Content-Type: text/html Communication & Community Involvement: The Link Between Media Use & Civic Participation in a Changing Urban Environment In 1965, Scott Greer noted that the creation of community, which he defines as the sharing of values and objectives, requires a communicative process.[1] A growing body of research has helped reveal the role news media play in facilitating this process. Many of these studies have focused on the association between individuals' community ties, or level of involvement in a community, and their use of the local newspaper. This study builds on that earlier work, expanding on these basic concepts in several ways. Today, newspapers no longer wear the mantel of sole, or even principal, community media. Therefore, this study extends the analysis beyond newspapers to look at a broader range of information variables. But more importantly, changes in culture and lifestyle as well as technological advances have helped redefine the very concept of community. Therefore, the community-involvement measure has been refined to help reflect some of these changes. While most prior research in this area has looked only at the number of groups to which an individual belongs, this measure incorporates the number of hours spent with local groups as a second indicator of community involvement. In addition, the researcher examines a qualitative dimension of community involvement -- the types of ties individuals have to their communities -- more thoroughly than in past studies. A number of specific research questions are addressed: Do people with different levels of involvement in the community differ in their use of information sources? Do individuals with strong community ties depend primarily on newspapers for community news, or do they rely on other news sources? Which members of a community are most likely to rely on "external" news sources? And finally, does type of group involvement -- whether one belongs to a local chapter of the Sierra Club or a neighborhood watch -- affect an individual's use of news sources? Research Background & Literature Review Sociologists explored the importance of newspapers as mechanisms for integrating people into their communities as early as the 1920s. Robert E. Park argued that individuals' "community ties" were associated with newspaper reading.[2] I turn, newspaper use then served to maintain those ties. Robert K. Merton later expanded on this concept, distinguishing between two types of community ties or "orientations" -- "local" and "cosmopolitan."[3] He operationalized community ties as membership in local organizations. A local orientation was associated with interest in local newspaper content. Cosmopolitans displayed some interest in local matters, but were primarily oriented toward the larger society. They were more likely to read both newsmagazines and prestige newspapers external to the community than localities. In the 1950s, work by Janowitz corroborated these findings.[4] He proposed that the greater individuals' involvement in their communities, the greater their use of local newspapers. Janowitz operationalized community involvement as participation in and identification with local organizations and institutions. More recent studies by communication scholars have supported these earlier findings. Sobal and Jackson-Beeck, for instance, found that newspaper readers were more active in their communities than non-readers.[5] Viswanath, Finnegan, Rooney and Potter showed that involvement in local volunteer organizations was associated with subscribing to local and regional newspapers, suggesting that people involved in local groups were more likely to be readers.[6] Their results also suggest that community "elites" who are active in civic or political organization are often exposed to "external" media, such as regional newspapers, as a result of their community roles. However, the authors did not find a significant relationship between subscription to local cable television and community involvement. Stamm and Fortini-Campbell also identified community involvement as among the most consistent correlates of newspaper readership.[7] Some association has also been found between types of community ties and media use. For instance, Stamm reports research results which showed local church involvement correlated positively with subscriptions to both church and commercial newspapers.[8] And the results of several studies suggest that people who are politically active and participate in local elections are more likely to be newspaper readers.[9] Another group of studies explains demographic linkages to media use in terms of individuals' community ties. Newspaper readership has frequently been linked to age, with very young adults more likely to be non-readers.[10] Stamm and Fortini-Campbell suggest that this can be explained by the trend among Americans to establish families at an older age.