Does Streaming Undermine Mainstreaming? Finding Common Cultural Ground in Divisive Times

Abstract

This study assesses whether the mainstreaming hypothesis, derived from cultivation frameworks developed during the mass audience era, remains operative in a digital media environment characterized by fragmenting media and cultural taste publics. In particular, we consider evolving conceptions of mainstreaming that stimulated our research questions and hypotheses in four surveys conducted from 2015 to 2024. We broaden our view of media to see if entertainment content—especially film genres—can provide common ground in attracting people with little else in common. Results suggest that such “cultural mainstreaming” may occur by providing common gratifications and impact global indictors of our lives—happiness, community attachment, feelings about our quality of life, and perceived cosmopoliteness. But the results are limited to a general adult population, not the younger students studied. The findings apply only to the general adult population and not to the younger student sample examined. Overall, the results indicate that the cultivation effect is relatively weak; the small number of significant relationships observed does not appear to exceed what might be expected by chance. Taken together, these findings suggest that mainstreaming and media influence operate as more complex processes in the digital era.

1. Introduction

In the heyday of cultivation research in the 1970s and 1980s, mainstreaming was advanced as the idea that mass media—especially television—essentially reported the “same thing,” pulling viewers toward the center of the political spectrum (e.g., Gerbner et al. 1980). The “long tail” of new media continues to fragment audiences (e.g., Anderson 2006; Atkin and Litman 1986). This narrowcasting—in concert with algorithmically-driven ideological platforms that further polarize voters (Kubin and von Sikorski 2021)—fuels rising concerns that media are pulling us apart, rather than bringing us together (e.g., Jeffres et al. 2001).

The present examination looks elsewhere for “common ground,” moving from the news media that are the subject of so much current analyses, to entertainment platforms that have the potential for bringing together people with different political views. We suggest that “cultural mainstreaming” may occur when people with little else in common are attracted to films, videos, television programs, or live concerts that provide common gratifications and perhaps impact more global indicators of our lives—happiness, attachment to our communities, how we feel about the quality of our life, and whether we see ourselves as parochial or more cosmopolitan. Here, we will examine the historical literature that undergirds the cultivation and mainstreaming framework, before considering entertainment media and the diverse media platforms of today, which stimulated our research questions and hypotheses tested in surveys across the past decade.

2. Literature Review

2.1. Cultivation and Mainstreaming

Mainstreaming was part of media cultivation, a “grand theory” that sees media images molding society through the long-term presentation of relatively uniform versions of social reality. According to the theory, originally proposed by George Gerbner (1969, 1998), media define what is normal, or deviant, and those who are heavier viewers of television (in particular) are more likely to have views and beliefs similar to those televised. Cultivation theory grew out of studies of violence in the media and the relationship between television exposure and viewers’ perceptions of social reality (e.g., Gerbner and Gross 1976). Studies showed that heavier viewers were more likely to harbor beliefs or knowledge that matched those evident on television. Gerbner concluded that television viewing cultivated a distorted view of a “mean and scary world” (Gerbner et al. 1982). His team argued that it did not matter much what one watched because television channels offered pretty much the same thing.

Although TV received the most attention from researchers, other media also have been examined. Moreover, early research found some support for the theory. Perse et al. (1994) confirmed, for example, that heavy TV viewers are more likely to express alienation and fear of crime (Gerbner and Gross 1976; Gerbner et al. 1978, 1979), are more likely to describe their lives as less satisfying (Morgan 1984), are more likely to see the elderly as feeble and ineffectual (Gerbner et al. 1980) and are more likely to express more sex-stereotyped beliefs (Morgan and Rothschild 1983).

Cultivation research assumes that people who do not watch TV—or who watch less TV—have a “better” chance of giving more accurate perceptions of reality (e.g., the percentage of people in various occupational categories) or more accurately estimating the probability that they will be victims of crime, etc. However, sources of influence are always complex, encompassing not only mass media but also interpersonal communication and personal observation. Without TV, people must rely on those other sources, none of which may enhance their estimates. As Jeffres et al. (2001) note, the problem is one of comparisons “People with little media use are likely to depend on conversations with friends and neighborhood observations, neither guaranteed to improve their performance on cultivation measures. What is the “relevant point” with which to compare TV answers?” (pp. 409–10). They echoed Potter’s (1991) key criticism of cultivation research: determining the “TV answer” is nearly impossible, because viewers draw on numerous contextual elements within portrayals when interpreting meaning from TV (p. 569).

For instance, Kim (2025) reports data suggesting that exposure to entertainment TV shows like Shark Tank—featuring rags-to-riches storytelling—perpetuate a meritocratic ideology that may “glorify the winners of the economic system, increase tolerance for income inequality, and dampen mass support for politics that could support those left behind” (p. 13). Such findings broadly support mass-society conceptions that see media serving a hegemonic function (e.g., Gitlin 1980), as even early conceptions of cultivation theory cast media as homogenizing forces of social control (e.g., Gerbner 1969). The rather weak and inconsistent cultivation effects found more recently (e.g., Chaffee and Metzger 2001; Jeffres et al. 2023) may reflect the moderating influence of fragmenting daily me environments that polarize public opinion (e.g., Kim 2025).

As Jeffres et al. (2001) note, however, we do not face the same comparison problem nor rely on the same assumptions applying Gerbner’s mainstreaming concept to issues of public opinion or perception. Mainstreaming proposes that attention to mass media—especially TV—draws people to the center. Although once used in a pejorative sense, the case can be made that mainstreaming should occur regardless of the values projected on TV or in the media. Indeed, if moderates, middle-of-the-roaders, dominate society, then they also should seek out access as authoritative sources. Thus, they would be over-represented if the media were merely doing their job of trying to “represent” society (Jeffres et al. 2001, p. 410). Interestingly, in the 1970s and 1980s, more “radical” critics saw cultivation and mainstreaming processes as foundations for media bias against more progressive or leftist views and towards the center, “common ground.” Today, the concern shifts as media pull people apart and traditionalists wonder if the center will hold because extremes are succeeding and observers wonder if there is a “common ground” anymore.

Evidence in support of mainstreaming was found on the prevalence of racial integration (Matabane 1988). Following earlier work by Gerbner and his colleagues (Gerbner et al. 1982), Piepe et al. (1990) examined the notion of mainstreaming in Britain, finding a relationship between heavy TV viewing and political centrism. Results confirmed that heavy viewers were more likely to place themselves in the center rather than on the left or right and to occupy the mid-point on a scale tapping attitudes toward health, education and welfare. Those in the center also were more likely to name TV as their favorite source of political information and said they were more dependent upon TV during election campaigns. Heavy viewers of current affairs were less likely than light viewers to be in the center, measured by either self-placement or by attitude.

It may be profitable to rephrase the mainstreaming hypothesis so that we focus on whether attention to media increases the likelihood that someone “shares” in a consensus on important issues or perceptions. Additionally, we could examine whether heavy users of certain media are more homogeneous, i.e., whether the variance on a perceptual or attitudinal measure is reduced with increasing exposure. Jeffres et al. (2001) sought to expand the range of measures examined in mainstreaming and cultivation analysis beyond the estimates of violence or crime, or occupational distributions most frequently used at the time. Using three surveys conducted in a metropolitan area of the Midwest, they used not just TV but other media exposure variables as independent measures and included a broad range of dependent measures focusing on both local and national levels, general attitudes, beliefs about human nature and the state of society, and beliefs about one’s personal status. In their mainstreaming analyses, a substantial number of differences were found only for newspaper reading, and eight of the nine comparisons were in the reverse direction—meaning that heavier readers were less likely to agree with the norm of the distribution.

A pair of differences were found for TV viewing and for movie viewing, and all were in the reverse direction, meaning that heavier viewing groups were more likely to be in the extremes rather than the modal center expected by mainstreaming theory. With 42 items compared for cultivation analysis and 25 items compared in mainstreaming analysis for seven media groups, chance alone should produce some differences between media exposure groups; thus, the pattern is less than overwhelming for cultivation and mainstreaming theory as generally stated. Comparing cultivation and mainstreaming differences for local and national/international issues, they found a slightly stronger pattern for local issues, where eight items generated seven cultivation differences and four mainstreaming differences between media exposure groups; all four mainstreaming differences were in the reverse direction, suggesting that heavy media exposure (newspapers, radio and books) lead to less commonality rather than more agreement—a “scatterstreaming” rather than a “mainstreaming.”

Overall, the media did a better job of creating diversity, of “scatterstreaming” rather than “mainstreaming.” Some 11 differences between media use groups ran counter to the mainstreaming thesis, with three other differences approaching statistical significance. In only one case was a difference consistent with the mainstreaming hypothesis—more frequent newspaper readers were more likely to be in the “middle” on a general attitudinal item, while less frequent readers were more likely to be in the extremes.

In the current media environment, do cultivation and mainstreaming processes operate to pull people together, or tear them apart? The present study examines this research question before shifting in search of common ground elsewhere. This is no trivial question. Maintaining cohesion is essential to a functioning democracy, but that becomes increasingly difficult as diversity increases, and the public fragments into news silos that weaken our cultural common ground (e.g., Anderson 2006; Jeffres et al. 2023). What remains to hold people together, to enable institutions to solve problems in the face of conflict and divisiveness?

