow panelist and Newton Street Block Watch member Josue Salmeron had a similar view: “For me personally, e-mail or text messages work great, but not all our neighbors [use those technologies]…We need to get back to one-on-one” conversation. Ward 5 Court Watch Leader and panelist Kathy Henderson tied crime prevention to a sense of community, observing, “[In] neighborhoods where crime is at a minimum, their neighbors are cohesive.”[2]

The NWCHCA uses a variety of hyperlocal media to get its message out. Its Yahoo! Group-powered list-serv was included in the content analysis, although Cecilia Jones also e-mails a list-serv of users who have signed up to receive messages via the NWCHCA website,[3] which serves as a resource for crime prevention information. Additionally, there is YouTube channel for the NWCHCA that features of a video with highlights from an April 2010 public forum titled “Youth and Our Community.”[4]

ANCs 1A and 1B

In contrast to the community and civic associations, ANCs exist less to deal with specific issues of concern to the community than to handle all of a neighborhood’s public business. That is not to say that single member districts (SMDs), a subcategory within ANCs represented by each ANC commissioner, cannot handle more specific concerns. In the South Columbia Heights SMD of ANC 1B03, Commissioner Sedrick Muhammad holds a monthly meeting with members of the Metropolitan Police Department from the third district to address residents’ concerns. Because this part of Columbia Heights has crime rates that are higher than the city’s average (as well as higher than those in some other parts of the neighborhood), these meetings address a concern specific to Muhammad’s SMD that would not necessarily be as prevalent at the monthly ANC 1B meeting.[5] Muhammad also started the “southcolumbiaheights” Yahoo! Group, which has, as mentioned earlier, over 400 subscribers (S. Muhammad, pers. comm.), but while the Yahoo! Group is active, the February 2011 ANC 1B03 meeting attracted only 8 attendees (excluding myself), only three of whom were attending as members of the public, rather than in a professional capacity. Of those three, only two were Columbia Heights residents (both lived in the apartment building where the meeting was held). In this case, it appears offline and online participation in South Columbia Heights have different uses, and issue-specific meetings in person may find less demand than in other parts of Columbia Heights.[6]

At the monthly ANC 1A and 1B meetings, though, it was not uncommon for people to come for a specific purpose and leave when their area of interest had been covered. This kind of selective participation corresponds to the structure of these meetings, in which official business must be conducted related to local businesses, real estate development, and other issues that are brought before the ANC requiring some kind of public response or approval. Perhaps because attendance at ANC meetings can be compulsory for many who come with official business before the city, the nature of community engagement through an ANC is somewhat different. However, special-interest committees allow members of the public to participate in greater depth, and ANCs do publicize their meeting agendas and minutes on their websites and via a number of list-servs. At their most basic, though, ANC meetings are a way of communicating the business of the neighborhood directly, without editorializing.

ANC 1B Commissioner Tony Norman said he relies primarily on word of mouth to stay informed about news in Columbia Heights, and while he follows several list-servs, he rarely uses them to get information out. He does read some hyperlocal media, such as Prince of Petworth, but says, “I take those with a grain of salt, just like when I watch The [Washington] Post. They only communicate a certain perspective and a certain group’s perspective…In fact I don’t even use [media] as a major source of information about what’s going on. But I know enough people in the neighborhood, in this area, to contact various different people that I know that live in the neighborhood [for local information]” (T. Norman, pers. comm.). For those who have the interest to attend, ANC meetings can fill the role of seeing the city’s governance processes at work before them; for everyone else, published meeting minutes can fill some of the information gaps left by media that do not cover such granular issues.

North Columbia Heights Civic Association

In contrast to concerns addressed by leaders in the associations previously discussed, the North Columbia Heights Civic Association (NCHCA) embraces new media unreservedly. The NCHCA’s website includes a blog with community announcements and the occasional video. As evidenced in the content analysis, it does not regularly update its content, but when it does, updates are linked to the NCHCA Facebook page.  Jeff Zeeman, president of the NCHCA for the past three and a half years, said he relies upon online communication—through the blog, Facebook page, and the NCHCA list-serv—because of its ease of use. Zeeman is aware solely online communication will leave some out and acknowledged, “I think there’s going to be a disproportionate number of people who know of things happening that happen to be online than those who are not” (J. Zeeman, pers. comm.). While NCHCA’s varied use of online media has the potential to reach users with different new media use habits, the comments presented by other interviewees above suggest it may be premature to ignore members of community who are not online.