[11] As a result, family ties, home ownership and other variables that contribute to community involvement occur more often among older people. And Neuwirth, Salmon and Neff found that higher income, education and organizational membership were associated with reading newspapers and magazines external to the community.[12] In other research, Stamm and others found that community involvement may encompass multiple dimensions.[13] These "communities within communities" could refer to both places and social structures. The former might include a suburb within a metropolitan area. An example of the latter would be ties to a social structure, such as a church. Similarly, Tichenor, Donohue and Olien proposed that a community may comprise a variety of interacting subsystems, including the mass media.[14] According to Stamm, the causal order of the association between community involvement and media use has also been subject to debate.[15] In some cases, researchers have treated community involvement as an antecedent of media use. Other studies have treated media use as an independent factor that strengthens community ties. Still a third approach suggests that the two are reciprocal. However, Stamm suggests that few, if any, studies have shown "a temporal relationship in which newspaper use precedes construction of a community tie." In summary, a strong body of evidence links newspaper reading and community involvement. However, few studies have sought to determine whether this same association exists with any other information variables. In addition, the concept of community involvement has traditionally been fairly narrowly defined. Thus, this study builds on previous research in this area by refining both the communication and the community involvement variables. Research Method Data for this study was collected in a telephone survey during March 1994 in Austin, Texas, a city of approximately one-half million residents. Students in two communications classes at the University of Texas at Austin conducted the survey. The study population included all Austin residents age 18 and older. Using a systematic random sampling method, students drew a sample of one original telephone number and three alternates from each page of the Austin telephone book, for a total of 500 original numbers. Before beginning the surveys, interviewers added a "1" to the last digit of every phone number to ensure that unlisted phone numbers would be included in the sample. A total of 474 respondents comprised the final sample, with a sampling error of 4.2. Graduate students then used the Statistical Package for the Social Sciences (SPSS) to analyze the data. The survey covered a range of different topics. This report was based on respondents' answers to questions about media use (including discussion of issues with family and friends) and their involvement in Austin-area groups and organizations. Results: Frequency Analysis A frequency analysis constituted the first phase of data analysis. The study sample split nearly evenly in terms of gender, with around 49 percent male and 51 percent female. Anglos comprised the largest racial group (78 percent), followed by Latinos (10 percent), African-Americans (6 percent), and Asians (2 percent).[16] Approximately 2 percent of the interview subjects responded "other" when asked to give their race. The respondents ranged from 18 to more than 70 years in age, but the majority were middle-aged or younger. Almost half (44 percent) were under 35. And more than three-fourths (80 percent) were under 50. Only 13 percent of those interviewed were older than 60. The Greater Austin Chamber of Commerce claims that Austin is the most highly educated among U.S. cities of its size. The poll results reflected that. [17] Nearly half of the respondents (48 percent) had received college degrees; 14 percent of this group also held graduate degrees. Another third of the sample (30 percent) had completed some college or technical training. And approximately one-fifth (18 percent) were high school graduates. Most of the respondents lived in middle-income households. About 22 percent reported annual incomes between $10,000 and $19,000, and about 30 percent between $20,000 and $59,000. Another 14 percent lived in households with incomes of $60,000 or above. Only about 6 percent of annual household incomes fell below $10,000.[18] Although five colleges and universities are located in Austin, with a combined student population of around 87,000, only 12 percent of respondents said they were full-time students.