During the mass audience era, Cotteret (1972) noted that the media had become the common denominator, and some have suggested that the entertainment sphere may be the remaining place that people come together. However, Chaffee and Metzger (2001) questioned “whether mass communication is a fleeting idea, a purely 20th-century phenomenon” (p. 375). At the same time, Jeffres et al. (2001) found empirical support for the notion that fragmentation wrought by emerging media has replaced mainstreaming conceptions with a dynamic more reminiscent of “scatterstreaming. Today’s increasingly balkanized “post truth” media environment may thus be eroding the foundations of shared meaning to which our democracy is tethered (e.g., Kim 2025).

2.2. Theorizing on the Loss of Cohesion with Fragmentation

Heinberg (2022) notes that social cohesion is vital and it is being lost, and it is happening not only at the national level but also within communities, large cities and urban areas. “In order to maintain cohesion, we humans engage in near-constant signaling of group membership and participation. Choices of entertainment, clothing, and food—and, today, whether to wear a mask and get vaccinated—are all fraught with in-group, out-group symbolism. Further, the groups (or publics) we align with may range from country to political party, religion, military organization, criminal organization, and corporation, all the way down to the family (Graan 2022). Each of these groups and sub-groups has its own requirement for loyalty signaling—which can be problematic if these groups do not have aligning values or goals” (para 1). He adds that social cohesion can slip away for a society that becomes internally divided, people increasingly distrust authority figures, and society loses its ability to solve problems.

This dynamic is unfolding alongside the social media revolution, which has prompted fragmentation through news algorithms that enhance news silos (McCarthy 2022), stoking conspiracy theories and extremist views—the “echo-chamber effect.” Beniger (1987) noted some time ago that the proliferation of media options would lead to fragmentation which would reduce cohesiveness. Edsall (2023) surveyed analyses and opinions of five political scientists with relevant expertise and concluded America is approaching a tipping point because nothing in the system is pulling things back towards the middle.

Concern over the loss of social cohesion is not unique to the United States but exists around the world (Jenson 2019). This notion struggles to co-exist with pluralism, which recognizes and values differences across ethnicity, religion, language and other distinctions, not only at the national level but also within communities. Many are dissatisfied with democracy as a system, in part because it has not delivered the quality of life they seek (Sozan 2024).

Survey researchers have relied on a single measure of people’s perceptions—right/wrong track polling data—to capture the American public’s sentiment about the country’s current trajectory, offering insights into the collective mood and potential desire for political change or stability. Though often linked to how people feel about the economy and their own financial circumstances, there is also recognition that this includes sentiment about social changes. Thus, while someone may be comfortable financially, they may be conservative and fearful about LGBT+ issues and the popularity of minorities in popular culture.

Separated from the political and economic implications, people’s reactions to pollsters’ queries should be extended to broader measures of how they feel about the quality of life, their own feelings of happiness, and attitudes toward their communities and key domains such as media and leisure. We all live in a physical community and spend considerable time engaging in media use and taking advantage of opportunities for leisure available. Indeed, this accompanies parallel concerns that we are spending too much time paying attention to “screens” on cell phones or other media (e.g., Rideout et al. 2022).

Although some see the public today being detached from their communities, we have considerable evidence that community continues to make important contributions to our well-being and feelings of satisfaction, maybe even stability. We moan when favorite institutions stumble, schools have to merge, neighborhood stores close, and even shopping centers empty as shopping shifts to the Internet. While traditionalists hang on to their physical newspapers, broad sections of the public have shifted to mobile media (if any source) for their news, so the changing digital environment forces us to reconsider established relationships from past research. Moreover, as audiences segregate into their news silos, the public shares less news in common but also less popular culture fare, leading to fears that our divisions will feed on themselves. If a concern with stability of our country, our culture, and respect for our values is reflected in the right/wrong track measure of the direction the country is headed, should we expect it to also be related to people’s feelings of attachment to community, their perceptions of the quality of life, reports of happiness, and their assessments of media and leisure opportunities?

Research over several decades has documented positive relationships between people’s attachment to their communities, quality of life perceptions, and leisure pursuits that include mass media and public leisure activities. Rothenbuhler et al. (1996) found newspaper reading related to community attachment, while Stamm et al. (1997) found several media use variables (newspaper reading, TV viewing, radio listening and interpersonal communication within the community) related to community involvement. Yamamoto (2011) found that community newspaper use promoted social cohesion. Jeffres (2002) notes that, “Neighborhood relationships and communication are linked to neighborhood attachment expressed as use of local facilities, belonging to organizations and feelings of attachment. Studies provide much evidence for the interrelationships among media use, measures of community involvement, and measures of community attachment” (p. 182).

In addition to the literature on quality of life and community attachment, we need to examine leisure theory and explanations for the relationship between participating as a spectator in public leisure behaviors and mass media use. Film and leisure pursuits, especially the arts, have been linked to people’s happiness, their quality of life, and even the pursuit of the “American dream.” Researchers from various academic disciplines focusing on happiness have examined arts and cultural activities (Brajša-Žganec et al. 2011; Hand 2018; Pagán 2015; Wheatley and Bickerton 2017; Zhang et al. 2017), finding that attendance at arts and cultural activities has a positive impact on the individual’s happiness and life satisfaction.

Using three surveys in a metropolitan area, Jeffres and Dobos (1993) found that people’s perceptions of the leisure opportunities available in their communities and the importance of such opportunities were related to perceived quality of life as well. Actual media use—where people learn via the news about leisure activities—also was related to leisure perceptions and leisure values. Commercial entertainment activities in urban areas have proliferated at the same time that new technologies have increased the mass media menu from which people may select. While witnessing these two trends, mass communication scholars have employed theory that portrays media behaviors as both facilitators and competitors for opportunities to fill people’s free time. First, research shows that media are most effective in conveying information. Thus, since the media provide copious amounts of information about public leisure options, we would expect such instrumental uses of the media to be positively related to participation in such public activities, largely as spectators, since those aware of the options are more likely to participate—other factors being equal.

However, media also represent leisure options, so that time spent viewing TV, for example, competes with attendance at sporting events or the theater. Earlier research shows that media take about half of people’s free time, while sports and various leisure pursuits take about 8%, our social life 18%, and walking, resting and attending “spectacles” accounting for the rest. In another subsequent study, Jeffres et al. (2003) discovered that the level of media use was positively related to the level of leisure activity even when social categories were controlled, suggesting that media are strongly integrated into the modern entertainment system, that the media system has changed dramatically over the past three decades, as the digital system has dramatically changed people’s access to mobile technology.

Wheatley and Bickerton (2017) investigated the effects of arts and culture activities and sports on general happiness. Using 2010 and 2011 data from the Understanding Society Survey, they found that attending arts events, visiting museums and historic sites, and sports had a positive relationship with life satisfaction and general happiness. Zhang et al. (2017) found that performing arts activities had a positive relationship with life satisfaction and happiness and that people are happier when they are attending various leisure activities in China.

Michalos and Kahlke (2010) looked at the impact of arts-related activities on perceived quality of life in British Columbia and compared it to a similar study of five B.C. communities in 2006. In both surveys, participation in arts-related activities was correlated with respondents reports of happiness and perceived quality of life measured in several different ways, with attendance at live theater having the strongest correlation.

Lee and Heo (2021) used the 2014 Seoul Survey data to gauge viewer happiness and how often they participated in arts and cultural activities. Findings suggest that both the frequency and diversity of attending arts and cultural activities were positively related to individual happiness. These findings are consistent with the results of previous studies (Brajša-Žganec et al. 2011; Hand 2018; Pagán 2015; Wheatley and Bickerton 2017; Zhang et al. 2017). Study findings suggest that participation in these activities results in aesthetic value, giving individuals psychological satisfaction which leads to pleasure and happiness (McCarthy et al. 2001; Throsby 2001). They found that attendance at movies was extremely high. Attending movies has lower time and financial constraints than other arts and cultural activities, and movies have greater popular appeal and entertainment value than visual arts and performing arts.

Are specific genres related to personal happiness or quality of life? Hefner and Kretz (2021) conducted an experiment to test the relationship between types of romantic beliefs and life satisfaction using film viewing. Participants viewed full-length popular motion pictures and completed a set of measures related to idealistic beliefs and life satisfaction. Results showed that positive portrayals of romance influenced viewers to report greater endorsement of beliefs and stronger life satisfaction; idealistic content led to stronger romantic beliefs and greater life satisfaction than did realistic content. Pannu and Goyal (2024) look at cinematherapy, an innovative therapeutic approach that leverages the emotional and psychological impact of films to aid in the treatment of depression and emotional well-being.

Similar to personal happiness, broader concepts of how things are going also have been related to cultural activities. For example, Pautz and Lumpkin (2020) look at the ability of film to influence its audience about the nature of the “American Dream” and one’s ability to pursue it. We would expect this to be related to people’s feelings about whether the country is headed in the right direction. They recruited audiences to watch two films offering a perspective on the American Dream, Forrest Gump (1994) and Idiocracy (2006), and an additional film (The Italian Job) as a control. Results found that film appears to influence audience attitudes about the American Dream and that a quarter to half of audience members changed their responses to particular questions after the movie. A more positive portrayal in a movie—Forrest Gump—seems to align with a more optimistic audience response.

