Cultivation, a theory for TV that explains media effects on worldview

For all the times that Cultivation theory has come up thus far, Morgan et al. (2009) finally lay out the breadth and nuance of the perspective. These discussions on the effects of television and of cultivation intrigue me far more than the results of any one experiment exactly because of the breadth and nuance that can be brought forward.

I was a teenager during the rise of violent video games, and many of my first forays into news and politics were through the lens of someone on the receiving end of a moral panic. Don’t take away my violent video games, Senator; I can tell the difference between fantasy and reality. I didn’t understand why video games were always being bullied in the traditional media, when TV, movies, and books had as much violence. I blamed political scapegoating and an entrenched media that feared and distrusted an emerging artform. It is good to revisit these thoughts from a nuanced understanding, and to be fair to my younger self, the portrayal of these topics in the media always takes on a simplistic framing. Morgan et al. suggests that violence comes up again and again in media effects research because it is easy to study, gets funding, and makes the work of researcher more accessible to the general public through a concrete topic. Violent media and its effects lend themselves to being easily conceptualized and discussed. They, as well as Jacobs et al. (2017), suggest that the clearest effects of violent media are on a person’s worldview, not necessarily on their likelihood to commit violent acts themselves.

My love of video games is likely what led to my unchecked optimism for technology. Televisions were getting bigger and cheaper, with high‐definition on the horizon, while video games, computing, and graphics technologies were rapidly advancing, and I thought this was great — it was great for everyone and everything, always. Current events have tempered this feeling — current events or that, oh no, I’m getting older. For my youth, I was steeped in online video game forums, gaming‐industry and technology magazines, television shows about science‐and‐technology, and technology‐and‐business related news. Without realizing it, I’d been pushed toward a worldview that normalized a perspective on technology that was extreme in optimism, trusting both industry‐leaders and start‐ups implicitly, and one that expected and demanded a high degree of morally, economically, and environmentally unsustainable obsolescence. New, new, new; the next big thing always meant something better.

From these new understandings, it’s easy to see how my media selection and exposure cultivated a specific outlook. The realities of growing up before the end of Moore’s Law, and being an end‐user to the industries driven by it, gave me heuristic models for understanding the world that privileged the newest things as the best things. Even just a few years ago, my thoughts on technology were somewhat nuanced but still firmly in the tech‐enthusiast camp. The criticism of technology always seems to be framed around “Millennials,” and I see this as a trope. Cries that a current generation is ruining civilization are as old as civilizations themselves while the voracity of undue criticism can mask actual social issues — for everyone. I still believe this last bit, but at the time I also thought that there was a certain xenophobia around the new forms of cultural expression that devices like cellphones permit. Like emojis; it’s probably not worse, just different. At the time, I made the argument that, yes, while everyone is walking through public with their faces glued to their devices and ignoring the people near them, those same devices permit them to be, not anti‐social, but hyper‐social. Internet‐connected devices permit choosing quality over proximity in social interactions. Today, however, these thoughts are overtaken by just how different the current environment is from anything that came before, and the effects that emerge as the new media environment has matured. As I go next into the media effects literature addressing new media, I’m curious to see how my current positions are informed or changed by the research.

Literature reviewed
  • Jacobs, Laura, Marc Hooghe, and Thomas de Vroome. “Television and Anti‐Immigrant Sentiments: The Mediating Role of Fear of Crime and Perceived Ethnic Diversity.” European Societies 19, no. 3 (May 27, 2017): 243 – 67. https://doi.org/10.1080/14616696.2017.1290264.
  • Morgan, M., J. Shanahan, and N. Signorielli. “Growing up with Television: Cultivation Processes.” In Media Effects: Advances in Theory and Research, edited by Jennings Bryant and Mary Beth Oliver, 3rd ed., 34 – 49. Communication Series. Communication Theory and Methodology. New York: Routledge, 2009
  • Riddle, Karyn, W. James Potter, Miriam J. Metzger, Robin L. Nabi, and Daniel G. Linz. “Beyond Cultivation: Exploring the Effects of Frequency, Recency, and Vivid Autobiographical Memories for Violent Media.” Media Psychology 14, no. 2 (May 31, 2011): 168 – 91. https://doi.org/10.1080/15213269.2011.573464.

Designer’s perspective on health communication research

This week I watched some conveniently relevant news reports on health and advertising (perhaps this could be considered a priming effect?) while reading on health information effects. The videos, special reports put out by Vox, covered the effects of today’s advertising on public health — specifically the use of nicotine by minors and young adults, which had been nearing zero until e‐cigarettes started being marketed and framed separately from smoking cessation (Vox, 2018, “How Juul made nicotine go viral”), as well as prescription drug ads (Vox, 2016, “How Americans got stuck with endless drug ads”).

Initially, these reports interested me because they touch on what effect the interpretive aspects of an ad’s visual vernacular have on the audience — like model ages, gestures/poses, scenery, facial expressions, and other visual connotations or cues that build “expectancy.” These reports and the media effects literature on health follow‐up on areas of my interest that were last influenced by Jean Kilbourne and Sut Jhally. They also provide some scholarly sources that I can add to a growing selection of literature I’ve been collecting — searching for a perspective to address “graphic design effects.”

But, back to health. Messages related to health and medicine are particularly worrisome for the fatal implications of being misinformed and worrisome for the realities of medical research: null results are not published while news reporting exaggerate and misconstrue the underlying medical literature.

The media effects perspectives on internet health resources are particularly interesting, for both their positive and negative outcomes. This makes me wonder to what extent media effects researchers can investigate historical periods. I want to know if the spread of public libraries (“Carnegie libraries”) in towns across the US had similar consequences for public health as digitally democratized sources.

Randolph and Viswanath’s treatment on public health effects (2004) was slightly bizarre to read. It is the first paper that I’ve read to directly discuss practical, goal‐directed applications of media effects research — applying the models and research toward achieving some specific outcome. Even though the discussion is directed toward theory‐based campaigns for the public good, I nevertheless found it alarming. This reaction is somewhat hypocritical, of course, because that process — create an intervention, evaluate the resulting behaviors, then revise and repeat — is a design process by the most fundamental definitions of design.