[19] Married respondents comprised about half of the sample (49 percent). Another 34 percent said they were single, and around 16 percent were either widowed or divorced. Community Involvement Community involvement was measured on a scale ranging from "none" to "high." As conceived here, the variable comprises two indicators -- "Membership" and "Hours." The Membership factor indicated the number of local organizations to which an individual belonged. A rating of none meant that the respondent belonged to no organizations. Low involvement was defined as membership in one or two groups. High involvement meant membership in three to 10 groups, however, only 6 percent of respondents belonged to more than five. The Hours indicator of community involvement measured the total number of hours per month individuals spent with the group in which they were most involved. Only respondents who belonged to at least one group could be rated on this factor. Low involvement was operationalized as spending up to 10 hours a week in group activities including meetings, committees and projects. High involvement was defined as spending more than 10 hours a month on organizational activities, although only 10 percent of respondents spent more than 40 hours with any one group. Half of the 474 respondents in this survey did not belong to any local organizations. About a third belonged to one or two, and the remainder belonged to between three and 10 groups. Of the respondents who belonged to at least one group, 60 percent spent one to 10 hours a month with the organization in which they were most involved. About 40 percent spent more than 10 hours a month with one group. Another aim of the study was to determine the types of community groups to which people belonged and how that affected their use of news sources. Out of 234 locally "involved" respondents, around 19 percent belonged to a religious group (includes church membership), 15 percent to a community service group, 10 percent to a school-educational group, 10 percent to a business-professional group, 9 percent to a political-governmental group, 8 percent to a recreational group, 6 percent to a social group, 6 percent to a youth group and 5 percent to an environmental group. Use of News Sources The news sources examined in this study included a local newspaper, local television newscasts, national television-network newscasts, news magazines, radio, and discussion with family and friends. The researcher conceived of radio as a local news medium, assuming that radio news programming would include local items. The exception might be a nationally produced news show such as National Public Radio's "All Things Considered." But the specific shows respondents listened to could not be determined from the study data. The variable was measured on a three-level scale which ranged from "never/seldom" use to use "every day or nearly every day." More than half of the respondents (51 percent) ranked as high-frequency readers (read every day or nearly every day). Only 14 percent of the sample subjects counted as "non-readers" (never or seldom read the newspaper). About 35 percent were low-frequency readers (one to four days a week). The frequency with which respondents watched local television news showed a distribution pattern similar to that for newspapers. A majority (59 percent) said they watched daily or nearly every day. Another 29 percent watched with low frequency. Only 13 percent of respondents seldom or never watched. In contrast, 24 percent of subjects never or seldom watched network newscasts. Forty-two percent of those who said they were viewers fell at the low end of the viewing scale. Another 34 percent watched with high frequency. About two-thirds of respondents said they never or seldom watched CNN, the Cable News Network, or watched it with low frequency. About 28 percent watched daily or nearly every day. More respondents said they never or seldom used newsmagazines than any other news source. Some 58 percent of subjects fell into this category. Another 23 percent read them one to three times a month and 20 percent read them every week. The distribution pattern for radio use was just the reverse. About 30 percent seldom or never listened to radio news, 16.3 listened one to four days a week and more than half listened daily or nearly ever day. Very few respondents (13 percent) said they never or seldom discussed the news with their family or friends. And almost half (44 percent) said they discussed the news nearly every day or daily. Results: Cross-Tabulations To complete the second phase of data analysis, three sets of cross-tabulations were run and tested for significance with Pearson's Chi-square. The first set compared community involvement, measured both as number of group memberships (Membership factor) and amount of time spent with group (Hours factor), with demographic variables including race, sex, age, education, income and marital status. The researcher treated the demographics as independent variables. These tests detected several significant relationships. TABLE 1 Community Involvement Measured as Number of Group Memberships* by Education Memberships Education None (%) Low (%) High (%) High school 71.6 21.6 6.9 (N=102) Some college 50.3 32.2 17.5 (N=143) College grad 41.2 32.9 25.9 (N=228) Missing Cases=1 Pearson Chi-square=29.66, DF=4, p<.01 *Community Involvement is based on two factors. "Membership"=total number of local groups to which respondent belongs. (Question: Now I'd like to ask some questions about your involvement in Austin-area