Attention and Media Research
The following is an adaptation of a paper I wrote at the request of my advisor. This paper is an overview of attention and mass communication research. As such, findings from cognitive science and neuroscience literature are discussed as a means for coming up with meaningful media research questions. Accordingly, much of this post is actually a brief review of literature and methods. A shorter section is devoted to research opportunities (expect more of this in upcoming posts).
Attention is one of the oldest and most studied areas within the behavioral sciences (D. Anderson & Kirkorian, 2006). While modern inquiry dates back to the mid 1800s, attention can be traced back to Aristotle and Lucretius (D. Anderson & Kirkorian, 2006; Potter & Bolls, 2011). Accordingly, one might reasonably expect that there is a universally shared definition of what exactly attention is. As early as 1890 it was assumed that “everyone knows what attention is” (James, 1890, p. 403). However, this simplistic characterization does little to further our understanding of attention.
Decades of inquiry have yielded multiple, and sometimes competing, conceptions of attention. This paper will contemplate modern attention research, considering why the study of attention is important, both generally, and within the field of communication. Common methods utilized to study attention will be considered, as well as ways in which attention is typically operationalized. Finally, this paper will discuss new directions for interactive media research.
What is Attention?
If attention is so universally understood, the question “what is attention” seems inconsequential. Why bother with defining that which is already well known? But consider for a moment exactly what attention is. While thinking about this question, ask yourself: is attention the same as awareness, as memory, as consciousness? If these concepts are different, on what dimensions are they different? Our “shared” understanding of attention is becoming less clear.
While attention, awareness, memory, and consciousness are all directly related, each is distinct. More precisely, attention, awareness, and memory are three subsystems of consciousness (Potter & Bolls, 2011). In a brief review, Csikszentmihalyi (1988) explains that attention allows us to recognize information, awareness is our ability to interpret information, and memory is our ability to store information. Consciousness is our ability to focus, construct meaning, and store details related to a given stimulus. Without attention, the consciousness would not be able to recognize information. Moreover, without all three of these systems, the consciousness would lack control over individual behavior (p. 20-21). Finally, while these three subsystems can operate concurrently, they do not always operate congruently. For example, the act of attending to a stimulus does not mean we actively commit it to memory (Potter & Bolls, 2011, p. 69). Related to media research, just because attention is a precondition for entertainment does not mean it is not the same thing as entertainment (D. Anderson & Kirkorian, 2006, p. 35).
While the above paragraph helps explain what attention is not, it does little to explain what attention is. Unfortunately, there is not an all-encompassing definition of attention. However, Anderson & Kirkorian (2006) offer one of the better definitions. “The term attention refers to a psychological process by which information, usually from the external environment, is made available for cognitive and emotional analysis” (p. 35). An important component omitted from this definition is that attention is a finite resource. At any given moment, we encounter more information than we can possibly attend to (for a full review, see Stroessner & Hamilton, n.d.). Thankfully, there are several important ways in which we cope with this overabundance of information.
Conceptualizations of Attention
As mentioned earlier, attention has a long history of empirical inquiry. Within this scholarly tradition, cognitive scientists have made several important contributions (see Appendix for a historical overview of attentional research). Generally speaking, cognitive science considers questions related to different types of attention such as: vigilance, orientation, attention to objects, and location of attention (D. Anderson & Kirkorian, 2006, p. 35). These questions are often tested in a relatively short timeframe (e.g. less than a second), and require participants to engage in intrinsically unrewarding tasks. As such participants in these studies are compelled, or motivated to complete a specific task.
While cognitive science research has allowed for advancements in our understanding of attention, the constraints in which attention is studied has limited the applicability of these findings to mass communication research (D. Anderson & Kirkorian, 2006). One reason for this is that each field approaches attention research with different research questions. For instance, communication scholars tend to ask questions that consider attention in increments longer than a split second, and among subjects who may very well intrinsically enjoy the activity they are participating in. Communication research on attention has considered questions as diverse as: the impact of ingroup/outgroup on attention, visual salience and attention, personally relevant messages and attention, and social context on attention (for a extensive overview, see Stroessner & Hamilton, n.d.).