Message design is also discussed by Anket et al. (2016), though there is a misunderstanding behind their discussion of Noar (2006b) and of message design. Noar can’t be specific related to message design because that’s not how design solutions work — they are nearly always one‐offs. This is true of all design, but is especially true of communication design. You’re working with people; messy, complicated, and intensely irrational in one moment then intensely rational in the next. Worse, your only tool is language, which is both imprecise and endlessly descriptive. Even that word, language, makes my point for me because it is so ubiquitous and abstract that how I use it here may call up unintended understandings in a reader. All of this is to say that repeating campaigns and messages can ignore hidden contexts that allowed the initial ones to be successful, and it risks the message becoming stale, or worse, subverted with each use. There is nothing inherently wrong with the three message elements that they lay out (use of celebrity, community members, and audience participation), but there is nothing inherently right, either. The biggest factor ignored by their meta‐analysis is novelty, which I argue was a major contributor to the success of The Truth tobacco cessation campaign.

To the authors, or anyone hoping to evaluate such public health campaigns, I would offer this advice: Anker et al.’s three “inclusion criteria” are a good starting point for a design brief, but if you are searching for replicability between campaigns then I’m sorry, but communication design is not a social science. Design by committee, by formula, and by template will fail you, and there are so many confounding variables to deal with that you will have to rely on intuition. That’s not to say that you shouldn’t be informed, but establish your intent, research prior campaigns, research unrelated (even commercial) campaigns, and above all else work to gain an understanding of your target audience — their social, media, health, and internal contexts — and then try to do something different than what other people are targeting these people with. Your message will stand out by contrast. From this mindset, craft many, divergent versions of your message and then test and iterate on them. Helen Armstrong’s book, Participate, offers valuable insight on the practicalities of user testing. But also take inspiration from unexpected places. Paul Bennett, a partner in the design firm IDEO which has a prolific history of design within health contexts, discusses the mindsets necessary for this in his talk, “Design is in the Details.” Stanford’s Design program hosts a resource that might be helpful in creating better models for the design of public health campaigns (see “A Virtual Crash Course in Design Thinking”).

This week I watched some conveniently relevant news reports on health and advertising (perhaps this could be considered a priming effect?) while reading on health information effects. The videos, special reports put out by Vox, covered the effects of today’s advertising on public health — specifically the use of nicotine by minors and young adults, which had been nearing zero until e‐cigarettes started being marketed and framed separately from smoking cessation (Vox, 2018, “How Juul made nicotine go viral”), as well as prescription drug ads (Vox, 2016, “How Americans got stuck with endless drug ads”).

Initially, these reports interested me because they touch on what effect the interpretive aspects of an ad’s visual vernacular have on the audience — like model ages, gestures/poses, scenery, facial expressions, and other visual connotations or cues that build “expectancy.” These reports and the media effects literature on health follow‐up on areas of my interest that were last influenced by Jean Kilbourne and Sut Jhally. They also provide some scholarly sources that I can add to a growing selection of literature I’ve been collecting — searching for a perspective to address “graphic design effects.”

But, back to health. Messages related to health and medicine are particularly worrisome for the fatal implications of being misinformed and worrisome for the realities of medical research: null results are not published while news reporting exaggerate and misconstrue the underlying medical literature.

The media effects perspectives on internet health resources are particularly interesting, for both their positive and negative outcomes. This makes me wonder to what extent media effects researchers can investigate historical periods. I want to know if the spread of public libraries (“Carnegie libraries”) in towns across the US had similar consequences for public health as digitally democratized sources.

Randolph and Viswanath’s treatment on public health effects (2004) was slightly bizarre to read. It is the first paper that I’ve read to directly discuss practical, goal‐directed applications of media effects research — applying the models and research toward achieving some specific outcome. Even though the discussion is directed toward theory‐based campaigns for the public good, I nevertheless found it alarming. This reaction is somewhat hypocritical, of course, because that process — create an intervention, evaluate the resulting behaviors, then revise and repeat — is a design process by the most fundamental definitions of design.

Message design is also discussed by Anket et al. (2016), though there is a misunderstanding behind their discussion of Noar (2006b) and of message design. Noar can’t be specific related to message design because that’s not how design solutions work — they are nearly always one‐offs. This is true of all design, but is especially true of communication design. You’re working with people; messy, complicated, and intensely irrational in one moment then intensely rational in the next. Worse, your only tool is language, which is both imprecise and endlessly descriptive. Even that word, language, makes my point for me because it is so ubiquitous and abstract that how I use it here may call up unintended understandings in a reader. All of this is to say that repeating campaigns and messages can ignore hidden contexts that allowed the initial ones to be successful, and it risks the message becoming stale, or worse, subverted with each use. There is nothing inherently wrong with the three message elements that they lay out (use of celebrity, community members, and audience participation), but there is nothing inherently right, either. The biggest factor ignored by their meta‐analysis is novelty, which I argue was a major contributor to the success of The Truth tobacco cessation campaign.

To the authors, or anyone hoping to evaluate such public health campaigns, I would offer this advice: Anker et al.’s three “inclusion criteria” are a good starting point for a design brief, but if you are searching for replicability between campaigns then I’m sorry, but communication design is not a social science. Design by committee, by formula, and by template will fail you, and there are so many confounding variables to deal with that you will have to rely on intuition. That’s not to say that you shouldn’t be informed, but establish your intent, research prior campaigns, research unrelated (even commercial) campaigns, and above all else work to gain an understanding of your target audience — their social, media, health, and internal contexts — and then try to do something different than what other people are targeting these people with. Your message will stand out by contrast. From this mindset, craft many, divergent versions of your message and then test and iterate on them. Helen Armstrong’s book, Participate, offers valuable insight on the practicalities of user testing. But also take inspiration from unexpected places. Paul Bennett, a partner in the design firm IDEO which has a prolific history of design within health contexts, discusses the mindsets necessary for this in his talk, “Design is in the Details.” Stanford’s Design program hosts a resource that might be helpful in creating better models for the design of public health campaigns (see “A Virtual Crash Course in Design Thinking”).