By contrast, a more negative portrayal—Idiocracy—seems to align with more pessimism. Viewers across all films believed that the United States was headed in the wrong direction. When asked before the movie if they thought the nation was headed in the right or wrong direction, 46 percent indicated wrong direction with 38 percent unsure and the remaining 16 percent said the country was headed in the right direction. Twenty-three percent of participants changed their opinion after watching a movie, with the largest percentage change from viewers of Idiocracy; 34 percent of participants who watched this film changed their opinion; about half were more optimistic about the direction of the nation and the other half were less optimistic. On the whole, after the movie, 46 percent said the nation was headed in the wrong direction, 26 percent said right direction, and 28 percent unsure.

The biggest change involved a decrease in the number of viewers who were unsure. Overall, results suggest that most viewers in the study were not all that confident in the direction that the country is heading and offered only moderate trust in government. After viewing the movies, there was some slight improvement in these assessments, and the biggest change came from viewers of Idiocracy, “which might be explained because that film takes a particularly grim view of the future so perhaps viewers improved their opinions slightly thinking that, in reality, things are not that bad. Attitudes about the direction of the nation and trust in government are likely to be fairly deeply held beliefs and a change in those opinions should not be all that likely after watching a movie for a few hours” (Pautz and Lumpkin 2020, p. 51).

In our pluralistic society, with all of its divisions, do media and leisure opportunities add to those divisions or can they pull people together, providing some common ground? The literature examined suggests that media have strengthened people’s attachment to their communities, but with changes in the media landscape, is that relationship sustained? Moreover, while the media may strengthen ties to community, they also provide news about politics and the economy, fueling attitudes about where the country is headed. We also know from the literature examined that reports of happiness, perceived quality of life available in communities, and cosmopoliteness—which itself tells us something about whether people exhibit a degree of parochialism or broader outlook on life—figure in this mix of relationships. We will treat this set as “criterion variables” examined in our query of whether media use and leisure activities can act as common ground during a time of great divisions and media fragmentation using a series of studies conducted in the past decade. First, we will describe the studies and measures employed.

We examine issues of mainstreaming and cultivation in a series of studies that provide a much broader range of dependent measures than traditionally used. The criterion items encompass not only public opinion items, but also measures of attitudes, values and beliefs, allowing us to see whether the mainstreaming hypothesis extends applies in variegated public opinion domains. In addition, we focus on one issue of particular significance to cultivation—censorship or regulation of TV violence. The conventional guiding perspective is that heavier media consumers are more likely to concur with “mainstream” opinions on current issues (e.g., crime) than are those who watch less TV, read newspapers less frequently, and so forth. We are interested in exploring the extent to which exposure to specific genres and other content types may predict beyond this mainstreaming hypothesis, showing the “scatterstreaming” of audience exposure and effects (Jeffres et al. 2001).

Chaffee and Metzger (2001) suggest that traditional mass-audience perspectives like cultivation reflected a rather anomalous era, as the splintering media audience impel us to reinterrogate our penitential mass-society frameworks in light of an increasingly fragmenting media menu. This raises the issue of whether the potency of classic S-R media dynamics borne of the 20th-century (e.g., cultivation/mainstreaming) will continue to erode, as the larger media environment fragments. We thus pose the following research questions to be considered in light of the increasingly complex digital media environment:

RQ1: To what extent do legacy media, newer digital technologies, and entertainment preferences predict the life attitudes of community attachment (RQ1a), quality of life perceptions (RQ1b), happiness (RQ1c), and cosmopoliteness (RQ1d)?

RQ2: How do genre-specific “taste publics” relate to community attachment (RQ2a) and which genres are most strongly associated with other perceptions of well-being (QOL, happiness, cosmopoliteness, perception of where country is headed) (RQ2b)?

RQ3: Does participation in public leisure activities (e.g., attending concerts, plays) serve as a stronger source of common ground for community attachment than media use alone?

3. Methods

3.1. Four Studies Looking at Film and Genre Perceptions in the Digital Age

We conducted a number of online studies from 2015 to 2024 that allow us to look at some of the relationships between these measures—right track/wrong track perceptions, reports of community attachment, perceptions of the quality of life available in their communities, self-reports of happiness, and several measures of public participation in public leisure options, media use, and attitudes toward media, technology and popular culture—particularly film and TV offerings:

National survey from 2015 of 543 adults;

National survey from 2021 of 321 creators of video content for social media;

Survey from 2022 of 557 young, student creators of video content;

Survey from 2024 of 212 students at a large eastern public university.

3.2. Measures

Social Categories. Typical social category measures were included in all studies: Age, education, income, gender, ethnicity, political party affiliation, and political philosophy (liberal/conservative).

Life Attitude Measures. Five criterion “life attitude” measures were collected across all four surveys. Respondents were asked how much they agreed or disagreed with the following items on 1–7 response scales ranging from completely disagree to completely agree:

• Self-reports of community attachment: “I feel strongly attached to my community.”
• Self-reports of perceived quality of life: “I rate the quality of life in my community very high.”
• Self-reports of happiness: “I generally think of myself as a happy person.”
• Self-reports of cosmopoliteness: “I see myself as a citizen of the world.”
• Self-reports of state of country: “I think the country is headed in the wrong direction.”

Film Genre Viewing. Respondents used a 1–6 response scale (where 1 = never, 2 = almost never, 3 = once in a while; 4 = often, 5 = very often, and 6 = all the time) to tell how often they watched each of the following film genres: Musicals, westerns, horror films, science fiction, detective films, comedy films, film noir films, documentary films, action films animated films, mystery/suspense films, dark comedy films, biographical films, film parodies/spoofs, slasher films, drama films, fantasy films, adventure films, foreign films, romantic dramas, romantic comedies, gangster films, samurai films, epic films, sports films, historical films, super hero films, war films.

Use of the Media and Technologies. Respondents were asked to use the following response scale to indicate their use of legacy media and various newer technologies: several times each day, once a day, several times each week, about once a week, every couple weeks, less often than that, almost never, never.

• Watch television;
• Listen to the radio;
• Read a magazine;
• Read a book;
• Read a newspaper;
• See films in a theater;
• Read a newspaper online;
• Listen to podcasts;
• See films on phone, computer, or tablet;
• Stream movies on home TV;
• Streat TV programs on home TV;
• Surf internet for pleasure;
• Check my emails;
• Go on Facebook;
• Post videos on Facebook;
• Post photos on Facebook;
• Browse YouTube for videos to watch;
• Browse TikTok for videos to watch;
• Play video games on some device;
• Watch videos on smartphone;
• Text family or friends rather than call on phone;
• See live musical concerts (2015 survey; 2021 survey);
• See live plays in theater.

In the 2021, 2022, and 2024 studies, we also measured respondents’ streaming behaviors and use of newer media technologies: items asked respondents how often they stream a movie on home TV, stream a TV program on home TV, and watch a TV program on a tablet, computer, or cell phone (all measured on a 7-point response scale from “never” to “more than several times a week”), and the summative number of streaming services they report using (from a roster of 29 services).

Scales Measuring Traditionalist Preferences, Technology Enthusiasm, and Mobility/Social Media Enthusiasm (2015 national survey; 2021 creator survey; 2024 student survey). Scales were constructed using the items listed. Alphas here refer to the 2015 study.

• Traditionalist Preferences Scale. (alpha = 0.61)

• I’m more a traditionalist, preferring to read physical copies of books, magazines and newspapers.
• I’d still rather talk to people over the phone than text.
• I think that the new technologies have begun to dominate our lives, occupying too much of our time.
• I like the variety of entertainment available today but sometimes feel it’s too much.

• Technology Enthusiasm Scale. (alpha = 0.70)

• I love the options at my finger-tips today, watching videos on my phone, texting in a car, streaming films.
• I can hardly wait to see what technology comes next.
• I think I’m getting less patient and am glad I have a smart phone or other digital options to fill the time.

• Mobility/Social Media Enthusiasm Scale. (alpha = 0.70)

• I often watch videos on my cell phone.
• I often search for videos on YouTube to watch.
• I often share videos via Facebook.
• I often share videos on Instagram.

4. Results

Per the research questions, we compared community attachment (RQ1a), quality-of-life (QOL) perceptions (RQ1b), perceived happiness (RQ1c), and cosmopoliteness (RQ1d) across the surveys. As

Table 1. Descriptive statistics for key variables and correlations with social categories in 4 surveys.