NCHCA meetings typically attract newcomers to the neighborhood, in addition to a few veterans, and the Association’s president suggested that online outreach may increase NCHCA’s visibility to such populations. Compared with the leaders of some of the other civic associations discussed here, Zeeman himself is still a relative newcomer, and this may give him a different perspective on which media are most effective for reaching the public. Zeeman also noted that the issues of interest to the NCHCA are those that tend to catch the attention of new residents, such as the Columbia Heights Community Marketplace in Tivoli Plaza and a day-long community service event planned for the spring of 2011 (J. Zeeman, pers. comm.). The concerns of long-term residents, including crime and education, are beyond the scope of the issues the NCHCA sees as within its reach (ibid.). However, the distinct variations among civic and community associations reconcile the interests of veteran residents looking to maintain their way of life in the neighborhood and new arrivals trying to find theirs. As the following sections will show, this leads to a model for community engagement that is based not on demographic characteristics, but on other factors.

Methodology

This chapter will undertake the following research question:

RQ2: Is there a correlation between community engagement and hyperlocal digital media use in Columbia Heights?

The methodology used to explore this question involved surveying members of the Columbia Heights community, in online and offline public spaces. I developed a 28-item questionnaire, which asked respondents about their media consumption and online participation habits, their attendance at and involvement in events and organizations in the community, and their political attitudes and participation.[7]

The survey questions were developed with several things in mind. First, as established in Chapter 4, this study takes a broad view of what it means to participate in the public life of the community. Thus, the survey asked participants how often they vote in local elections, but it also elicited responses about their involvement in other aspects of community life, such as attendance at Advisory Neighborhood Commission (ANC) meetings and membership in civic associations and other religious, social, professional, and cultural groups in the neighborhood. Questions about the frequency of participation in such groups and whether or not respondents held leadership roles in the community measured both the quantity and quality of their participation. Second, the survey counted those who follow or contribute to online hyperlocal sources across media. The survey named specific hyperlocal media sources in Columbia Heights and also asked respondents to name others they use to seek out information on their neighborhood or communicate with their neighbors. These media spanned list-servs, Facebook groups and pages, Twitter feeds, blogs, and news or information-oriented websites. Two questions also asked about the uses and gratifications survey respondents sought from hyperlocal media; these findings are discussed in detail in Chapter 5.

In order to measure community engagement in offline contexts, a crucial part of my data collection process involved distributing surveys in person, in order to reach those who are engaged in the community but do not use hyperlocal media. To do so, I focused on two events that combine political and non-political sides of public life in Columbia Heights: ANC meetings and civic association meetings. In February 2011, I attended meetings of ANC 1A and ANC 1B, which between them cover all of Columbia Heights; ANC 1B03, a single member district (a smaller subset of the ANC) in South Columbia Heights within ANC 1B; the Pleasant Plains Civic Association; and the Northwest Columbia Heights Civic Association.[8] These meetings discussed real estate development and housing concerns, local businesses, local schools, community service initiatives, and new projects to enhance the community, such as the opening of a North Columbia Heights Community Garden and the return of the Columbia Heights Community Marketplace. Prior to the start of the meetings, I introduced myself to individuals as they were taking their seats, described my research as an examination of local media and community life in Columbia Heights, and asked them if they would be interested in taking my survey. At four of the meetings, I also had the opportunity to make an announcement about my survey and distribute copies to others who had entered later.

In addition to distributing paper copies of the survey, I also posted it through various online media. The survey was developed using SurveyMonkey, for privacy concerns and ease of analyzing data, which can be downloaded directly to SPSS software. To publicize the survey, I sent it with a note about my research to all of the Columbia Heights list-servs of which I was aware (i.e., those included in the content analysis in Chapter 6); posted it to the wall of the Facebook pages and groups listed in Methodological Appendix A.III, as well as to my own Facebook account; posted it to my own Twitter account; sent it to the list-serv of my graduate program at Georgetown University, which includes residents of Columbia Heights; e-mailed a small group of personal acquaintances in the neighborhood; and e-mailed it to the three blogs included in this study. My hope was that by using my own online social networks as well as those specific to Columbia Heights and this study, I would be able to take advantage of a snowball effect in which others passed the survey along to residents not reached through other means. However, I was not able to track this outcome except through anecdotal evidence from my own acquaintances. Four unique collector links were created in SurveyMonkey: one sent to the Columbia Heights list-servs; one sent to bloggers; one posted on Facebook; and one posted via Twitter and sent to my personal networks. The blogger behind New Columbia Heights agreed to post a link to the survey, after which point the number of online replies I had collected increased from 76 responses to an eventual n=351. However, the link he posted was the one I had sent to the list-servs, so it was not possible to determine discretely the number of respondents who came upon the survey through New Columbia Heights.