Literature reviewed

Connecting media priming theories to graphic design and gay rights

This week’s readings, in light of previous weeks, has given me an appreciation of the process behind social science and the challenges that media effects researchers and scholars face – the scope of the current models allow for an expanded breadth and specificity for investigating media effects, but significant limitations persist. The field fully admits that the models for understanding and explaining media effects are myopic. But, it seems that this is as much a feature of social science as it is a criticism of media effects literature. Addressing overly narrow and ungeneralizable models has to be done slowly, through incremental gains over decades, because each addition to a model is a complication that must be proven. The design process has something similar: iteration. It’s an important process for improving work, but it has a blind spot. You can slowly walk yourself down a dead‐end. So, early on in the ideation process, designers will inject chaos into their process and play with the results – this trial‐and‐error experimentation relies heavily on intuition.

I’ve been thinking about intuition a lot lately. In social science, intuition can obviously be a dangerous trap for researchers, leading to biases and unfounded assumptions that have to be mete out. Designer’s, however, rely on intuition to guide their work. The “effects” of graphic design can’t be reliably produced without it – formulaic design fails. This semester I’m teaching a freshmen‐level design class where students learn to identify and create gestalt “effects” with principles like closure, contrast, hierarchy, and implied movement. By creating a large set of studies using very basic shapes to elicit these principles, they begin to intuit how to do this. Sometimes intuition, inspiration, and other creative things are couched in airs of mysticism, and this is unfortunate.

Even if we, as individuals, can’t articulate or be fully aware of how and why we have certain perceptions and feelings, intuition can be externally understood and explained. Intuition is a type of heuristic thinking. Gaining access to that terminology was important in my professional development because a perceived mysticism originally pushed me away from art and design.

Rosko‐Ewoldsen et al. have expanded my lexicon further. The general affective aggression model qualifies intuition as primary appraisal based on mental models or schema. So, to rephrase, this Bauhaus‐style of design pedagogy helps students build good, versatile mental models for interpreting visual information. Then, over the course of their education, these mental models are developed into more abstract – and thus, more widely applicable – schema that they can use to apply their craft to cross‐ and interdisciplinary means.

This leads me to wonder how we generate and train our own our mental models at the casual level. Rosko‐Ewoldsen et al. state that our mental models are under some amount of self‐control. We can direct changes to them, swap out and test different ones, and apply them to information at will. They can be viewed as fast and efficient or as quick and dirty. While this fails to explain the role of circumstance, rather than intention, in the formation of mental constructs, it explains an area of filmic studies that I’ve been fascinated with since my undergraduate studies, which is the gay read.

Before the papers themselves primed my rants and tangents on intuition, I was looking forward to digging into the topic of media stereotyping because I wanted to talk about one thing: The relationship between film and the Gay Rights movement. First, these were people who did not have characters that they could identify with, but that didn’t stop them. Gay affections and cross‐dressing are frequently played for laughs, and today we could (perhaps should) condemn such stereotypical portrayals, but at the time characters and scenes with these elements gave the only mirrors from which an LGBT person could see themselves. They also gave heteronormative films a “gay read,” which is applying the constructs and mental models from their personal and romantic lives to films in order to understand the characters differently and gain further mirrors for their self‐identity. The best, however, were campy movies.

Campy is a term that’s hard to pin down, but it’s used to describe movies that are over‐the‐top in a way that’s appealing. Think, so bad it’s good. For LGBT persons in the 70s through the 90s, campy movies provided the only portrayals of gay and lesbian characters. They were usually low‐budget, which gave creators some freedom to include (what were then) subversive topics. This was crucial not just for their individual identities, but also for their group identities. Communities centered around viewing these films, they gave rise to cult fandoms like that of Rocky Horror, but they also gave much more than entertainment. These communities gave a sense of normalcy, defined what it meant to be something other than straight, and were a source of the collective courage needed to come out and to pursue the political action necessary to secure a better future at a point in history when admitting this meant risking your job, your family, and your safety.

​How is that for the power of media effects? Or, for the monumental importance of fair and diverse portrayals of societal groups?

Literature reviewed
  • Mastro, Dana. 2009. “Effects of Racial and Ethnic Stereotyping.” In Media Effects: Advances in Theory and Research, edited by Jennings Bryant and Mary Beth Oliver, 3rd ed. Communication Series. Communication Theory and Methodology. New York: Routledge.
  • — —  — . n.d. “Effects of Racial and Ethnic Stereotyping,” 13.
  • Rosko‐Ewoldsen, D. R., B. Rosko‐Ewoldsen, and F. D. Carpentier. 2009. “Media Priming: An Updated Synthesis.” In Media Effects: Advances in Theory and Research, edited by Jennings Bryant and Mary Beth Oliver, 3rd ed, 74 – 93. Communication Series. Communication Theory and Methodology. New York: Routledge.
  • Skinner, Allison L., and Jacob E. Cheadle. 2016. “The ‘Obama Effect’? Priming Contemporary Racial Milestones Increases Implicit Racial Bias among Whites.” Social Cognition 34 (6): 544 – 58. https://doi.org/10.1521/soco.2016.34.6.544.
  • Tukachinsky, Riva, Dana Mastro, and Moran Yarchi. 2015. “Documenting Portrayals of Race/Ethnicity on Primetime Television over a 20‐Year Span and Their Association with National‐Level Racial/Ethnic Attitudes: TV Portrayals and National‐Level Attitudes.” Journal of Social Issues 71 (1): 17 – 38. https://doi.org/10.1111/josi.12094.