Attention and mass communication. Pertaining to mass communication, scholars generally focus on four key aspects of attention: onset, continuation, intensity, and offset (D. Anderson & Kirkorian, 2006). When observing onset and offset, researchers ask questions such as: what features of a message cause an individual to attend to a TV show and what causes this person to stop attending to the TV? Answering these questions typically involves observing looking behavior, which will be discussed in greater detail in the methods section of this paper. For now, we know that movement on a screen triggers looking, that we are more likely to look after a cut scene takes place, and that overly complex content can cause us to stop attending to mediated messages (see D. Anderson & Kirkorian, 2006 for a review of literature). Similarly, we know that there are several looking strategies (i.e. monitoring, orienting, engaged, stares) we employ when watching television, each with important implications for memory and recall (Hawkins et al., 2005).
The continuation of attention to mediated messages is somewhat intertwined with the amount of resources we allocate to attending to a given stimulus (Lang, 2000, 2006). The continuation of attention will be discussed greater detail in subsequent sections of this paper. At present, it is meaningful to point out that we attend to trivial and significant messages depending on motivational factors as well as the amount of cognitive resources available at a given moment (D. Anderson & Kirkorian, 2006; Potter & Bolls, 2011).
Finally, media scholars consider the intensity of attention to mediated messages. Typically, attentional intensity is measured by assessing secondary task response times (STRTs) as well as physiological measures such as cardiac activity (both methods will be discussed in greater detail in subsequent sections of the paper). In a review of related literature, Potter and Bolls (2011) note that formal features, such as cuts, are associated with decreases in heart rate and increases in response times (RTs) to a secondary task; both indicators of greater attention. Similarly, arousing messages yield increased RTs and exposure to negative messages result in decreased hart rate. Beyond message characteristics, larger TV screens have been shown to produce greater attention and arousal than smaller screens (Reeves, Lang, Kim, & Tatar, 1999). By now it should be clear that, when it comes to understanding how media impacts attentional intensity, “it’s not just what you say, but how you say it” that matters (Potter & Bolls, 2011, p. 66).
One of the ways we attend to a multitude of information is by lumping stimulus together into chunks of information (Csikszentmihalyi, 1988). For instance, when speaking with a partner, we do not specifically attend to the unique sound of each syllable. Instead, we mentally combine these syllables together to formulate words. While words are still a relatively small unit of information, this chunk is an order of magnitude larger than the individual parts. This represents an important dichotomy associated with information we attend to. As chunk size increases, the level of detail decreases (Stroessner & Hamilton, n.d.). This concept extends beyond verbal attention. When attending to visual stimuli, say watching a movie, we chunk actions into groupings of various sizes (e.g. villain runs down a flight of stairs, villain rounds a corner while rapidly descending a flight of stairs, while rounding a corner as he descends a flight of stairs villain looks over his shoulder to see that he is not successfully escaping from the heroine).
The above example hints at a second aspect of attention. Attention is selective. At any given moment, there is more information than we can possibly process, so we focus on a limited subset of stimuli (Potter & Bolls, 2011). This process, known as selective attention, is analogous to a spotlight or zoom-lens (Handy, Hopfinger, & Mangun, 2001). Since attention is a limited resource, there is a cognitive benefit for the attended stimulus, and a cognitive cost for unattended stimuli. With that said, the spotlight analogy is only so helpful as it implies that attention is directed at only one stimulus at any given moment. This is not true; we have the capacity to attend to multiple inputs at any given time. However, as the cognitive requirements associated with these inputs increase, our ability to successfully attend to multiple tasks decreases. For example, most of us can listen to music while reading neuroscience literature, but we might struggle to read neuroscience literature while having a meaningful conversation with a friend. Here, the spotlight model would predict that we would selectively attend to either the reading, or the conversation, but not both.