Characteristics of Survey Respondents

Before we turn to the analysis of the data, it is useful to keep in mind some of the demographic characteristics of the survey respondents, as many of these attributes will be tested as independent variables in the hypotheses presented in this chapter. Figure 6.1 provides the most relevant information, and frequencies showing more detailed information can be found in Methodological Appendix A.

Figure 6.1: Demographics of Survey Respondents (n=407)

Live in Columbia Heights 78.9%
Average years lived in Columbia Heights 4.2 years
Average age 41 years old
Highest level of education completed (mode) Bachelor’s degree (48.3%)
Type of Internet access (mode) High-speed broadband (45%)
Respondents without Internet access at home 2.7%

Demographic information on race is presented separately in Table 6.1, in a comparison of the sample and the Columbia Heights population.

Table 6.1: Racial Demographics: Survey Sample vs. Columbia Heights Population

Sample Columbia Heights[9]
Other/Did not respond 25.8% N/A
White – non-Hispanic 58.2% 26.5%
White – Hispanic/Latino 3.9% 27.7%
Asian/Pacific Islander 3.2% 4.3%
Black/African-American 8.8% 40.6%

Interviews

To complement the data gathered at public meetings, I also conducted a limited number of targeted, semi-structured interviews in the hopes of gathering information that would shed light on the findings from my empirical data. The interviewees included an ANC commissioner, two civic association presidents, and a blogger.[10] The interviews were designed to elicit information on the way these individuals use new media to do their work and the role of hyperlocal media in the community. Insights acquired through these interviews will be discussed in Chapter 8.[11]

Preparing the Data

Data collected through the online survey were downloaded to SPSS, and data collected through the paper surveys were added by hand to create one data set for both data collection methods. Because questions were exactly the same in each survey, the two methods of data collection will not be differentiated. The two methods of data collection were only employed to ensure a sample that reached more of the population; responses to the survey questions themselves show differences in community engagement online and offline. The data were prepared to reflect missing data and other inconsistencies.[12]

Hypotheses

In Chapter 4, I laid out a theoretical framework for examining the dynamics between the concepts that are tested in this study, including their applications to this specific situation. Within the context of the literature, we can understand hyperlocal media to refer to online media with a narrow geographic focus that provide the public an opportunity to participate or interact. In this case study, hyperlocal media cover one neighborhood. Because one of the attributes that sets hyperlocal media apart from other traditional media is its high level of interactivity, this study also differentiates between users who follow or passively consume hyperlocal media and those who contribute to or actively interact with these media. Community engagement in this study is considered from a broad perspective as behavior that encompasses many ways of participating in the public life of the community, whether political, social, cultural, religious, or professional in orientation. In analyzing the data, I brought these concepts together to test the following hypotheses:

H6: Contributors to hyperlocal media are more likely to exhibit a high degree of community engagement in Columbia Heights than those who do not contribute.

H0: Contributors to hyperlocal media are no more likely to exhibit a high degree of community engagement in Columbia Heights than those who do not contribute.

H7: People who live in Columbia Heights follow that neighborhood’s hyperlocal media more closely than people who live in other parts of the city.

H0: People who live in Columbia Heights are no more likely to follow that neighborhood’s hyperlocal media closely than people who live in other parts of the city.

H8: People who are long-term residents of Columbia Heights are more likely to exhibit a high degree of community engagement than are people who have not yet lived in Columbia Heights on a long-term basis.

H0: People who are long-term residents of Columbia Heights are no more likely to exhibit a high degree of community engagement than are people who have not yet lived in Columbia Heights on a long-term basis.