Media Effects in Aggregate: Confluences and limitations

In prior updates, these posts have addressed the various complications of media effects research; across models, eras, and paradigms, media effects research is a complicated field trying to study, figuratively, already fidgety subjects within contexts that are moving faster than scholarly papers can be produced (Valkenburg, Peter, and Walther 2016, 331). The field either needs new theories (Bennett and Iyengar 2008, 708) or to find the underlying assumptions and biases which are undermining their perspectives (Neuman 2018, 376). With this aim in mind, Valkenburg & Peter (2013) and Valkenburg, Peter, & Walther (2016) strive to reassess the literature. These papers deconstruct the literature to make cross‐comparisons and groupings between independently developed models, to find unifying descriptors for areas of seemingly infinite variety, and to collect and re‐integrate their findings into a digestible, succinct‐yet‐comprehensive form.

Valkenburg, Peter, and Walther identify five “global features” of media that are shared amongst different models, which they use to mark out the contexts which media effects models address (2016, 319). The features are audience selectivity, media properties (like mode, format, and genre), an assumption of indirect media effects, conditional effects based on individual differences, and transactional or reciprocal media effects (Valkenburg, Peter, and Walther 2016, 318 – 26). From their analysis, they find that both media and nonmedia factors are inexorably linked (Valkenburg, Peter, and Walther 2016, 332). Media effects take form through a web of interdependent factors that exist within and between media, outside the media, and even outside the context of media‐use. Since the factors involved in media use, influence, and selection cannot be taken in isolation, more sophisticated and comprehensive models are required. They posit that inconsistencies and minimal media effects might be caused by undertheorizing (Valkenburg and Peter 2013, 222). However, this poses methodological challenges, as “theories in the social sciences are not applicable irrespective of context” (Busse, Kach, and Wagner 2017, 576).

Media effects models as a whole are at risk of narrowing generalizability in the face of technology‐driven social changes (Valkenburg, Peter, and Walther 2016, 332). Other researchers are even more damning, with Bennett and Iyengar (2008) believing that the validity of current research methods and theories faces risks that can only be addressed through new theorizing. Other researchers believe the models may be valid, but incomplete, and instead they call for structural changes to the publishing and funding of research that would permit researchers to pursue their work with critically different mindsets: “We may be asking the right questions but have a paradoxical paradigm‐induced blind spot that leads us to ignore or explain away null finding or reverse effect” (Neuman 2018, 376).

Literature reviewed

Audience Selectivity: Uses‐and‐gratifications and selective exposure

Audience selectivity is an area of research that looks to explain the reason that people make the media use choices that they do. The two dominant approaches, uses‐and‐gratifications and selective exposure, share a perspective shift from other types of media research — they focus, not on what the media does to audiences, but on what audiences do with the media (Rubin 2009, 168; Valkenburg, Peter, and Walther 2016, 320); however, the two approaches to selectivity research hold incompatible perspectives on collecting and interpreting data.

The shift to audience‐centered research requires a psychological perspective (Camaj 2019, 1), and for the models to consider the factors that influence selectivity “within the context of other influences” (Rubin 2009, 165). With these distinctions, selectivity research gains a more holistic understanding of the layers of intervening variables that guide and inform the selection process. Both approaches explain selection choices as being based on the needs and desires of the individual as moderated by psychological and social factors (Valkenburg, Peter, and Walther 2016, 320).

From these points of agreement, they diverge in ways with significant implications for data collection and validity. Uses‐and‐gratifications looks to explain the factors that influence an individual’s selection choices because audiences are active participants in both selecting and interpreting media. It further presumes that audiences are fully self‐aware and able to accurately account for the selection‐making process. This is important because uses and gratifications relies on self‐reports as its principal form of data collection. If this assumption is wholly wrong, then the entire corpus of uses and gratifications studies is called into question. Selective exposure, however, argues the opposite; it assumes that audiences are “not fully aware of their selection motives” (Valkenburg, Peter, and Walther 2016, 320). Self‐report data, then, is used minimally (Knobloch‐Westerwick 2015, 8), which comes with the risk of discounting important data related to the selection experience. Instead, selective exposure researchers use behavioral observation in the hopes of collecting more objective measures (Knobloch‐Westerwick 2015, 8). However, these methods bring their own issues because behavioral observation can be invasive, overt observations can alter or bias selection motives and choices, and observational studies — functionally — imply a reliance on laboratory contexts that may hamper generalizability. Being unable to study audiences in‐situ, which would be observing media use and selection as it actually occurs and as it is influenced by differing contexts and environments, may undermine the implications and conclusions of selective exposure research. If you build up models and theories that can only explain selection in a laboratory context, then what have you gained — or more crucially, what have you missed?

The two branches of audience selectivity will persist until one or both of their limitations can be addressed. Setting aside philosophical implications regarding the mind and the impossibility of neutrality, in the practical sense the two branches of research cannot be unified until they are able to accurately qualify the limitations of self‐reporting or to achieve designs of behavioral observation that demonstrably minimize intrusion while remaining ethically sound.

Literature reviewed
  • Camaj, Lindita. 2019. “Week 3: Media Audiences & Selectivity.” Notes. University of Houston. COMM 6317: Media Effects.
  • Knobloch‐Westerwick, Silvia. 2015. “Building Blocks of the Selective Exposure Paradigm.” In Choice and Preference in Media Use: Advances in Selective Exposure Theory and Research, 3 – 24. New York: Routledge.
  • Rubin, A. M. 2009. “Uses‐and‐Gratifications Perspective of Media Effects.” In Media Effects: Advances in Theory and Research, edited by Jennings Bryant and Mary Beth Oliver, 3rd ed, 165 – 84. Communication Series. Communication Theory and Methodology. New York: Routledge.
  • Valkenburg, Patti M., Jochen Peter, and Joseph B. Walther. 2016. “Media Effects: Theory and Research.” Annual Review of Psychology 67 (1): 315 – 38. https://doi.org/10.1146/annurev-psych-122414 – 033608.