groups or organizations. In the past three years, about how many groups, clubs, or organizations have you been a member of?) "Hours"=Hours per month respondent spends with group in which he/she is most active. (Question: Think of the group that you've been most active in. About how many hours a month do you spend with that group, including meetings, committees, projects and similar activities?) Low=1-2 local group memberships High=3-10 local group memberships Cross-tabulations of the Membership factor of community involvement with the set of demographic variables detected significant relationships between the Membership variable and both education and race. The former relationship appeared to be the stronger of the two (p<.01). As shown in Table 1, one-quarter of respondents with college or graduate degrees belonged to three or more community groups (high involvement), compared with around one-fifth of those with some college or technical training, and less than one-tenth of those who had attended or graduated from high school. Approximately one-third of all respondents with a college degree and one-third of all respondents with some college or technical training, versus one-fifth who had a high school degree or less, belonged to one or two community groups (low involvement). The majority of respondents at each educational level said they were not involved in any group. But nearly three-fourths of those subjects with a high school education or less were not involved, compared with half of those with some college/technical school and less than half of those with a college degree or more. In summary, the more highly educated respondents appeared to fall higher on the community involvement scale when measured by number of group memberships. No significant relations was found between education and the Hours measure of community involvement. TABLE 2 Community Involvement Measured as Number of Group Memberships by Race Memberships Race None (%) Low (%) High (%) Anglo 48.2 32 19.98 (N=369) Black 60 26.7 13.3 (N=30) Latino 65.3 16.3 18.4 (N=49) Asian 45.5 18.2 36.4 (N=11) Other 36.4 63.6 ---- (N=11) Missing Cases=4 Pearson Chi-square=16.13, 8DF, p<.05 Table 2 shows that a cross-tabulation of race by the Membership factor of community involvement also proved significant (p<.05). Some 36 percent of Asians belonged to more than two community groups (high community involvement). Lower percentages of Anglos (20 percent), Latinos (18 percent) and African-Americans (13 percent) belonged to more than two groups. About a third of Anglos and blacks belonged to one or two organizations (low involvement), compared with less than one-fifth of Asians and Latinos. More than half of Latinos and blacks and less than half of whites and Asians were not members of any groups. Thus, Asians and whites appeared to be involved in more community groups than Latinos and blacks. TABLE 3 Community Involvement Measured as Hours Spent With Group Per Month* by Race Hours per Month Race Low (%) High (%) Don't know (%) Anglo 60.7 38.2 1 (N=191) Black 30.8 53.8 15.4 (N=13) Latino 66.7 33.3 ---- (N=18) Asian 80 20 ---- (N=5) Other 71.4 28.6 ---- (N=7) Missing Cases=240 (Includes all respondents who did not belong to at least one group.) Pearson Chi-square=19.25, 8DF, p<.05 *Community Involvement is based on two factors. "Membership"=total number of local groups to which respondent belongs. (Question: Now I'd like to ask some questions about your involvement in Austin-area groups or organizations. In the past three years, about how many groups, clubs, or organizations have you been a member of?) "Hours"=Hours per month respondent spends with group in which he/she is most active. (Question: Think of the group that you've been most active in. About how many hours a month do you spend with that group, including meetings, committees, projects and similar activities?) Low=1-10 hours a month High=More than 10 hours a month The researcher also found a significant association with race using the Hours factor of community involvement, or the hours per month spent with the organization in which the respondent is most active (p<.05). As shown in Table 3, more than three-fourths of Asians and more than half of Latinos and Anglos said they spend between one and 10 hours a month on group activities (low involvement). More than half of blacks, in contrast, spend in excess of 10 hours a month involved with one organization (high involvement). In summary, African-American respondents appeared to devote more time than members of other ethnic groups to the community organizations in which they were most involved. Asians appeared least likely to spend more than 10 hours a month with any one group. Finally, the Hours factor measure of community involvement showed some association with age (p<.01). Nearly three-fourths of respondents in the 25- to 39-year-old age group and more than half in both the 40- to 54-year-old group and the 55 and over group spent fewer than 10 hours a month with any one group (low