H9: Increased use of hyperlocal media is positively correlated with increased community engagement.

H0: Increased use of hyperlocal media is no more likely to be correlated with increased community engagement than is low use of hyperlocal media.

H7 and H8 test the idea that hyperlocal media target a geographically-defined audience and set of users, and that people become most involved in the public life of their immediate neighborhoods. H6 suggests that those who are active users of hyperlocal media exhibit a higher degree of community engagement than those who follow such media passively or not at all, while H9 proposes a positive relationship overall between community engagement and hyperlocal digital media use.

These hypotheses presuppose certain definitions for the terms used to explain the relationships, described as follows:[13]

  • Contributors: Contributors to hyperlocal media are those who indicate they have e-mailed a list-serv, posted a comment on a blog or website, or posted to the wall of a Facebook page or group related to Columbia Heights.
  • Followers: People who follow Columbia Heights’ hyperlocal media are those who read, subscribe to, or, in the case of Twitter and Facebook, “follow” and “like” the media sources included in this study.
  • Hyperlocal media users: This refers to the combined group of contributors and followers of Columbia Heights’ hyperlocal media, as indicated by respondents.
  • Community engagement: For the purposes of this study, community engagement includes attendance at ANC and civic association meetings, including frequency of attendance; participation in social, religious, political, cultural, volunteer, and professional organizations in Columbia Heights; and level of involvement in Columbia Heights as indicated by the respondent. Less engaged respondents tallied up to 10 points for each behavior or level of involvement they indicated; more engaged respondents had scores of 11 to 18 on the same scale.
  • Long-term residents and short-term residents: Based on the frequencies of survey responses, long-term residents are defined here as those who have lived in Columbia Columbia Heights for more than 10 years.[14]

Operations

These variables were constructed as scales using the responses to multiple survey questions. The contributors, followers, and hyperlocal media users variables were constructed by adding similar behaviors across different media. The “community engagement” variable combined different behaviors that are related on a theoretical basis; the index was statistically reliable. Initial attempts at constructing this index included a variable that measured how frequently respondents voted in local elections, but this lowered the reliability of the index to an insignificant level; thus, voting was discarded. However, as Chapter 4 pointed out, community engagement need not be restricted only to the acts of political participation that have previously been the focus of political communication scholarship, and the poor fit of voting in this index suggests that the behaviors that were included are in fact distinct from other forms of political participation. Figure 6.2 shows in more detail the survey questions from which data were drawn to create each of the indices used in this chapter’s hypotheses.

Figure 6.2: Index construction[15]

Contributors Q15 + Q17 + Q19

(Cronbach’s α = .449)

Followers Q12 + Q14 + Q16 + Q18 + Q20

(Cronbach’s α = .83)

Hyperlocal media users Q12 + Q14 + Q15 + Q16 + Q17 + Q18 + Q19 + Q20

(Cronbach’s α = .653)

Community engagement Q4 + Q5 + Q8 + Q10

(Cronbach’s α = .623)

Findings

The first hypothesis tested considers the character of hyperlocal media use with respect to community engagement:

H6: Contributors to hyperlocal media are more likely to exhibit characteristics of community engagement in Columbia Heights than those who do not contribute.

The findings show that we can reject the null hypothesis and support H6. Respondents who are more engaged in the community actively contribute to more hyperlocal media sources than those who are less engaged. Specifically, among those who are more engaged, significantly more people contribute to two or three hyperlocal media than do those among the less engaged (n=145; Chi square, p=.003). The less engaged are more likely to contribute to no media sources, but they are also slightly more likely to contribute to just one hyperlocal source than the more engaged respondents. In summary, those who are less engaged in the community contribute to fewer hyperlocal media sources than do those who are more engaged in the community.

Table 6.2: Community engagement and contribution to hyperlocal media

Contribute to 0 media Contribute to 1 medium Contribute to 2 media Contribute to 3 media
Less engaged 43.6% 41.0% 14.1% 1.3%
More engaged 22.4% 38.8% 28.4% 10.4%

While we can understand contributing to any kind of medium to be related at a theoretical level—here, the distinction is being made between active participation and passive consumption of media—a reliability analysis of the “Contributors” index does not confirm this empirically. The value of Cronbach’s alpha for the “Contributors” index is 0.449 (n=323), a weak relationship at best. There is a connection to be made here to knowledge-based development, discussed in Chapter 4. We can think of those who contribute to hyperlocal media as participating in knowledge production in the community. If these people are also engaged offline, there exists an opportunity for that knowledge production to inform knowledge-based development, particularly when the offline activities in which hyperlocal media contributors engage are ones oriented towards urban development, politics, or other areas with implications for local policies.