Classifying Infinite Variety

Once scholars shifted to incorporate how audiences themselves effect media, they were confronted with the messy realities of humanity; that our individual differences complicate, confound, and defy but that those differences also “[represent] the very thing that makes humans interesting, unique, and infinitely worthy of our research attention” (Oliver and Krakowiak 2009, 517). Models had to become to more sophisticated to adequately explain media use and influence as new technologies upset old paradigms and as media audiences generated new trends and variety in media use. Using a negative outlook for illustrative purposes, media use has consequences that can be described in terms of where they take place or manifest, the forms those consequences take, and finally the severity of their outcomes.

Media effects can manifest at multiple levels. Overtime, models had to incorporate different levels of analysis to explain apparent effects because influences and effects take form both within and between people. Broadly, these are the micro‐ and macrolevels of media effects (McLeod, Kosicki, & McLeod, 2010, via Potter 2011, 903). The micro is the intrapersonal level — it includes factors like pre‐existing beliefs and amount of activity or engagement — whereas the macrolevel is concerned with the societal effects of media (Potter 2011, 903 – 4). However, this initial division ignored how effects manifest differently within societal groups (as well as institutions, organizations, and cultures) than they do within relatively smaller groups of individuals (like social groups and co‐workers), so researchers like Chaffee and Berger (1987), McLeod, Kosicki, and McLeod (2010), and Potter (2011) have suggested additional levels of analysis that fall between the individual and the many (Potter 2011, 903 – 4). This is the mesolevel, though some research further distinguishes between person‐to‐person and network scales of interaction (Potter 2011, 904). Across the different levels, models also need to consider the results of media use: their forms and the impact that they have.

First, Potter (2011) frames the outcomes as coming in distinct types of effects — the media can act on our “cognitions, attitudes, beliefs, affects, physiology, and behaviors” (2011, 904), though the exact typology varies across the literature (Potter 2011, 897). Valkenburg and Peter’s Differential Susceptibility Model of Media‐use expands the means of discussing outcomes by reframing the aforementioned types as conditional effects and adding four other categories of indirect media effect — collected and synthesized from the media effects literature (2013, 222 – 23). Those categories consider how outcomes can manifest differently based on: (1) how/why/for‐what individuals are using the media, (2) how individuals mentally and physiologically respond to that media, (3) the “second‐order” media effects resulting from other media effects, and (4) the reciprocity between outcomes and further media use (Valkenburg and Peter 2013, 222 – 24).

Finally, each outcome of media use has an impact that can be described in terms of varying strengths of effects. Implicit in this is change. Effects can result in a change to an attribute which can vary in strength — take a jumpscare, which triggers an orienting reflex, elicits a strong (magnitude of the change) physiological response (type of effect) in a person (level of effect) that increases their heart rate (a specific measure of change) — but this is a fundamentally limited perspective. Potter expands types of changes to include other dimensions. He intuits three properties to describe the changes resulting from an effect: “kind, magnitude, and weight” (2011, 904). This permits him to describe amounts like minuscule to large changes (magnitude: a change occurred, by how much did the attribute change?), changes that reinforce existing factors and resist further change (weight: does the change reinforce/strengthen existing beliefs?), as well as more radical shifts which overcome prior insusceptibility (kind: true persuasion, the change to an attribute was from one type to another). That said, the distinction between magnitude and weight can be confusing — they are a mixed metaphor — and Potter’s later use of waveform metaphor (2011, 907 – 9) or, alternatively, terms from social influence scholarship like change, formation, and reinforcement (Holbert, Garrett, and Gleason 2010, 17) may provide better clarity when quantifying and qualifying what results from a media effect.

Literature reviewed
  • Holbert, R. Lance, R. Kelly Garrett, and Laurel S. Gleason. 2010. “A New Era of Minimal Effects? A Response to Bennett and Iyengar.” Journal of Communication 60 (1): 15 – 34. https://doi.org/10.1111/j.1460 – 2466.2009.01470.x.
  • Oliver, B., and K. M. Krakowiak. 2009. “Individual Differences in Media Effects.” In Media Effects: Advances in Theory and Research, edited by Jennings Bryant and Mary Beth Oliver, 3rd ed, 517 – 31. Communication Series. Communication Theory and Methodology. New York: Routledge.
  • Potter, W. James. 2011. “Conceptualizing Mass Media Effect.” Journal of Communication 61 (5): 896 – 915. https://doi.org/10.1111/j.1460 – 2466.2011.01586.x.
  • Valkenburg, Patti M., and Jochen Peter. 2013. “The Differential Susceptibility to Media Effects Model.” Journal of Communication 63 (2): 221 – 43. https://doi.org/10.1111/jcom.12024.

Polarity in Effects Research

How much impact does mass communication have on audiences? This central question of media effects came about during the early Twentieth Century when the proliferation of broadcast and “industrialized” media was underway and the consequences of “European totalitarian propaganda” were becoming apparent (Neuman and Guggenheim 2011, 171). Since its inception as a formal field of study — and perhaps before (see Greenberg and Salwen 2009, 61) — Media Effects research has gone through phases that can be typified as dealing with, and occasionally assuming, significant effects and minimal effects. This has its origins in the specific contexts of different periods and places, it has shifted from contrasting extremes to nuanced perspectives, and it has presented a narrative that media effects researchers have relied on to orient their field for both critical and detrimental purposes.

If it was the industrialization of mass media that began the study of media effects, then it was for what industrialization entailed. In broader contexts, it came with the newfound uniformity and ubiquity of goods that mass production offered (Geddes 1932, 13 – 15). For communication researchers, industrialization of the media meant that state‐sponsored propaganda could be served directly to the people instead of filtering through error‐prone messengers and the well‐informed (Neuman and Guggenheim 2011, 171; Valkenburg, Peter, and Walther 2016, 316). The Modern era was a societal paradigm shift — this was a period when everyone shifted from purchasing artisan handiwork to buying identical teapots. Just like the artisans imparted imperfections and humanistic qualities to their work, messengers imparted their own (Neuman 2018, 370 – 71, 374). Applying this analogy, what changes when messages lose their humanistic qualities and imperfect transmission? Greenberg and Salwen posit that in the aftermath of WWI there was an intuitive belief that a very few, powerful individuals would be able to use an industrializing and modernizing media to send propaganda with uniform effects (2009, 62). Neuman and Guggenheim indicate that the early “magic bullet” perspective of media effects may have had more nuance than it is typically credited (2011, 172), which may also indicate that early media effects research was prompted by concerns over these new technologies and media contexts, mirroring today’s similar, if opposing, concerns (compare Bennett and Iyengar 2008, 716; Chaffee and Metzger 2001, 367). However, some researchers and the public at‐large held the impression that media had a direct influence over “vulnerable audiences” (Chaffee and Metzger 2001, 366 – 67), that the message need only be received to have impact, and that audiences were homogeneous, passive, and “malleable” (Greenberg and Salwen 2009, 62).