	2015 National Survey of General Population (n = 364)	2021 National Survey of Content Creators (n = 321)	2022 Survey of Young, Student Creators (n = 557)	2024 Survey of Students (n = 212)
Attachment to community Mean Std deviation	4.13 1.7	5.0 1.5	4.5 1.4	4.54 1.3
Correlations with social categories	Age (0.12 *) Income (0.16 **)	Educ (0.13 *) Dem (0.13 *) Indep (−0.18 ***) Pol ph (−0.12 *)	Income (0.12 **) Rep (0.15 ***) Pol ph (−0.14 **)	Educ (−0.24 ***) Pol ph (−0.16 *) Income (0.15 *)
Perceived QOL Mean Std deviation	4.58 1.6	5.21 1.3	4.82 1.3	4.67 1.2
Correlations with social categories	Age (0.09 #) Educ (0.10 *) Income (0.26 ***)	Age (−0.16 **) Educ (0.13 *) Dem (0.11 *) Indep (−0.15 **)	Educ (0.10 *) Income (0.14 **) White (0.18 ***) Rep (0.09 *) Pol ph (−0.08 #)	Educ (−0.17 *) White (0.16 *) Income (0.17 *)
Perceived happiness Mean Std deviation	5.22 1.6	5.36 1.4	5.04 1.4	4.99 1.4
Correlations with social categories	Age (0.10 #) Educ (0.15 **) Income (0.21 ***)	Educ (0.13 *) Income (0.14 *) Dem (0.16 **) Indep (−0.11 *)	Income (0.14 ***) White (0.12 **) Rep (0.11 *) Pol ph (−0.09 *)	Age (−0.19 **) Educ (−0.19 **) Pol ph (−0.22 **) Dem (−0.14 *)
Cosmopoliteness Mean Std deviation	4.7 1.8	5.3 1.5	4.9 1.4	4.98 1.3
Correlations with social categories	Age (−0.14 **) Income (−0.09 #)	Educ (0.12 *) Dem (0.15 **) Rep (−0.12 *)	None sig.	Educ (−0.20 **)
Country headed in wrong direction Mean Std deviation	4.38 1.7	4.97 1.6	4.60 1.3	4.53 1.4
Correlations with social categories	Educ (−0.12 *)	Educ (0.10 #) Income (0.12 *) Dem (−0.21 ***) Rep (0.20 ***) Pol ph (−0.17 **)	Rep (0.09 *) Pol ph (−0.13 **)	Rep (0.14 *) Income (0.15 *)

Note. On 1–7 scale, 1 = strongly disagree, 4 = neither agree nor disagree, 7 = strongly agree; Self-report of attachment to community: “I feel strongly attached to my community”. Self-report of perceived quality of life: “I rate the quality of life in my community very high”. Self-report of happiness: “I generally think of myself as a happy person”. Self-report of state of country: “I think the country is headed in the wrong direction”. Cosmopoliteness: “I see myself as a citizen of the world”. # 0.05 < p < 0.10, * p < 0.05, ** p < 0.01, *** p < 0.001.

shows, there is a relatively similar pattern across all four surveys, with slightly positive ratings of the quality of life (RQ1a) in their communities (4 = neutral, so all are in the positive direction) and cosmopoliteness (all in positive direction), an even more positive pattern of perceived happiness (all in the positive direction of ~5 or greater), and a slightly negative pattern on what direction the country is headed (all means greater than 4, agreeing that it is headed in the wrong direction). So, the national measure—country headed in wrong direction—diverges from the more positive assessments of community attachment, QOL perceptions, personal happiness, and cosmopoliteness. This finding holds from 2015 through 2024, a time period marked by great political, social, and public health changes and challenges. The negative finding of where the country is headed is consistent with that found by Pautz and Lumpkin (2020), who note that it is consistent with widespread surveys of public opinion.

We also examined relationships between our criterion variables and social categories. Context is important for understanding the impact of social and political variables. In 2015, Barack Obama was president of the United States, the unemployment rate was 5.0 percent, and inflation was below 2 percent. Donald Trump was elected president in 2016. In 2020, the Covid pandemic hit the United States, and the government ended the public health emergency in May, 2023. The actual unemployment rate averaged 13 percent in the second quarter of 2020 and then fell sharply, averaging 4.2 percent in the fourth quarter of 2021. COVID-19 inflation was a supply shock as inflation surged. Average inflation in 2022 reached 8 percent. Donald Trump was president in 2016, and Joe Biden was elected in 2020.

Significant correlations are found in Table 1. The relationships are consistent with the literature, so that older people and those with more education and higher incomes tend to be more attached to their communities, think the quality of life available is higher, are happier, and see themselves as more cosmopolite, though there are differences across the different surveys. Partisan identification and political philosophy appear to be significant in some instances. Since our interest is in the relationship between our criterion measures and media/leisure activities, we will hold the social categories constant for a clearer look at the relationship. Fewer measures of social categories were obtained in the 2015 study than in subsequent surveys.

We expected our measure of community attachment, QOL perceptions, and several other criterion variables would be interrelated. As

Table 2. Correlations among community attachment, perceived QOL, happiness, cosmopoliteness and country headed in wrong direction.

	Community Attachment	Perceived QOL	Happiness	Cosmopoliteness
Perceived QOL	0.58 *** (2015) 0.55 *** (2021) 0.36 *** (2022) 0.51 *** (2024)
Happiness	0.52 *** (2015) 0.47 *** (2021) 0.50 *** (2022) 0.50 *** (2024)	0.54 *** (2015) 0.45 *** (2021) 0.45 *** (2022) 0.49 *** (2024)
Cosmopoliteness	0.17 *** (2015) 0.36 *** (2021) 0.26 *** (2022) 0.43 *** (2024)	0.12 * (2015) 0.30 *** (2021) 0.19 *** (2022) 0.30 *** (2024)	0.14 ** (2015) 0.39 *** (2021) 0.25 *** (2022) 0.36 *** (2024)
Country headed in wrong direction	−0.01 (2015) 0.15 ** (2021) 0.04 (2022) −0.02 (2024)	−0.05 (2015) 0.07 (2021) 0.04 (2022) −0.07 (2024)	−0.05 (2015) 0.05 (2021) 0.05 (2022) 0.04 (2024)	−0.02 (2015) 0.12 * (2021) 0.02 (2022) −0.03 (2024)

Note. 2015 = national survey of adults; 2021 = national survey of content creators; 2022 = survey of young, student creators; 2024 = survey of students. * p < 0.05, ** p < 0.01, *** p < 0.001.

shows, there is a strong pattern of positive correlations between community attachment, perceived quality of life available in the community, self-reports of happiness, and cosmopoliteness, but reports of the state of the country are not related to much of anything, and the pattern is clear across all of the studies.

4.1. Legacy Media Use and Criterion Variables

First, we examined legacy media use and our criterion variables. As noted in our literature review, newspaper use earlier was positively related to community attachment. At the same time, Putnam (2000) indicated television as a debilitating factor for social capital. Since social categories are related to not only our criterion variables but also use of legacy media, we examined the relationships controlling for them. As

Table 3. Partial correlations between use of legacy media and criterion variables—community attachment, QOL perceptions, reports of happiness, cosmopoliteness and beliefs the country is headed in the wrong direction—controlling for social categories.

	Community Attachment	Perceived QOL	Happiness	Cosmopoliteness	Country Headed in Wrong Direction
TV viewing	(2015) 0.10 * (2021) 0.31 * (2022) 0.05 (2024) 0.05	(2015) 0.04 (2021) 0.26 * (2022) 0.10 * (2024) 0.02	(2015) 0.07 (2021) 0.19 * (2022) −0.01 (2024) −0.01	(2015) 0.03 (2021) 0.21 * (2022) 0.06 (2024) 0.10	(2015) −0.08 (2021) 0.12 * (2022) −0.05 (2024) −0.13 #
Radio listening	(2015) 0.08 (2021) 0.27 * (2022) 0.08 # (2024) 0.17 *	(2015) 0.08 (2021) 0.19 * (2022) −0.01 (2024) 0.19 **	(2015) 0.02 (2021) 0.26 * (2022) 0.00 (2024) 0.20 **	(2015) −0.03 (2021) 0.13 * (2022) 0.08 # (2024) 0.17 *	(2015) 0.02 (2021) 0.19 * (2022) −0.09 # (2024) 0.04
Magazine reading	(2015) 0.13 * (2021) 0.36 * (2022) 0.13 * (2024) 0.02	(2015) 0.00 (2021) 0.22 * (2022) 0.08 (2024) −0.08	(2015) 0.00 (2021) 0.18 * (2022) 0.02 (2024) −0.05	(2015) 0.03 (2021) 0.32 * (2022) 0.01 (2024) 0.09	(2015) −0.05 (2021) 0.18 * (2022) 0.00 (2024) −0.07
Reading books	(2015) 0.01 (2021) 0.27 * (2022) 0.08 (2024) −0.06	(2015) 0.02 (2021) 0.16 * (2022) 0.04 (2024) −0.05	(2015) 0.00 (2021) 0.12 * (2022) 0.02 (2024) −0.13 #	(2015) 0.04 (2021) 0.24 * (2022) 0.02 (2024) −0.02	(2015) 0.03 (2021) 0.17 * (2022) −0.04 (2024) −0.04
Reading newspapers (physical copies)	(2015) 0.18 * (2021) 0.35 * (2022) 0.09 # (2024) −0.02	(2015) 0.02 (2021) 0.19 * (2022) 0.03 (2024) −0.02	(2015) 0.02 (2021) 0.18 * (2022) −0.01 (2024) −0.14 *	(2015) 0.05 (2021) 0.22 * (2022) 0.08 # (2024) 0.01	(2015) −0.07 (2021) 0.18 * (2022) 0.01 (2024) −0.10
Go out to see film in a theater	(2015) 0.11 * (2021) 0.36 * (2022) 0.05 (2024) −0.02	(2015) 0.02 (2021) 0.36 * (2022) 0.06 (2024) 0.08	(2015) 0.03 (2021) 0.15 * (2022) 0.00 (2024) 0.08	(2015) 0.02 (2021) 0.29 * (2022) 0.06 (2024) 0.06	(2015) −0.19 * (2021b 0.23 * (2022) −0.05 (2024) −0.05
Communication Traditionalist Scale	(2015) 0.11 * (2024) 0.08	(2015) 0.11 * (2024) 0.04	(2015) 0.04 (2024) 0.10	(2015) 0.06 (2024) 0.07	(2015) 0.34 * (2024) 0.07