Until now, this paper has considered attention primarily from a motivational perspective. We are cognitively motivated to lump information together and selectively attend to a given stimulus due to our limited capacity for attention. But this does not provide a complete perspective on how we attend to stimuli. Scholars are now advocating an approach that considers attentional allocation as the result of the perceived significance of both motivational and emotional factors (Potter & Bolls, 2011). Thinking back to the example about reading neuroscience and talking with a friend will help elucidate this point. Since both are cognitively challenging tasks, a selective attention framework tells us that we are cognitively motivated to focus on our reading, or on the conversation. But how will we choose what to focus on, and can we assume that we always make the same decision? Understanding the way emotions motivate our attentional process will help answer this question. Assume that you are reading neuroscience literature because of an impending exam. In this instance, you may very well opt to ignore the conversation and continue attending to the reading. This example demonstrates that, even though our cognitive processing may differ from our emotional processing, it is impossible to have one without the other (p. 66-67).
This dual model is useful for demonstrating how attentional allocation is an evolutionary construct. Since the onset of the human species, that which we experience was actually represented in our physical environment. One hundred and fifty years ago, being chased by a tiger was a very real experience. Today, TV shows, movies, and videogames dramatically alter what it means to be chased by a tiger. Whereas being chased by a tiger 150 years ago might have been a traumatic experience that we would avoid at all costs, we might now actively choose to be chased by a tiger if this is an integral component of our favorite videogame. The important thing to consider is that, from a purely cognitive perspective, our brain processes the tiger in real life and the tiger on the screen in a similar way (Reeves & Nass, 1996). Our emotional processing allows us to make sense of the two tigers: run or you will die or, keep playing the videogame – this is fun. This is also an important component related to the suspension of disbelief. Being chased by a tiger in a videogame can feel qualitatively similar to being chased by a tiger in real life (Reeves & Read, 2009, p. 194).
As noted to earlier, we attend to stimuli for various lengths of time and at differing levels of intensity. When sitting in a movie theater, we might pay minimal attention to onscreen advertisements and trailers. Conversely, once the movie starts, we might focus intently by allocating substantial resources to a gripping drama (after all, you paid to see the movie, not the advertisements). Another concept to consider here is referred to as attentional inertia (D. Anderson & Kirkorian, 2006). At onset, attention is easily disrupted. When watching movie trailers, we are easily distracted by everything from the person in front of us texting, to our neighbor shuffling in his seat, to our own thoughts about how annoying movie trailers are. But once the movie starts, we allocate more resources to processing the messages on screen (assuming we find the movie captivating). Soon, we stop noticing the person in front of us, or our fidgety neighbor. As we allocate more cognitive resources to attending to a specific stimulus, it becomes increasing difficult for competing stimuli to distract us. Our attention to the movie has developed a sort of self-perpetuating inertia. Attentional inertia is the metaphorical “glue” that keeps us engaged with what is happening on the screen (p. 50).
Controversies Associated With Studying Attention
Given the long and prolific history of attentional research, one might assume that canonical assumptions, terminologies, typologies, and methodologies are fully realized. To the contrary, a brief review of research on attention is reveals that the field is disjointed, disorganized, and fraught with competing terminologies. While these assessments might be applied to many fields of study, a recent critique contends that current conceptualizations of attention are largely misguided (B. Anderson, 2011). B. Anderson contends that decades of research, including recent advancements in physiological and fMRI measures, have dramatically improved our understanding of the “how of attention” but have done little to explain what attention is (p. 2).
B. Anderson’s (2011) critique of attentional research is rooted in three key concerns. First, the field treats attentional research as if it can be understood by examining binary research questions. He contends that the litany of dichotomies readily observable in attention research (e.g. voluntary vs. involuntary, automatic vs. controlled, pre-attentive vs. attentive) are better understood not as categorical variables, but as continuous variables. Second, research treats attention as a cause, when in actuality attention is an effect. This reification of attention is meaningless when the attentional construct is invoked to explain empirical observations we do not fully understand (p. 4). Finally, B. Anderson contends that research on attention all too often treats facts as synonymous with theory. To demonstrate this point, he remarks: “there is no EKG long enough to tell us what love is, nor can any number of spike trains tell us what attention is” (p. 4). As a resolution, B. Anderson advocates for theory based research and analysis, and poses Bayesian modeling as a way forward.