Researchers eventually abandoned this perspective — and in light of Congressional interest, researchers like Klapper heavily criticized it (Neuman and Guggenheim 2011, 172). Mass media doesn’t have such clear effects; nor are audiences homogenous, passive, or malleable. The earlier notions were predicated on a mechanistic approach to the study of mass media (Rubin 2009, 165), but these simplistic models presumed uniform and immediate effects and were therefore ill‐equipped to explain the factors that lie between communication and effect (Carey, 1989, via Neuman 2018, 370). The minimal effects hypothesis was championed by Klapper in response to political science research which revealed that political persuasion was heavily mediated by external factors like predisposition, interpersonal relationships, and the influence of thought‐leaders (Neuman and Guggenheim 2011, 172). Klapper argued that the media’s influence was confirming and reinforcing existing beliefs (Greenberg and Salwen 2009, 66 – 67). This era of minimal effects was an attempt to explain why there were not uniform and immediate effects resulting from media use by using “two‐step flow” which models how such external factors as personal influence might mediate messages and persuasive effects (Bennett and Iyengar 2008, 707 – 8; Greenberg and Salwen 2009, 72; Valkenburg, Peter, and Walther 2016, 319). This area of research quickly lost its footing — Klapper was lampooned with a procession of examples with “not‐so minimal effects” (Iyengar, Peters, & Kinder, 1982, via Neuman and Guggenheim 2011, 172), and the field began embracing an era of new models which can be loosely called “moderate effects” (Greenberg and Salwen 2009, 67).

Moderate effects, agenda setting in particular (Bennett and Iyengar 2008, 708), is the prevailing view of media effects research. It envelopes models like “agenda setting, knowledge gap, [and] gatekeeping” (Greenberg and Salwen 2009, 67). However, these models are as much an artifact of their time and place as the models from the 1920s (Bennett and Iyengar 2008, 707). Does agenda setting pass muster when the number of media sources is orders of magnitude larger than those in 1968? Can knowledge gap make sense of knowledge distribution when digitally‐democratized information with always‐on access presents a total and fundamental upheaval to how information was accessed in the 1970s? Is gatekeeping meaningful when audiences become their own gatekeepers, or when the role is delegated to algorithms? In light of new technology and new paradigms of media use, Bennett and Iyengar argue a return to minimal effects — or, more accurately, a return to the minimal effects paradigm as a perspective for readdressing contemporary models and research (2008, 707 – 8). The old models, prior to Klapper’s minimal effects era, were insufficient to describe a changing media environment, and the minimal effects era sparked the creation of new methods, measures, and models to describe the “strong effects” that were manifesting in new contexts (Bennett and Iyengar 2008, 708). Bennett and Iyengar’s call for a return to minimal effects is, at least in part, a provocation to again generate new models for new contexts. It’s also a callback to that era; an opportunity to do it better, this time considering the social and technological changes within comprehensive models which address the oversights that Bennett and Iyengar consider the greatest failure of the literature from that earlier period (Bennett and Iyengar 2008, 708). In the contemporary landscape, they see society as being hyper‐niche and disconnected from the social institutions which once bore considerable influence (Bennett and Iyengar 2008, 707), and they see the challenges that threaten to break, not just current models, but also the relevance and validity of current designs for media effects data collection and research (Bennett and Iyengar 2008, 724). Central to Bennett and Iyengar’s argument is that audience structure is radically different due to changes in communication technology (2008, 717) which allow audiences to self‐select media (2008, 717) while largely avoiding inadvertent media‐use (2008, 717 – 18). The authors assert that audiences are becoming increasingly politically fragmented through selective exposure and partisan echo‐chambers (Bennett and Iyengar 2008, 717, 719 – 20). In this environment, current measures and models of persuasive media effects begin to fail (Bennett and Iyengar 2008, 724) — they show as minimal effects. Holbert, Garrett, and Gleason largely agree about the need for new theorizing, but they also contend that this focus on news and political information hides some unfounded assumptions (2010, 15 – 16). They caution that a preoccupation with examples, questions, and measures based around news sources in the study of political media effects is itself an assumption to be questioned (2010, 15, 17 – 18, 31). The critical discourse between the groups reveals that Bennett and Iyengar aim to uncover the field’s assumptions at‐large (2010, 38) and that their true concern is current research methods and models being insufficient to describe and measure the effects taking place in the new media environment (2008, 708, 2010, 37).