Note. 2015 = national survey of adults; 2021 = national survey of content creators; 2022 = survey of young, student creators; 2024 = survey of students. Note. Social categories controlled by study: (2015) age, education, income and gender; (2021 creators) age, education, income, gender, white ethnicity, Democratic, Republican and Independent affiliations, political philosophy; (2022 students) age, education, income, gender, white ethnicity, Democratic, Republican and Independent affiliations, political philosophy. # 0.05 < p < 0.10, * p < 0.05, ** p < 0.01.

shows, though there are differences by study, in general, there is a strong pattern of positive relationships between the use of legacy media and our criterion variables controlling for social categories. The scale tapping communication traditionalism also is positively related to our criterion variables.

Next, we examined the relationships between our criterion variables and use of more modern technologies. As

Table 4. Partial correlations between use of newer technologies and media and criterion variables controlling for social categories.

	Community Attachment	Perceived QOL	Happiness	Cosmopolite- Ness	Country Headed in Wrong Direction
Read a newspaper online	(2021) 0.23 * (2022) 0.04 (2024) 0.01	(2021) 0.08 (2022) 0.02 (2024) 0.04	(2021) 0.10 # (2022) −0.05 (2024) −0.08	(2021) 0.23 * (2022) −0.01 (2024) 0.04	(2021) 0.13 * (2022) 0.07 (2024) 0.00
Listen to podcasts	(2021) 0.26 * (2022) 0.10 * (2024) 0.06	(2021) 0.26 * (2022) 0.10 * (2024) 0.05	(2021) 0.21 * (2022) 0.08 (2024) 0.03	(2021) 0.22 * (2022) −0.02 (2024) 0.05	(2021) 0.20 * (2022) 0.02 (2024) 0.06
Watch a film on tablet, computer, or cell phone	(2015) 0.02 (2021) 0.31 * (2024) 0.04	(2015) 0.04 (2021) 0.19 * (2024) 0.01	(2015) 0.01 (2021) 0.23 * (2024) −0.03	(2015) 0.08 (2021) 0.24 * (2024) 0.05	(2015) 0.04 (2021) 0.13 * (2024) −0.12 #
Watch a TV program on tablet, computer, or cell phone	(2021) 0.27 * (2024) 0.06	(2021) 0.18 (2024) 0.03	(2021) 0.23 * (2024) −0.02	(2021) 0.20 * (2024) 0.12 #	(2021) 0.12 * (2024) −0.10
Stream a movie on home TV	(2021) 0.28 * (2024) −0.02	(2021) 0.24 * (2024) −0.04	(2021) 0.19 * (2024) −0.05	(2021) 0.27 * (2024) 0.12 #	(2021) 0.05 (2024) −0.12 #
Stream a TV program on home TV	(2021) 0.27 * (2024) 0.07	(2021) 0.21 * (2024) 0.05	(2021) 0.18 * (2024) 0.03	(2021) 0.27 * (2024) 0.15 *	(2021) 0.04 (2024) −0.10
Surf the internet for pleasure	(2015) 0.01 (2021) −0.01 (2022) −0.05 (2024) 0.11	(2015) 0.05 (2021) 0.02 (2022) 0.06 (2024) 0.03	(2015) 0.02 (2021) 0.10 # (2022) −0.04 (2024) 0.10	(2015) 0.04 (2021) 0.05 (2022) 0.01 (2024) 0.14 *	(2015) −0.09 # (2021) 0.02 (2022) −0.02 (2024) −0.01
Check email	(2015) 0.01 (2021) 0.03 (2022) 0.02 (2024) 0.15 *	(2015) 0.06 (2021) 0.12 * (2022) 0.05 (2024) 0.07	(2015) 0.13 * (2021) 0.13 * (2022) 0.07 (2024) 0.00	(2015) 0.00 (2021) 0.13 * (2022) 0.02 (2024) 0.14 *	(2015) −0.04 (2021) 0.08 (2022) −0.00 (2024) 0.00
Go on Facebook	(2015) 0.05 (2021) 0.29 * (2022) 0.08 # (2024) 0.05	(2015) 0.01 (2021) 0.20 * (2022) −0.02 (2024) −0.04	(2015) −0.01 (2021) 0.20 * (2022) 0.07 (2024) 0.07	(2015) −0.04 (2021) 0.20 * (2022) 0.08 (2024) −0.02	(2015) −0.07 (2021) 0.04 (2022) −0.02 (2024) 0.14 *
Post videos on Facebook	(2021) 0.33 * (2022) 0.04 (2024) 0.12 #	(2021) 0.25 * (2022) −0.01 (2024) −0.09	(2021) 0.18 * (2022) 0.01 (2024) 0.05	(2021) 0.24 * (2022) 0.01 (2024) 0.06	(2021) 0.14 * (2022) −0.05 (2024) 0.05
Post photos on Facebook	(2021) 0.33 (2022) 0.04 (2024) 0.20 **	(2021) 0.22 (2022) 0.02 (2024) −0.04	(2021) 0.16 * (2022) 0.01 (2024) 0.06	(2021) 0.18 * (2022) 0.02 (2024) 0.09	(2021) 0.17 * (2022) −0.05 (2024) 0.08
Browse YouTube for videos to watch	(2021) 0.20 * (2022) −0.02 (2024) 0.01	(2021) 0.17 * (2022) 0.04 (2024) 0.05	(2021) 0.14 * (2022) 0.05 (2024) 0.04	(2021) 0.18 * (2022) 0.05 (2024) 0.04	(2021) 0.19 * (2022) 0.03 (2024) 0.05
Browse TikTok for videos to watch	(2021) 0.29 * (2022) 0.12 * (2024) 0.15 *	(2021) 0.19 * (2022) 0.06 (2024) 0.10	(2021) 0.19 * (2022) 0.05 (2024) 0.00	(2021) 0.14 * (2022) 0.07 (2024) 0.08	(2021) 0.10 # (2022) −0.06 (2024) −0.06
Play video games on some device	(2015) −0.04 (2021) 0.24 * (2022) −0.05 (2024) −0.03	(2015) −0.07 (2021) 0.22 * (2022) 0.04 (2024) 0.05	(2015) −0.06 (2021) 0.18 * (2022) 0.01 (2024) 0.11	(2015) 0.04 (2021) 0.30 * (2022) 0.05 (2024) 0.01	(2015) −0.02 (2021) 0.12 * (2022) −0.05 (2024) −0.02
Watch videos on a smart phone	(2015) 0.09 # (2021) 0.28 * (2022) 0.04 (2024) 0.03	(2015) 0.07 (2021) 0.27 * (2022) 0.13 * (2024) 0.15 *	(2015) −0.03 (2021) 0.24 * (2022) 0.12 * (2024) 0.16 *	(2015) 0.00 (2021) 0.21 * (2022) 0.10 * (2024) 0.04	(2015) −0.02 (2021) 0.12 * (2022) −0.03 (2024) −0.03
Text family or friends rather than call them on the phone	(2015) 0.15 * (2021) 0.21 * (2022) 0.04 (2024) 0.11	(2015) 0.16 * (2021) 0.21 * (2022) 0.11 * (2024) 0.10	(2015) 0.15 * (2021) 0.18 * (2022) 0.10 * (2024) 0.03	(2015) 0.04 (2021) 0.18 * (2022) 0.10 * (2024) 0.14 *	(2015) −0.06 (2021) 0.10 # (2022) 0.06 (2024) 0.02
Technology Enthusiasm Scale	(2021) 0.38 * (2024) 0.10	(2021) 0.38 * (2024) 0.19 **	(2021) 0.40 * (2024) 0.15 *	(2021) 0.31 * (2024) 0.11	(2021) 0.14 * (2024) 0.02
Mobility/Social Media Enthusiasm Scale	(2015) 0.15 **	(2015) 0.14 *	(2015) 0.02	(2015) 0.16 *	(2015) 0.10 #

Note. 2015 = national survey of adults; 2021 = national survey of content creators; 2022 = survey of young, student creators; 2024 = survey of students. Note. Social categories controlled by study: (2015) age, education, income and gender; (2021 creators) age, education, income, gender, white ethnicity, Democratic, Republican and Independent affiliations, political philosophy; (2022 students) age, education, income, gender, white ethnicity, Democratic, Republican and Independent affiliations, political philosophy. # 0.05 < p < 0.10, * p < 0.05, ** p < 0.01.

shows, again, we find a pattern with many positive partial correlations controlling for social categories, though fewer significant relationships are found for perceptions that the country is headed in the wrong direction.