involvement). The group of respondents between 18 and 24 split nearly evenly between the low and high community involvement levels. In summary, respondents between 25 and 34 years seemed less likely than other age groups to spend more than 10 hours a month with a local group. Type of Group and Media Use The second set of cross-tabulations examined the relationship between types of community memberships and media use. The type of group in which an individual was most active served as the independent variable. Here the analyses detected two significant relationships. Both involved television viewing. The most significant association appeared between group type and viewing of network newscasts (p<.01). About half of those respondents who were most active in a social group said they never or seldom watched network news. A third of those in an education-school group, a quarter of those in a business-professional group, a quarter of those in a community service group, and about a fifth of those in a religious group said they did not watch. The majority of members who were involved in a youth group (69 percent), a recreation group (61 percent), a religious group (54 percent) and a political-governmental group (50 percent) said they watched one to four days a week (low frequency). Those group types which had large percentages of members who said they watched network news every or nearly every day (high frequency) included political-government (41 percent) and community service (49 percent). Thus, members of a social group appeared least likely to watch network news on a daily basis, while members of a political-governmental or community service organization were most likely to watch daily or nearly daily. Watching local television news also varied significantly with group type (p<.05). About 36 percent of respondents who were involved in a social group did not watch, while 21 percent of those who were members of an educational-school group said they were non-viewers. Thirty-nine percent of the members of social and youth groups said they watched one to four days a week (low frequency), followed by religious (34 percent) and educational-school (33 percent). A majority of the members of a number of groups said they watched daily or almost daily (high frequency), including community service (77 percent), political-governmental (68 percent), youth (62 percent), recreational (56 percent), business-professional (54 percent) and religious (53 percent). In summary, people who were most active in a social group appeared least likely to be daily viewers of local TV news, and members of political-governmental or community service groups were most likely to be daily viewers. These results closely resemble those found in the cross-tabulation between network news viewing and group type. Community Involvement and Media Use The final set of cross-tabulations compared community involvement (Membership and Hours factors) with use of news sources, treating the former variable as independent. Several significant relationships were found. Not surprisingly, given the association between these variables documented in the literature, newspaper use showed a strong positive relationship with community involvement operationalized as number of group TABLE 4 Community Involvement Measured as Number of Group Memberships by Newspaper Use Memberships Newspaper use None (%) Low (%) High (%) Seldom/never 19.1 12.7 3.31 Low frequency 39 31.7 28.6 High frequency 41.9 55.6 68.1 (N=236) (N=142) (N=91) Missing Cases=5 Pearson Chi-square=24.07, 4DF, p<.01 Low=1-2 local group memberships High=3-10 local group memberships Low frequency=1-4 days a week High frequency=every day or nearly every day memberships (p<.01). For example, as shown in Table 4, just 3 percent of those respondents who belonged to between three and 10 local groups (high involvement) said they never or seldom read a newspaper. In contrast, 13 percent of those who belonged to one or two local groups (low involvement) and 19 percent of those who belonged to no local groups said they seldom or never read a newspaper. Differences were less pronounced at the low frequency readership level, with about a third of both high community-involvement respondents and low-involvement respondents and 40 percent of respondents who belonged to no community groups saying they read one to four days a week. Nearly three-fourths of the high involvement group said they read the paper nearly every or every day versus just over half of the low involvement group and less than half of subjects who held no memberships. These results support earlier research which suggests that people who are involved in their communities are interested in local newspaper content. However, newspaper readership was not found to be significantly related to the number of hours individuals spend with one group. TABLE 5 Community Involvement Measured as Number of Group Memberships by Radio Use Memberships Radio use None (%) Low (%) High (%) Never/seldom 34.7 25.9 26.4 Low