The media effects research is abundantly clear that effects do occur and that they can be anything from minimal to strong (Bennett and Iyengar 2008, 708; Greenberg and Salwen 2009, 67; Valkenburg and Peter 2013, 222), leading to a new era of theorizing broadly called “moderate” (Greenberg and Salwen 2009, 67). However, this name predisposes the literature to a specific narrative. The models and research implications have moved well beyond minimal versus strong effects, there is no polarity nor debate, yet Neuman and Guggenheim accuse the scholarship of continuously casting these new models against Klapper’s minimal effects (2011, 173). A polarity between minimal and significant effects belies the complex interconnectedness of mass communication, media effects, and audiences. It also belittles Klapper’s contribution to the field, integrating selectivity and nonmedia factors into mass communication studies (Greenberg and Salwen 2009, 66; Neuman 2018, 373). In a field already preoccupied with responding to public fears regarding harmful media effects (Potter 2011, 897), this polarity (rather, the narrative of it) is a false dichotomy between one‐dimensional options when a greater breadth of perspectives exist. Even with the addition of a “moderate effects” perspective, this narrative conflates what is significant (in the sense of noteworthy, non‐trivial) with what is measurable (Neuman and Guggenheim 2011, 172), and it hides an assumption that minimal and significant are mutually exclusive when this need not be the case. Conceivably, effects may be both minimally measurable through current methods and models while still having significant consequences on media audiences. Conversely, if we imagine a pair of Maximal Effects — to abuse the language — the first may weaken over time, while the other appears as a flash‐in‐the‐pan. Neuman and Guggenheim strive to move the literature beyond this fictitious debate, and they chastise the field for perpetuating and engaging in the narrative:

[T]he minimal‐effects/significant‐effects polarity we believe is a demonstrable impediment to the design and interpretation of media effects. […] It would appear that even after 50 years, simply to demonstrate a statistically significant effect in the ongoing battle against the vestiges of Klapper’s evil empire is sufficient justification for celebration and publication. (2011, 173)

And, Neuman later adds to this line of reasoning:

Perhaps our paradigm would be strengthened if we recognized that media effects are neither characteristically strong nor are they characteristically minimal: they are characteristically highly variable. (2018, 370)

The narratives of polarity and of conquering minimal effects are simplistic comforts in the face of the crucial issues and limitations that the field must contend with. It is time for media effects researchers to put them aside.


Literature reviewed
  • Bennett, W. Lance, and Shanto Iyengar. 2008. “A New Era of Minimal Effects? The Changing Foundations of Political Communication.” Journal of Communication 58 (4): 707 – 31. https://doi.org/10.1111/j.1460 – 2466.2008.00410.x.
  • — —  — . 2010. “The Shifting Foundations of Political Communication: Responding to a Defense of the Media Effects Paradigm.” Journal of Communication 60 (1): 35 – 39. https://doi.org/10.1111/j.1460 – 2466.2009.01471.x.
  • Chaffee, Steven H., and Miriam J. Metzger. 2001. “The End of Mass Communication?” Mass Communication and Society 4 (4): 365 – 79. https://doi.org/10.1207/S15327825MCS0404_3.
  • Geddes, Norman Bel. 1932. Horizons. Boston, Little, Brown, and Company. http://archive.org/details/horizons00geddrich.
  • Greenberg, Bradley S., and Michael B. Salwen. 2009. “Mass Communication Theory and Research: Concepts and Models.” In An Integrated Approach to Communication Theory and Research, edited by Don W. Stacks and Michael Brian Salwen, 2nd ed, 61 – 74. Communication Series. Communication Theory and Methodology. New York: Routledge.
  • Holbert, R. Lance, R. Kelly Garrett, and Laurel S. Gleason. 2010. “A New Era of Minimal Effects? A Response to Bennett and Iyengar.” Journal of Communication 60 (1): 15 – 34. https://doi.org/10.1111/j.1460 – 2466.2009.01470.x.
  • Neuman, W. Russell. 2018. “The Paradox of the Paradigm: An Important Gap in Media Effects Research.” Journal of Communication 68 (2): 369 – 79. https://doi.org/10.1093/joc/jqx022.
  • Neuman, W. Russell, and Lauren Guggenheim. 2011. “The Evolution of Media Effects Theory: A Six‐Stage Model of Cumulative Research.” Communication Theory (1050 – 3293) 21 (2): 169 – 96. https://doi.org/10.1111/j.1468 – 2885.2011.01381.x.
  • Potter, W. James. 2011. “Conceptualizing Mass Media Effect.” Journal of Communication 61 (5): 896 – 915. https://doi.org/10.1111/j.1460 – 2466.2011.01586.x.
  • Rubin, A. M. 2009. “Uses‐and‐Gratifications Perspective of Media Effects.” In Media Effects: Advances in Theory and Research, edited by Jennings Bryant and Mary Beth Oliver, 3rd ed, 165 – 84. Communication Series. Communication Theory and Methodology. New York: Routledge.
  • Valkenburg, Patti M., and Jochen Peter. 2013. “The Differential Susceptibility to Media Effects Model.” Journal of Communication 63 (2): 221 – 43. https://doi.org/10.1111/jcom.12024.
  • Valkenburg, Patti M., Jochen Peter, and Joseph B. Walther. 2016. “Media Effects: Theory and Research.” Annual Review of Psychology 67 (1): 315 – 38. https://doi.org/10.1146/annurev-psych-122414 – 033608.

Notes from Love Data workshop

Since Hurricane Harvey, Rice University’s Kinder Institute for Urban Research has been collecting data and making it available to the public along with the tools and resources to use it.

They’ve created two repositories and sets of tools:

Houston Community Data Connections (HCDC)
  • datahouston.org
  • Targeted to non‐expert community officers
  • Hosts webinars and in‐person training
Kinder Urban Data Platform
Upcoming Events:
HCDC Data Talk: Understanding Gentrification in Harris County
Th, Feb 21, 1 – 2 pm
Online webinar
(urban disparity, urban planning)
 
HCDC Data Talk: Transportation, Infrastructure and Safety Concerns
Th, Apr 18, 1 – 2 pm
Online webinar
(placemaking, urban planning)
 
Urban Reads: I‐45 Meets the Walkable City
Feb 27, 2019
7:00 pm to 8:30 pm
Lecture
MATCH (Midtown Arts & Theater Center Houston) — Matchbox 4
3400 Main Street
(transportation, urban planning)
 
The Future of Urban Mobility
Apr 11, 2019
7:00 pm to 8:30 pm
Panel
Bioscience Research Collaborative
6500 Main Street
(transportation)