4.2. Do Entertainment and Leisure Variables Make a Difference?

The studies analyzed here focused on film genres, and in one instance included going out for live entertainment experiences. Respondents were asked how often they watched a series of film genres, which were then factor analyzed to identify taste publics. Respondents in the four surveys were asked how often they saw each of 31 film genres on a six-point response scale. These were then factor analyzed (Varimax rotation; principal components) and factor scores were retained for additional analyses (see Atkin et al. 2025; Jeffres et al. 2024). Though the four samples cross different groups and different periods of time, as well as a general population and creators, we found several genre taste publics repeating themselves, with slight variations. This suggests some genre taste publics endure regardless of the environment and individual differences. Taste publics labeled Romance & Drama, Big Screen Fantasy, and Intellectual Escape appear in all four studies. Sports & Conflict and History Buff appear in three of the four studies, and a unique, notable film genre taste pattern identified as Omnibus, Classic emerged in two studies—the national creators sample and the 2024 student sample. Several other unique patterns emerged in individual studies.

Per RQ2a, we examined relationships between the genre taste publics and our criterion variables. We also looked at relationships involving attendance at live entertainment events. As

Table 5. Partial correlations between criterion variables and genre taste publics and attending live entertainment activities controlling for social categories.

	Community Attachment	Perceived QOL	Happiness	Cosmopolite- Ness	Country Headed in Wrong Direction
Attending Live Entertainment Events
Going to see live plays performed in a theater	(2015) 0.14 *	(2015) 0.02	(2015) −0.04	(2015) 0.01	(2015) −0.09 #
Going to see live musical concerts/events	(2015) 0.19 *	(2015) 0.03	(2015) 0.00	(2015) 0.01	(2015) −0.12 *
Genre Taste Publics
Romance & Drama	(2015) 0.25 * (2021) 0.24 * (2022) 0.10 * (2024) 0.13 #	(2015) 0.15 * (2021) 0.16 * (2022) 0.04 (2024) 0.03	(2015) 015 * (2021) 0.14 * (2022) −0.00 (2024) 0.09	(2015) 0.11 * (2021) 0.09 (2022) 0.04 (2024) 0.20 **	(2015) 0.06 (2021) 0.10 # (2022) −0.03 (2024) 0.02
Big Screen Fantasy	(2015) −0.08 (2021) 0.17 * (2022) 0.04 (2024) −0.02	(2015) 0.11 * (2021) 0.22 * (2022) 0.09 # (2024) 0.05	(2015) 0.11 * (2021) 0.09 # (2022) 0.01 (2024) 0.05	(2015) 0.15 * (2021) 0.14 * (2022) 0.13 * (2024) 0.09	(2015) 0.04 (2021) 0.06 (2022) 0.03 (2024) −0.16 *
Intellectual Escape	(2015) 0.07 (2021) 0.14 * (2022) 0.06 (2024) 0.04	(2015) −0.03 (2021) 0.14 (2022) 0.03 (2024) 0.09	(2015) −0.03 (2021) 0.00 (2022) −0.05 (2024) 0.02	(2015) 0.12 * (2021) 0.04 (2022) 0.01 (2024) 0.07	(2015) −0.08 (2021) 0.17 * (2022) −0.03 (2024) 0.04
Sports & Conflict	(2015) 0.05 (2022) 0.14 * (2024) 0.19 **	(2015) −0.03 (2022) 0.07 (2024) 0.18 **	(2015) −0.03 (2022) 0.10 * (2024) 0.14 *	(2015) −0.06 (2022) 0.19 * (2024) −0.00	(2015) 0.02 (2022) 0.04 (2024) −0.00
History Buff	(2015) 0.12 * (2022) 0.11 * (2024) 0.00	(2015) −0.00 (2022) −0.01 (2024) 0.02	(2015) −0.00 (2022) 0.06 (2024) −0.06	(2015) 0.10 # (2022) 0.06 (2024) 0.08	(2015) −0.07 (2022) 0.10 * (2024) 0.00
Omnibus, Classic	(2021) 0.30 * (2024) −0.12 #	(2021) 0.09 (2024) −0.13 #	(2021) 0.23 * (2024) −0.08	(2021) 0.33 * (2024) −0.02	(2021) 0.08 (2024) −0.08
Mystery/Suspense & Action	(2015) 0.15 *		(2015) 0.24 *	(2015) 0.21 *	(2015) 0.09 #
Foreign & Film Noir	(2015) 0.03	(2015) −0.13 *	(2015) −0.13 *	(2015) 0.20 *	(2015) 0.11 *
Family-Oriented	(2015) 0.11 *	(2015) 0.14 *	(2015) 0.14 *	(2015) 0.05	(2015) −0.04
Documentary & Mystery/ Suspense	(2021) 0.12 *	(2021) 0.18 *	(2021) 0.17 *	(2021) 0.31 *	(2021) 0.11 #
Conflict & Weaponry	(2022) 0.01	(2022) −0.08	(2022) −0.10 *	(2022) 0.00	(2022) −0.01
Horror & Mystery	(2022) 0.04 (2024) −0.02	(2022) −0.02 (2024) −0.05	(2022) −0.04 (2024) −0.06	(2022) 0.07 (2024) −0.01	(2022) −0.05 (2024) −0.07

Note. 2015 = national survey of adults; 2021 = national survey of content creators; 2022 = survey of young, student creators; 2024 = survey of students. Note. Social categories controlled by study: (2015) age, education, income and gender; (2021 creators) age, education, income, gender, white ethnicity, Democratic, Republican and Independent affiliations, political philosophy; (2022 students) age, education, income, gender, white ethnicity, Democratic, Republican and Independent affiliations, political philosophy. # 0.05 < p < 0.10, * p < 0.05, ** p < 0.01.

shows, there are numerous significant relationships between genre taste publics and our criterion variables. The strongest pattern is found for community attachment, which is related to both measures of attendance at live entertainment events—concerts and plays—with 15 partial correlations statistically significant across the different genre taste publics. Per RQ2b, numerous partial correlations between genre taste publics and QOL perceptions (10), reports of happiness (12), cosmopoliteness (12), and beliefs the country is headed in the wrong direction (9) are statistically significant or approach significance.

4.3. Regression Models

We conducted regressions to see whether the genre taste publics and live entertainment variables had any impact on our criterion variables with social categories, use of legacy media, and use of newer media and technologies in the model. See

Table 6. Predicting criterion variables for 2015 study (general population).

Criterion Variables	Blocks of Predictors	R Sq Change	F Change
Community attachment	Social Categories Age (β = 0.11, p = 0.04), Income (β = 0.12. p = 0.02)	0.036	3.4, p < 0.01
	Legacy media use Newspaper reading (β = 0.16, p < 0.01)	0.042	3.2, p < 0.01
	Use of newer media & technologies Play video games (β = 0.11, p = 06), texting rather than phone call (β = 0.12, p < 0.05)	0.023	1.5, ns
	Genre taste publics + Go to live musical concerts (β = 0.19, p = 0.02) Romance & drama taste (β = 0.22, p < 0.001) History buff (β = 0.12, p < 0.02), Mystery, suspense & action (β = 0.16, p < 01), Family oriented (β = 0.09, p = 0.07)	0.105	3.7, p < 0.001
	Model R Sq = 0.21, F = 7.7(27,335), p < 0.001
Perceived QOL	Social Categories Income (β = 0.24, p < 0.001)	0.069	6.6, p < 0.001
	Legacy media use	0.008	0.59, ns
	Use of newer media & technologies Play video games (β = 0.12, p = 0.04) Texting rather than phone call (β = 0.17, p < 0.01)	0.034	2.2, p = 0.04
	Genre taste publics + History buff (β = 0.11, p = 0.05) Mystery, suspense & action (β = 0.18, p < 0.001)	0.053	1.8, p < 0.05
	Model R Sq = 0.16, F = 2.4, p < 0.001
Happiness	Social Categories Income (β = 0.17, p = 0.003)	0.061	5.8, p < 0.001
	Legacy media use	0.006	0.42, ns
	Use of newer media & technologies Check email (β = 0.13, p = 0.04) Texting rather than phone call (β = 0.20, p < 0.001)	0.047	3.1, p = 0.006
	Genre taste publics + Romance & drama (β = 0.15, p = 0.006) Big screen fantasy (β = 0.10, p = 0.056) Mystery/Suspense & Action (β = 0.24, p < 0.001) Foreign & film noir (β = −0.14, p = 0.013) Family oriented (β = 0.14, p = 0.004)	0.115	4.2, p < 0.001
	Model R Sq = 0.23, F = 3.7, p < 0.001
Cosmopoliteness	Social Categories Age (β = −0.13, p = 0.018)	0.025	2.3, p = 0.062
	Legacy media use None sig.	0.005	0.38, ns
	Use of newer media & technologies None sig.	0.008	0.46, ns
	Genre taste publics + Romance & drama (β = 0.12, p = 0.038) Big screen flash & fantasy (β = 0.16, p = 0.004) Intellectual escape (β = 0.11, p = 0.051) History buff (β = 0.11, p = 0.031) Mystery/suspense & Acton (β = 0.20, p < 0.001) Foreign & film noir (β = 0.21, p < 0.001)	0.137	4.6, p < 0.001
	Model R Sq = 0.18, F = 2.6, p < 0.001
Country headed in wrong direction	Social Categories Education (β = −0.13, p = 0.021)	0.023	2.1, p = 0.08
	Legacy media use	0.016	1.2, ns
	Use of newer media & technologies	0.010	0.62, ns
	Genre taste publics + Go out to see film (β = 0.22, p < 0.001) Romance & drama (β = 0.12, p < 0.05) Foreign & film noir (β = 0.10, p = 0.08)	0.086	2.8, p < 0.001
	Model R Sq = 0.14, F =1.9, p = 0.004

,

Table 7. Predicting criterion variables for 2021 creator study.