While more active participation in Columbia Heights corresponds with active participation in its hyperlocal media, the more passive activity of simply following the neighborhood’s hyperlocal media also corresponds with simply living in the neighborhood, whether an active member of the community or not. This correlation was tested in the following hypothesis:

H7: People who live in Columbia Heights follow that neighborhood’s hyperlocal media more closely than people who live in other parts of the city.

The data show support for H7 by disproving the null hypothesis that there is no difference in following Columbia Heights’ media between those who live in the neighborhood and those who do not (n=383; Chi-square, p=.023). Instead, the findings show that, regardless of the number of sources one follows, following hyperlocal media about Columbia Heights is more common among those who live in the neighborhood. Within the “Followers” index, there is also a stronger relationship among following various kinds of media than existed within the “Contributors” index. The reliability of this index was the highest of the four created in this chapter, with a Cronbach’s alpha value measuring 0.83 (n=407). This strong relationship suggests that respondents are less discriminating in the media they follow than in the media to which they contribute. Table 6.3 shows the percentages of hyperlocal media followers who live in Columbia Heights and in other areas. Some of the media included, such as Prince of Petworth and Ward 1 Councilmember Jim Graham’s Twitter feed (@JimGraham_Ward1), cover neighborhoods outside Columbia Heights, so this may explain these findings somewhat.

Table 6.3: Following Columbia Heights’ (C.H.) hyperlocal media and residence in the neighborhood

Live in Columbia Heights Live elsewhere
Follow less than 5 hyperlocal media on C.H. 78.5% 21.5%
Follow 5 to 10 hyperlocal media on C.H. 82.4% 17.6%
Follow more than 10 hyperlocal media on C.H. 86.1% 13.9%

Length of residence in Columbia Heights also proved to be a significant factor in this study, as the following hypothesis tested:

H8: People who are long-term residents of Columbia Heights are more likely to exhibit a high degree of community engagement than are people who have not yet lived in Columbia Heights on a long-term basis.

The findings support H8, which proposed that long-term residents of Columbia Heights (i.e., those living there for 10 years or more) are more likely to show community engagement than more recent transplants (n=125; Chi square=.011). As shown in the table below, 75% of those who have lived in the neighborhood for 10 years or more fall into the “more engaged” category:

Table 6.4 Level of community engagement according to length of residence in Columbia Heights

Less engaged More engaged
Residents for 1 year or less 67.9% 32.1%
Residents for 1 to 3 years 61.3% 38.7%
Residents for 3 to 10 years 65.2% 35.8%
Residents for 10 years or more 25% 75%

Given the influx of newcomers to Columbia Heights in the past decade, this is an important consideration. Other tests performed did not show age to be an important factor; in fact, in an ordinary least squares regression on community engagement, the independent variable of age was negatively correlated. Years from now, as Columbia Heights’ present residents age, it may be the case that these patterns are different. For now, though, the negative correlation of age with community engagement is an interesting contrast to Putnam’s (2000) claim that the older generations are the only ones who still participate in community associations.

Finally, all of the previous hypotheses lead us to the ultimate question: After taking into account the effects of length of residence in Columbia Heights and level of active participation in hyperlocal media, is there a positive relationship between hyperlocal media use and community engagement in Columbia Heights? The study’s findings allow us to reject the null hypothesis in support of H9, which predicted the following:

H9: Increased use of hyperlocal media is positively correlated with increased community engagement.

A contingency analysis of all hyperlocal media use (both following and contributing) and community engagement shows that at higher levels of hyperlocal media use, higher levels of community engagement also exist. This relationship is statistically significant (n=145; Chi-square p=.036) and can be observed in the table below.