frequency 18.8 14.7 12.1 High frequency 46.4 59.4 61.5 (N=239) (N=143) (N=91) Missing Cases=1 Pearson Chi-square=9.36, 4DF, p<.05 Low=1-2 local group memberships High=3-10 local group memberships Low frequency=1-4 days a week High frequency=every day or nearly every day A cross-tabulation for radio use and community involvement proved significant for the Membership factor (p<.05), but not the Hours factor. As shown in Table 5, using the Membership factor measure, almost two-thirds of the people who belonged to three or more groups (high involvement) listened to radio news daily or almost daily. A slightly smaller number (59 percent) of those who belonged to one or two groups (low involvement) were high-frequency listeners. In contrast, less than half of those who held no group memberships were high frequency listeners. Differences between group members and non-group members were not as pronounced at the other two levels of frequency for radio use. About 12 percent of those high in community involvement, 15 percent of those low in community involvement and 19 percent of non-members said they listened one to four days a week (low frequency). And around a quarter of both low- and high-community involvement respondents said they never or seldom listened to radio news, versus 35 percent of non-group members. In summary, respondents who belonged to local organizations tended to listen to radio news more frequently than non-members. The final cross-tabulation that proved to be significant compared the Membership Factor for community involvement and the frequency with which individuals discuss the news with their family and friends (p.<.01). As seen in Table 6, about 60 percent of respondents who belonged to three or more community groups discussed news events daily or nearly daily. In contrast, about 40 percent both of those who belonged to one or two groups and those who held no memberships discussed the news daily. Only a small percentage of all three groups of respondents said they never discussed news events -- about 9 percent of the high community-involvement group, 11 percent of those in the low-involvement group and 16 percent of those who were not involved in any TABLE 6 Community Involvement Measured as Number of Group Memberships by Discussion Memberships Discussion None (%) Low (%) High (%) Never/seldom 16.4 10.6 8.8 Low frequency 43.3 49.3 31.9 High frequency 40.3 40.1 59.3 (N=238) (N=142) (N=91) Missing Cases=3 Pearson Chi-square=13.95, 4DF, p<.05 Low=1-2 local group memberships High=3-10 local group memberships Low frequency=1-4 days a week High frequency=every day or nearly every day groups. In summary, individuals with many community ties, measured as group memberships, appeared more likely than individuals who were not as involved in the community to discuss the news with their family and friends on a daily basis. Discussion & Conclusions Many of the findings from this study support earlier examinations of the nature of the association between individuals' community involvement and their use of news sources. The significant relationship found between newspaper readership and community involvement presents perhaps the most important example. Judging from these results, newspapers remain an important instrument for integrating people into their communities. However, this relationship only appeared when the researcher, following past studies, used the number of local organizations to which an individual belonged as a measure of community involvement. It was not evident when involvement was measured as the number of hours per month survey respondents spent with the group in which they were most active. This and other data reported here suggest using only one community involvement indicator may produce an incomplete or even misleading picture of community involvement. For instance, the cross-tabulation between race and the Membership factor of involvement showed that Asian-Americans were more likely than other ethnic groups to be members of many local organizations. However, a test of race and the Hours factor of involvement found that African-Americans spent more hours per month with a single group. Stamm conceives community involvement as encompassing multiple dimensions.[20] In a city the size of Austin, people's affiliations may not extend community-wide. Instead, they may be linked to smaller "communities within communities," to use Stamm's phrase. And the nature of their involvement may vary accordingly, as a function of the culture dominant in their primary group or groups. These differences in the nature of communities and community involvement warrant further investigation, perhaps with a larger sample which could include a more substantial number of ethnic minorities. However, unlike the Membership factor of community involvement, the Hours factor did not appear to be associated with media use. This researcher suggests that this may be due to certain cultural and class-based biases. Within the Anglo community, people who belong