Criterion Variables	Blocks of Predictors	R Sq Change	F Change
Community attachment	Social Categories Demo party (β = 0.30, p = 0.002), Rep party (β = 0.26, p = 0.005)	0.103	4.0, p < 0.001
	Legacy media use Watch TV (β = 0.20, p < 0.001) Listen to radio (β = 0.10, p = 0.08) Read newspaper (hard copy) (β = 0.18, p = 0.04)	0.183	11.2, p < 0.001
	Use of newer media & technologies Go on Facebook (β = 0.16, p = 0.008) Watch videos on phone (β = 0.13, p < 0.05)	0.063	2.4, p < 0.001
	Genre taste publics + Omnibus, classic (β = 0.28, p < 0.001) Big screen action & fantasy (β = 0.15, p = 0.02) Intellectual escape (β = 0.20, p < 0.01) Romance & drama (β = 0.15, p = 0.02)	0.035	2.7, p < 0.001
	Model R Sq = 0.38, F = 5.3, p < 0.001
Perceived QOL	Social Categories Age (β = −0.16, p < 0.004, Demo party (β = 0.23, p = 0.02) Rep party (β = 0.18, p = 0.04)	0.087	3.3, p < 0.001
	Legacy media use Watch TV (β = 0.21, p < 0.001) Listen to podcasts (β = 0.18, p = 0.005) Read newspapers online (β = −0.14, p < 0.05)	0.121	6.7, p < 0.001
	Use of newer media & technologies Watch videos on smart phone (β = 0.15, p = 0.04)	0.053	1.7, p = 0.06
	Genre taste publics + Omnibus, classic (β = 0.14, p < 0.10) Big screen action & fantasy (β = 0.19, p = 0.006) Intellectual escape (β = 0.24, p = 0.003) Romance & drama (β = 0.11, p = 0.09)	0.036	2.4, p = 0.03
	Model R Sq = 0.30, F = 3.6(34, 288), p < 0.001
Happiness	Social Categories Demo party (β = 0.32, p < 0.001) Rep party (β = 0.20, p = 0.03)	0.081	3.1, p = 0.002
	Legacy media use Watch TV (β = 0.11, p = 0.06) Listen to radio (β = 0.19, p = 0.002) Listen to podcasts (β = 0.12, p = 0.052)	0.094	5.0, p < 0.001
	Use of newer media & technologies Go on Facebook (β = 0.12, p = 0.06)	0.046	1.5, ns
	Genre taste publics + Go out to see films (β = −0.20, p = 0.04) Omnibus, classic (β = 0.34, p < 0.001) Big screen action & fantasy (β = 0.13, p = 0.067) Intellectual escape (β = 0.14, p = 0.087) Romance & drama (β = 0.14, p = 0.036) Documentary & mystery/suspense (β = 0.13, p = 0.04)	0.048	3.2, p = 0.005
	Model R Sq = 0.27, F = 3.1, p < 0.001
Cosmopoliteness	Social Categories Education (β= 0.13, p = 0.023) Demo party (β = 0.20, p = 0.043)	0.052	1.9, p < 0.05
	Legacy media use Watch TV (β = 0.12, p = 0.04) Read magazines (β = 0.26, p = 0.005)	0.123	6.5, p < 0.001
	Use of newer media & technologies Post photos on Facebook (β = −0.20, p = 0.07) Play video games (β = 0.16, p = 0.02)		1.8, p = 0.045
	Genre taste publics + Omnibus, classic (β = 0.46, p < 0.001) Big screen action & fantasy (β = 0.15, p = 0.02) Intellectual escape (β = 0.18, p = 0.02) Romance & drama (β = 0.11, p = 0.07) Documentary & mystery/suspense (β = 0.28, p < 0.001)	0.101	7.3, p < 0.001
	Model R Sq = 0.33, F = 4.2, p < 0.001
Country headed in wrong direction	Social Categories None sig.	0.084	3.2, p < 0.001
	Legacy media use Listen to radio (β = 0.11, p = 0.087) Listen to podcasts (β = 0.12, p = 0.075)	0.062	3.2, p = 0.003
	Use of newer media & technologies Browse YouTube for videos to watch (β = 0.12, p = 0.087) Watch videos on smart phone (β = 0.19, p = 0.013)	0.053	1.6, p = 0.09
	Genre taste publics + Intellectual escape (β = 0.18, p = 0.03)	0.029	1.8, p < 0.10
	Model R Sq = 0.23, F = 2.48(34,288), p < 0.001

,

Table 8. Predicting criterion variables for 2022 student study.

Criterion Variables	Blocks of Predictors	R Sq Change	F Change
Community commitment	Social Categories Income (β = 0.10, p = 0.051) Pol philosophy (β = −0.14, p = 0.048)	0.047	2.4, p < 0.01
	Legacy media use None sig.	0.026	1.7, p < 0.10
	Use of newer media & technologies Browse TikTok for videos to watch (β = 0.14, p < 0.01)	0.040	1.2, ns
	Genre taste publics + Sports & conflict (β = 0.13, p = 0.016) History buff (β = 0.09, p = 0.076)	0.025	1.5, ns
	Model R Sq = 0.14, F = 1.7(40, 412), p = 0.009
Perceived QOL	Social Categories Education (β = 0.09, p = 0.068) White ethnicity (β = 0.13, p = 0.012)	0.055	2.9, p = 0.002
	Legacy media use Watch TV (β = 0.10, p = 0.049)	0.022	1.5, ns
	Use of newer media & technologies Watch videos on smart phone (β = 0.09, p < 0.10) Watch TV program on some device (β = 0.13, p = 0.062)	0.041	1.2, ns
	Genre taste publics + None sig.	0.010	0.57, ns
	Model R Sq = 0.13, F = 1.5(40,412), p = 0.02
Happiness	Social Categories Income (β = 0.11, p = 0.022)	0.041	2.1, p = 0.03
	Legacy media use None sig.	0.01	0.65, ns
	Use of newer media & technologies Surf internet for pleasure (β = −0.09, p = 0.09) Watch videos on smart phone (β = 0.13, p = 0.025)	0.047	1.4, ns
	Genre taste publics + Sports & conflict (β = 0.12, p = 0.032)	0.024	1.4, ns
	Model R Sq = 0.12, F = 1.44(40,412), p = 0.049
Cosmopoliteness	Social Categories None sig.	0.006	0.29, ns
	Legacy media use Read a newspaper (hard copy) (β = 0.10, p = 0.096)	0.017	1.1, ns
	Use of newer media & technologies Stream a film on home TV (β = 0.13, p = 0.079	0.047	1.3, ns
	Genre taste publics + Big screen action & fantasy (β = 0.12, p = 0.024) Sports & conflict (β = 0.21, p < 0.001)	0.047	2.7, p = 0.007
	Model R Sq = 0.12, F = 1.3, p = 0.082
Country headed in wrong direction	Social Categories Pol. Philosophy (β = −0.13, p = 0.059)	0.026	1.3, ns
	Legacy media use Listen to the radio (β = −0.10, p = 0.054) Read a newspaper online (β = 0.10, p = 0.068)	0.019	1.2, ns
	Use of newer media & technologies Text family rather than phone call (β = 0.12, p = 0.03) Watch TV programs on some device (β = −0.15, p = 0.04)	0.039	1.1, ns
	Genre taste publics + None sig.	0.017	1.0, ns
	Model R Sq = 0.10, F = 1.2(40,412), ns

and

Table 9. Predicting criterion variables for 2024 student study.