Table 6.5: Community engagement and hyperlocal media use in Columbia Heights

Less engaged More engaged
Do not use hyperlocal media 42.9% 57.1%
Follow or contribute to 1 hyperlocal media source 65.4% 34.6%
Follow or contribute to 2 hyperlocal media sources 70.4% 29.6%
Follow or contribute to 3 hyperlocal media sources 60% 40%
Follow or contribute to 4 hyperlocal media sources 36.7% 63.3%
Follow or contribute to 5 hyperlocal media sources 30% 70%
Follow or contribute to 6 hyperlocal media sources 20% 80%

Thus, while those who do not use hyperlocal media tend to include a slightly more engaged population, as hyperlocal media use increases, so does the likelihood that users show higher community engagement, with the turning point occurring in the middle of the scale of hyperlocal media use. Nonetheless, this is not necessarily a predictive, directional relationship.

To look for a more specific relationship between hyperlocal media use and community engagement, I performed several ordinary least squares (OLS) regression analyses. The first analyzed community engagement and hyperlocal media use to generate the following model:

Community Engagement = 0.223 Hyperlocal Media Users 

While the model shows no heteroskedasticity,[16] it is not necessarily the best fit (R Square=.050, Adjusted R Square=.043). However, it is statistically significant (p=.007). Models that incorporated race, age, Internet access, and education did not produce statistically significant relationships, either. From a methodological perspective, the absence of significant relationships among these variables can be explained in part by the small number of cases for some of the categories. Still, it is telling to look at the variables that were statistically significant in the OLS regression model. The best-fitting model (R Square = .103, Adjusted R Square = .088; ANOVA, p = .002) resulted in the following:

Community Engagement = .220 Hyperlocal Media Users + .209 Years in Columbia Heights

This second model shows hyperlocal media use is the strongest predictor of community engagement among study respondents, but it is closely followed by length of residence in Columbia Heights. A cross-tabulation of length of residence in Columbia Heights and hyperlocal media use also found that the longer respondents had lived in Columbia Heights, the more likely they were to have a higher level of hyperlocal media use (as seen in Methodological Appendix A.V). This reinforces several points discussed earlier in Chapter 4. First, because hyperlocal media use here includes both consuming and contributing to media, offline engagement is predicated on being informed about the community and being willing and able to participate in public discourse. Second, if we take longer residence in Columbia Heights as a sign of attachment to the community—a seemingly logical assumption—then attachment is also integral to encouraging community engagement.

Finally, although race and Internet access were not significant predictors of community engagement, the results reported about the two are nonetheless interesting to examine. Cross-tabulations did show significant relationships between race and community engagement, as illustrated in Table 6.6.

Table 6.6: Community Engagement by Race (n=153)

Less Engaged More Engaged
Other/Did not respond 50% 50%
White – non-Hispanic 60% 40%
White – Hispanic/Latino 50% 50%
Asian/Pacific Islander[17] 66.7% 33.3%
Black/African-American 29.6% 70.4%

As the table shows, those who identified as Hispanic/Latino or Other were equally likely to be less or more engaged. Non-Hispanic Whites were somewhat less likely to fall into the “more engaged” category, and Asians and Pacific Islanders were twice as likely to be less engaged as to be more engaged. However, it was the Black/African-American respondents who were overwhelmingly more likely to be more engaged. Chapter 8 will discuss further how this fits in the context of the neighborhood’s racial demographics, but as a preliminary observation, it should be noted that the racial make-up of survey respondents had far more representation of whites than the actual neighborhood does, fewer Latinos, and fewer African-Americans, as shown in Table 6.1. For this reason, the sample may not be exactly representative of the broader population. However, given the ways in which the survey was distributed online, there is no reason to believe the sample is not representative of hyperlocal media users in the neighborhood. As Table 6.7 shows, Internet access varies by race. However, the findings were only statistically significant for two races: White (non-Hispanic) (Chi square, p = .000) and Black/African-American (Chi square, p = .000). It is interesting to observe the stark difference in broadband access as well as lack of Internet access between the two races, particularly given that Black/African-American respondents had a greater likelihood to be more engaged in the offline community than White respondents.