Data collections
Library Databases and Data Resources
Social Explorer
  • current/historical census data
  • business patterns
  • health
  • crime
ICPSR
  • 500,000 social science research data
  • public use and restricted use data
  • different formats available
  • Learning: classroom exercises for teaching and resources for students.
  • Youtube channel with guides on using data
Gallup Analytics
  • Data from countries that are home to more than 98% of the world’s population
  • US Daily tracking and World Poll data to compare responses
  • Library has access to raw Gallup data
ReferenceUSA
  • Business data by name, industry, location, or a combination
  • Closed and historical business data
  • Longitude and latitude available for locations, ready for mapping
SimplyAnalytics
  • US demographic, business, marketing data
  • web interface for making maps, reports, and to cross‐compare data between geographic locations
  • Data is downloadable
HathiTrust
  • Humanities data
  • Digital Library with 16.7 million volumes
  • Provides tools for text mining
JSTOR Data for Research
  • provides datasets from JSTOR researchers
  • Define and submit desired dataset to be automatically processed
  • Metadata, n‐grams, and word counts for most content in JSTOR
  • No cost to researchers, includes data up to 25,000 requests
Digital Collections as Data
guides.lib.uh.edu/data
 
UH Library
digital.lib.uh.edu
site hosts digitized materials from special collections
 
Big 10 Academic Alliance
geo.btaa.org
 
Types of data:
  • cultural heritage
  • geospatial data
  • bibliographic data
  • text and images
Getting library metadata
How to access?
  • Ask UH librarian
  • Download from site
  • Find external libraries that offer metadata downloads and metadata profiles
OAI‐PMH
Open Archives Initiative Protocol for Metadata Harvesting
www.openarchives.org/oai
 
Digital Public Library of America
http://dp.la/
 
Always Already Computational‐ Collections as Data
collectionsasdata.github.io
 
Audiovisual Archives as Data
 
Archives of digitized films and videos
  • UH’s A/V Repository: av.lib.uh.edu
  • Kentuckyoralhistory.org
Netherlands Institute of Sound and Vision
  • FAIRview
    granular
  • MediaNow
    semantic search of media
  • MediaDNA
    media fingerprinting and tracking
    goal of media citation tracking

Debates in Media Effects Literature

The biggest take away from the contemporary scholarship is that Media Effects seems to be having its “Expanding Field” moment (a term I know from art history; see Krauss, 1979). Shit’s getting complicated, and things can’t be adequately explained without venturing into the theoretical frameworks of other disciplines or redefining some of the field’s foundational terms.

Following up the criticism from Holbert et al. to not toss the baby out with the bathwater, my instinct is to think that the discrepancies between Media Effects models that can describe effects and uses in earlier periods and the ones needed to describe today’s as requiring some overarching model or theory that can adequately explain effects from both periods. However, I don’t know if that’s possible, already happening, or within the scope of media effects study. The human ecological factors that audiences feel likely have confounding influences on media use, as the differences in early Modern (1890 – 1930) and late Modern (1930 – 1960) periods show — media use that was most informed and guided by social, political, and economic factors like urbanization and mass migration (Chaffee & Metzger, p. 367).

A second, counter‐perspective also comes to mind. To continue the parallel that Neumann draws between Media Effects models and the Heliocentric and Geocentric astronomical models, the Geocentric Model didn’t cease being useful for navigation, timekeeping, or for telling you where in the sky to point a telescope. It is important to note that a model doesn’t need to represent truth to be useful; it simply must have its limitations qualified. Maps and diagrams are another perfect example of this. Michael Beirut has a great video lecture (“The genius of the London Tube Map”) that explains how a geographically accurate map of London’s Underground train system was challenging to read and understand, but when they trialed a map which was abstracted from geography, it was instantly successful and quickly became the world standard for public transit maps.

Between this mess of mixed notions that I have, I think the key to evaluating any model is its simplicity and utility.

I think that there is a certain amount of alarmism around media fragmentation. While, yes, new technology is creating the infrastructure for extreme selectivity and individualized media use, there must be a finite ability for a population to create desirable, gratifying media content with diminishing quality and gratification for audiences and diminishing revenue and resources for creators as the audience narrows. Eventually, this media and audience fragmentation will equalize; though it may be at different levels depending on the nature of the platform. For example, I have a lower tolerance for low‐quality YouTube videos than I do for low‐quality links and discussion on Reddit. It takes more effort to evaluate the quality of a video (being a linear, durational medium), while Reddit’s text comments take minimal effort to skim and skip over. But in both platforms, I regularly reach the limits of the desirable content, and either become less‐selective or move onto other things. The only thing which concerns me are echo‐chambers and Eli Pariser’s Filter Bubbles, and not because they might completely isolate but because they might legitimize fringe ideas and largely seem to be passive, unintentional consequences of algorithms. Being driven by algorithms optimizing for clickthroughs and ad revenue, filter bubbles have the power to make us more extreme versions of ourselves in strange and unexpected ways.

Literature reviewed
Bennett, W. Lance, and Shanto Iyengar. 2008. “A New Era of Minimal Effects? The Changing Foundations of Political Communication.” Journal of Communication 58 (4): 707 – 31. https://doi.org/10.1111/j.1460 – 2466.2008.00410.x.
— —  — . 2010. “The Shifting Foundations of Political Communication: Responding to a Defense of the Media Effects Paradigm.” Journal of Communication 60 (1): 35 – 39. https://doi.org/10.1111/j.1460 – 2466.2009.01471.x.
Chaffee, Steven H., and Miriam J. Metzger. 2001. “The End of Mass Communication?” Mass Communication and Society 4 (4): 365 – 79. https://doi.org/10.1207/S15327825MCS0404_3.
Holbert, R. Lance, R. Kelly Garrett, and Laurel S. Gleason. 2010. “A New Era of Minimal Effects? A Response to Bennett and Iyengar.” Journal of Communication 60 (1): 15 – 34. https://doi.org/10.1111/j.1460 – 2466.2009.01470.x.
Neuman, W. Russell. 2018. “The Paradox of the Paradigm: An Important Gap in Media Effects Research.” Journal of Communication 68 (2): 369 – 79. https://doi.org/10.1093/joc/jqx022.