Criterion Variables	Blocks of Predictors	R Sq Change	F Change
Community commitment	Social Categories Education (β = −0.22, p = 0.006) Pol philosophy (β = −0.24, p = 0.076) Income (β = 0.15, p = 0.063)	0.120	3.0, p < 0.01
	Legacy media use Listen to radio (β = 0.16, p = 0.053)	0.036	0.9, ns
	Use of newer media & technologies Check email (β = 0.15, p = 0.057) Post photos on Facebook (β = 0.28, p = 0.042)	0.100	1.5, ns
	Genre taste publics + None sig.	0.031	0.9, ns
	Model R Sq = 0.29, F = 1.6(37, 146), p = 0.029
Perceived QOL	Social Categories Education (β = −0.15, p = 0.063) White identity (β = 0.16, p = 0.08) Income (β = 0.15, p = 0.047)	0.077	1.8, p = 0.076
	Legacy media use Listen to radio (β = 0.22, p = 0.010) Read magazines (β = −0.21, p = 0.061) Watch movies in theater (β = 0.18, p = 0.043)	0.099	2.5, p = 0.013
	Use of newer media & technologies Browse TikTok for videos (β = 0.19, p = 0.039)	0.070	1.0, ns
	Genre taste publics + None sig.	0.043	1.3, ns
	Model R Sq = 0.29, F = 1.6(37, 146), p = 0.026
Happiness	Social Categories Pol philosophy (β = −0.32, p = 0.013)	0.128	3.2, p = 0.002
	Legacy media use Listen to radio (β = 0.29, p < 0.001) Watch movies in theater (β = 0.14, p = 0.096)	0.113	3.1, p = 0.003
	Use of newer media & technologies Watch videos on smart phone (β = 0.15, p = 0.099) Stream a film on home TV (β = −0.23, p = 0.058) Stream a TV program on home TV (β = 0.22, p = 0.055)	0.070	1.1, ns
	Genre taste publics + None sig.	0.036	1.2, ns
	Model R Sq = 0.35, F = 2.1(37, 146), p < 0.001
Cosmopoliteness	Social Categories Education (β = −0.28, p = 0.001) Pol philosophy (β = −0.23, p = 0.088) Demo party (β = 0.19, p = 0.078) Income (β = −0.15, p = 0.078)	0.102	2.5, p = 0.015
	Legacy media use Listen to radio (β = 0.18, p = 0.032) Read books (β = −0.18, p = 0.031)	0.050	1.2, ns
	Use of newer media & technologies Go on Facebook (β = −0.26, p = 0.009)	0.076	1.1, ns
	Genre taste publics + Romance & drama (β = 0.18, p = 0.078)	0.045	1.3, ns
	Model R Sq = 0.27, F = 1.5(37, 146), p = 0.054
Country headed in wrong direction	Social Categories None sig.	0.034	0.8, ns
	Legacy media use None sig.	0.031	0.7, ns
	Use of newer media & technologies None sig.	0.043	0.5, ns
	Genre taste publics + None sig.	0.032	0.8, ns
	Model R Sq = 0.14, F = 0.6(37, 146), ns

.

We see from Table 6 and Table 7 that the film genre tastes perform well in predicting all five criterion variables for the 2015 general population sample and the 2021 creator sample. In general, the weakest prediction is found for whether the country is headed in the wrong direction. For the regression on whether the country is headed in the wrong direction in the 2015 study, only education appears as a significant predictor in the first block and then none of the variables are significant until the final block with genre taste publics. Re RQ3, attendance at public leisure activities serves as a significant predictor of community attachment, but media use also is significant.

The 2022 and 2024 student sample regressions (Table 8 and Table 9) differ from the other two in predictable ways. Students living on a campus live in a community to which they are attached for only four years, so community attachment predictors would be quite different. We find that political variables are much more important in several regressions, and only a couple genre taste public variables appear to be important, especially Sports & Conflict, though History Buff and Big Screen Fantasy make solo appearances.

5. Discussion

The present study examines whether the mainstreaming hypothesis, derived from cultivation frameworks developed during the mass audience era, still applies in a contemporary media environment characterized by fragmenting media and cultural taste publics. In particular, we consider evolving conceptions of mainstreaming and media influences in four surveys conducted from 2015 to 2024. We broaden our view of media to see if entertainment content—especially film genres—can provide common ground attracting people with little else in common.

Study findings suggest such “cultural mainstreaming” may occur by providing common gratifications and impact global indictors of our lives—happiness, community attachment, feelings about our quality of life, and perceived cosmopoliteness. However, these results are limited to a general adult population, not the younger students studied. Simply put, mainstreaming and media influence are more complex processes in the digital era.

On balance, study findings suggest that the cultivation effect is quite weak; that is, the modest number of significant relationships appearing seems no greater than that which would be generated by chance alone. Given the scope of study data, encompassing four sizeable samples across a decade, this dynamic reflects either “poor reception” of information (Wober 1990) via the media or we need to look at cultivation and mainstreaming as more complex processes.

In theoretical terms, the concepts of cultivation and mainstreaming are appealing, since they harken back to a more ambitious era of theory-building characterized by “unified field” perspectives on media influence (e.g., Jeffres et al. 2008). Although parsimony represents one of the theory’s key strengths, clearly, this conception is no longer sufficient to describe current media influence processes. However, if we reframe mainstreaming as a process of “homogenization”—comparing variances and distributions rather than mean differences—it is possible to detect more nuanced public opinion convergences than might be revealed by more conventional mean comparisons (e.g., Jeffres et al. 1995, 2001). If we assume that mainstreaming only operates across time, then longitudinal data will be needed to definitively track these centripetal forces over time. This movement should be better reflected in the static distributions than was apparent here.

One possible explanation for these marginal cultivation effects lies in the fragmented media environment, as Kubin and von Sikorski’s (2021) meta-analysis found that “pro-attitudinal media exacerbates polarization” (p. 188). As these partisan pro-attitudinal media generally supplant homogenizing legacy media channels, the net of media exposure may favor fragmentation over reinforcement. Taken together, our several regression analyses underscore the inconsistent role of digital media in predicting our five criterion life attitude variables. These influences extend beyond the traditional news media, to encompass emerging modalities like texting (per the 2015 survey)—and entertainment influences like film genre preferences—as determinants of Life Attitude variables, our specialized measures of the public in a “common ground.” Returning to our opening discussion on cultural fragmentation, study results underscore the centrifugal forces shaping American public opinion (e.g., Jeffres et al. 2023).

As a consequence, later work will need to reformulate the methods and conceptions of penitential accumulative effects theories—e.g., cultivation, agenda-setting—in light of these centripetal audience influence dynamics (e.g., Chaffee and Metzger 2001). Although the present research program tapped exposure to measures to extend beyond television to reflect digital media, this portrait fails to accurately capture the “media reality” to which audiences are exposed. In an era of unprecedented choice, it will be critical to move beyond broad assumptions of media homogeneity and conduct rigorous program-specific profiles—across a range of digital channels—to establish these media environment baselines.

Furthermore, we need to seek conceptualizations which integrate more short-term “effects” (as audiences process newscasts, TV programs and newspaper articles on a daily basis) with long-term consequences such as enduring images, beliefs and attitudes. Clearly, we will need more panel studies than surveys to accomplish this task. As streaming affords audiences access to all virtually any movie or series on demand—from across the globe—the inertia for conducting matching content analyses may wane.

Taken together, our findings present a dual picture of media influence in the digital age. On the one hand, the scatterstreaming trend—replete with the fragmentation of audiences and diversity of fare—dominates much of the media landscape (especially news and political information). On the other hand, looking only at the broader adult surveys rather than student responses, our findings suggest we may be looking in the wrong place for “common ground.” Within the realm of entertainment content, we see evidence that some newer technologies and genre taste publics (Graan 2022) may actually pull together audiences otherwise fragmented by news media. Beyoncé’s Cowboy Carter Concerts—where an African American entertainer drew enthusiastic country fans—illustrates how entertainment can capture broad audiences. We need research that increases our attention to entertainment media, especially social media such as TikTok, to see how media impact audiences.

The “golden era” of theorizing in the late 20th century produced penitential mass-society perspectives like Agenda Setting and Cultivation (e.g., Atkin et al. 2015). However, the fading role uncovered by legacy media here reinforces Chaffee and Metzger’s (2001) notion of the end of mass communication; this era—forged during a rare moment when mass-audience vehicles delivered homogenized content to undifferentiated audiences—needs to be reformulated to accommodate “displaced effects” perspectives wrought by fragmenting digital environments (e.g., Nowak et al. 2010).

The very complexity of contemporary streaming environments thus presents logistical challenges to the feasibility, if not the advisability of conducting matching content analyses to establish the content structures to which audiences are exposed; that is, in a chaotic program environment—where streamers enjoy access to virtually every film/series in global circulation—gaining an accurate profile of the “TV answer” may prove prohibitively difficult. Researchers trying to estimate these idiosyncratic streaming habits, amidst a streaming menu of abundance, would be hard-pressed to accurately record their viewing. As the feasibility (and utility) of gauging fractionated exposure continues to wane researchers may, perforce, replace content analyses and self-report data with aggregate big-data metrics that can capture exposure to variegated digital platforms.

Although these balkanizing trends seem to contradict the public opinion convergences described by venerable mass-society perspectives, their late 20th century genesis may stand out a rare anomaly over the longer arc of history. It will be important in later work, then, to expand our analysis beyond different media modalities and focus on content-specific measures of exposure (e.g., videos and film/television genres). Future research could profitably track corollary public opinion influences accompanying the long-tail of digital media.