to many local groups generally share certain other demographic characteristics such as high levels of education and income, which also correlate with high (news) media use. In the African-American community, some people may be more involved in terms of the amount of time they spend with any one group, but they may not share these same demographic characteristics. This research also detected some previously unexamined relationships. For example, the amount of time people spend discussing news events with family and friends related significantly to community involvement, measured as number of group affiliations. Lazarsfeld, in his landmark study The People's Choice, determined that people receive much of their information from other people, in a "two-step flow of communication."[21] This results when less-informed people seek information from "opinion leaders" who use other news sources to keep up with politics (and presumably other local affairs). Other researchers have found that community involvement measured by number of memberships often is associated with community leaders who are active in civic affairs. These findings suggest that these opinion leaders may comprise the most involved segment, using Membership as the measurement criteria, of the population. The positive relationship between group membership and education also seems to support this interpretation, since research has shown that newspaper readers also tend to be among the most educated members of a community.[22] In past years, the growth of talk radio has renewed interest in this news source. The fact that community involvement, using the Membership factor, proved to be significantly related to radio use, suggests that further work should be done here as well. Finally, past studies have detected few significant relationships between television news viewing and community involvement measured with quantitative indicators such as number of group memberships held. The findings presented here indicate that type of membership may be a more important determinant of television use. A larger sample would improve further research into these associations, since responses can easily comprise a dozen or more categories of memberships, from "environmental" to "social" to "military." All of these findings suggest that there is a great deal of research yet to be done in this area of the field. [1] Scott Greer, The Emerging City (New York: Free Press, 1956). [2] Robert E. Park, "The Natural History of the Newspaper," in W. Schramm, e d. , Mass Communication, 2d ed. (Urbana: University of Illinois Press, 1975, pp. 8-23). [3] Robert K. Merton, "Patterns of Inf luence: A Study of Interpersonal Influences and of Com municaiton Behaviors in a Local Community," in P. Lazarsfeld and W. Stant on, eds, Communication Research, 1948-49 (New York): Harper & Co., 1950 . [4] Morris Janowitz, The Community Press in an Urban Setting (Chicago : Chicago University Press, 1952). [5] Jeff Sobal & M arilyn Jackson-Beeck, "Newwspaper Non-Readers: A National Profile," Journalism Quarterly, 58:9-14 (Spring, 1981). [6] Kasisoma yajula Viswanath, John R. Finnegan Jr., Brenda Rooney & John Potter, "Community Ties in a Rural Midwest Community and Use of Ne wspapers and Cable Television," Journalism Quarterly,, 67:899-911 (Winter, 1990). [7] Keith Stamm & Lisa Fortini-Campbell, "Th e Relationship of Community Ties to Newspaper Use," Jo urnalism Monographs, 84 (August, 1983). [8] Keith Stamm, Newspaper Use a nd Community Ties: Toward a Dynamic Theory (Norwood, N .J.: Ablex, 1985). [9] Sobal & Jackson, op. cit.; Viswanath, Finnegan, Rooney & Potter, op. cit. [10] Paula M. Poinderxer, "Daily Newspaper Non -Readers: Why They Don't Read," Journalism Quarterely, 56:764-770 (Winter, 1979). [11] Stamm & Fortini-Campbell, op. cit. [12 ] Kurt Neuwirth, Charles T. Salmon & Mary Neff, "Community Orientation an d Media Use," Journalism Quarterly, 66:31-39 (Spring, 1989). [13] Stamm, op. cit. [14] Philip J. Tichenor, George A. Donohue & Clarice N. Olien, Community Conflict and the Press (Newbury Park, CA: Sage, 1989). [15] Keith Stamm, "Community Ties and Me dia Use," Critical Studies in Mass Communication, 5:35 7-361 (September, 1988). [16] The number of Latinos interviewed was low er than expected, given Austin's demographic profile. The difficulties students encountered interviewing Spa nish-speaking heads of households may account for this. [17] In fact, th e survey found a higher level of education than expected among respondents. Researchers speculated that a lower response rate a mong the less-educated segments of the population migh t account for this. [18] This percentage was smaller than expected. But low-income subjects probably were among the respondent s who refused to answer this question. [19] The survey did not account f or part-time students. [20] Stamm, op. cit. [21] Shearon A. Lowery and Melvin L. De Fleur, Milestones in Mass Communication Research (New York: Longman, 1983). [22] Neuwrith, Salmon & Neff, op. c it.