Table 6.7: Internet Access by Race

No Internet at Home Dial-up DSL Mobile[18] High-speed broadband
White – non-Hispanic 1% 0.5% 26% 1.4% 71.2%
Black/African-American 20.7% 3.4% 34.5% 6.9% 34.5%

To conclude, in developing a model for predicting community engagement, this study found length of time living in Columbia Heights to be the strongest predictor, followed by hyperlocal media use, then Internet access and race. What was interesting about these findings was that the demographic predictors that we often use to characterize political participation were not significantly correlated with community engagement as defined in this study. Correlations between hyperlocal media use, community engagement, and education were not significant; nor were those for hyperlocal media use, community engagement, and age. At the three ANC and two civic association meetings I attended, the mean age was 41 years old, but the lower 45% of the distribution ranged in age from 20 to 35 years old; attendees certainly included their share of younger residents. Similarly, there was a range of ages represented among ANC commissioners. However, the effects of race on the model developed can likely be attributed to the overwhelming likelihood of African American survey respondents to be more engaged (70.4% were more engaged; 29.6% were less engaged).

Using the survey data, we are able to support all the hypotheses tested in this chapter. The finding that contributors to hyperlocal media have a higher degree of community engagement than those who do not corresponds with the suggestion that there are many ways for residents to participate in the public life of their communities today, and not all need take place offline. Additionally, the finding that followers of Columbia Heights’ hyperlocal media are more likely to live in that neighborhood than others provides evidence that these hyperlocal media reach a geographically targeted audience. Long-term residence in Columbia Heights as a predictor of community engagement shows that attachment to place does come into play in participating in one’s community, although there may be a host of other factors associated with long-term residence that this study has not considered, which also influence community engagement. Finally, the positive correlation between community engagement and hyperlocal media use shows that online and offline engagement can go hand-in-hand, and despite Putnam’s (2000) best attempts to argue otherwise, the Internet need not always have a deleterious effect on community ties. Now that we have looked at the intersection of online and offline engagement, the next chapter will focus specifically on the online engagement that exists within these hyperlocal media by analyzing the social networks formed by Columbia Heights’ Twitter users.


[1] Despite his perception, this may not be the case. Of the eight census tracts in Columbia Heights, Pleasant Plains falls in the area with the second-highest poverty level and third-lowest average family income (“Tract 35”). Additionally, research on broadband adoptions in Wards 1 and 8 in Washington showed cost to be the highest deterrent to having Internet access at home (Leach 2011).

[2] All quotes from personal observation at the 15 Feb. 2011 meeting described.

[3] See http://www.nwchca.org, accessed 27 March 2011.

[4] See http://www.youtube.com/user/NWCHCAinDC, accessed 27 March 2011.

[5] This is based on data from census tract 35, as indicated on the website NeighborhoodInfoDC (http://www.neighborhoodinfodc.org/censustract/census.html, accessed 2 April 2011).

[6] All the other meetings I attended during the study had more than 20 attendees.

[7] The full questionnaire as distributed in paper form can be found in Methodological Appendix A; the same questions were included in the online version.

[8] See full list of meetings dates, times, and locations in Methodological Appendix A.II.

[9] All of these demographic figures were calculated using 2010 data from the 8 census tracts that made up Columbia Heights in 2000. Data were used from tract 28.1, tract 28.2, tract 29, tract 30, tract 31, tract 35, tract 36, and tract 37 as indicated on the website NeighborhoodInfoDC (http://www.neighborhoodinfodc.org/censustract/census.html, accessed 2 April 2011).

[10] See a full list of interviewees in the Methodological Appendix B.I.

[11] Unedited audio files of interviews available upon request.

[12] Among the paper surveys, many respondents chose multiple answers for Q. 22, which asks respondents to choose their one, primary use for hyperlocal media. Only the first checked answer was recorded in the data set.

[13] Operational definitions for these variables can be found in the Methodological Appendix A.

[14] From the time of this study, ten years ago dates back to 2001. Since Metro only came to Columbia Heights in 1999, those who lived in the neighborhood prior to 2001 arrived prior to or in the very early stages of the gentrification that has slowly changed the neighborhood since the start of the twenty-first century.

[15] A complete list of survey questions appears in the Methodological Appendix B.II.

[16] The Durbin-Watson value for this model is 1.662.

[17] This was the only finding that was not statistically significant (Chi square, p = .5). However, this is because two cells in the cross-tabulation had fewer than 5 cases.

[18] Note that survey response “mobile access” was intended to indicate mobile Internet as the sole means of access. When “mobile” was given as a response in addition to another form of wire line access, as on a few of the paper surveys, the response was not counted.


Be Sociable, Share!