B. Anderson is not the only researcher to express concern at the lack of theory associated with attention research. Among media researchers, there is a persistent concern that attentional research is often atheoretical, and occasionally rooted in media industry conventions (D. Anderson & Kirkorian, 2006, p. 51; Potter & Bolls, 2011, p. 64). These concerns are of critical importance. At issue is a field that seems to lack a clear distinction between what research questions expand our understanding of the brain, and those that utilize what we already know about the brain to expand our understanding of communication theories. While both are important pursuits, the role of mass communication scholars remains poorly defined, and wanting for direction.
Merits of Studying Attention
Despite the aforementioned concerns, a robust understanding of attention is critical to our understanding of media research. Attention is the means by which we come to interpret media messages and the first component in a system that enables consciousness. And, if B. Anderson (2011) is to be believed, there is much ground to cover as we work to improve our fundamental understanding of attention.
Moreover, much of the existing media and attention research has focused on more passive forms of entertainment such as TV and radio. This research has done much to elucidate the way we attend to messages on a TV screen and serves as an important contribution to the field of media research. However, comparatively new forms of entertainment, such as interactive media, seemingly challenge many of the assumptions underlying traditional attention research.
For instance, we know that individuals do not consistently attend to TV. In fact, TV viewers may look away from the screen as many as 120-150 times in a single hour of viewing (D. Anderson & Kirkorian, 2006). When individuals do attend to messages on a TV screen, they employ several looking strategies, some as brief as a quick glances, and others as long as stares (Hawkins et al., 2005). One simple, but interesting question tests our assumptions about looking at mediated messages: does our understanding of looking change when we observe videogame players? It seems unlikely that focused videogame players will look away from the screen as frequently as TV viewers. But this does not mean that our understanding of looks should be thrown out. Instead, it should serve as a guide for future empirical work. Thinking about videogame players, we might ask where they look on the screen and what sort of monitoring strategies they employ for attending to dynamic content.
Despite decades of research, we have just scratched the surface of our understanding of attention. In the next few sections, this paper will examine the ways we measure attention, including a brief discussion on the neural correlates of attention. This should help equip the reader with a basic overview of how to conduct attention research. Finally, with a methodological understanding of attention, future media research opportunities will be discussed.
Attention – A Triparted System
We know that attention, awareness, and memory are the three subsystems of consciousness. But our understanding of attention does not end there. A widely held perspective is that attention can be further broken down into three additional systems: alerting, orienting, and executive (Raz & Buhle, 2006). Each of these systems, including the relevant neural correlates, will be considered in greater detail in the following sections.
There is some controversy about which networks have received the most or least empirical observation. For now, it seems reasonable to say that, across all the fields researching attention, the alerting network has received the least study (Raz & Buhle, 2006). Nevertheless, advancements in brain imaging are allowing scientists to map the neural correlates of the alerting network. When observing the alerting network with an fMRI machine, one typically sees activation in the frontal lobe as well as the parietal lobe. This activity is generally constrained to the right hemisphere.
This network is responsible for alerting us to new stimuli and can be broken down into two subcomponents: phasic alertness and intrinsic alertness (Raz & Buhle, 2006). Phasic alertness represents our ability to respond to a goal directed task and intrinsic alertness represents our general levels of alertness. To clarify the distinction between phasic and intrinsic alertness, think back to the participants in cognitive science experiments that have been motivated to complete a tedious, or otherwise intrinsically unrewarding, task. A classic experiment, known as a vigilance task, asks research participants to stare at a computer monitor and press a button when an “X” appears on the screen. In the condition that tests phasic alertness, participants are warned moments before the “X” appears. In the intrinsic alertness condition, participants are not warned before the “X” appears. Comparison between these two groups suggests that those in the phasic alertness condition have faster RTs, and are less prone to making errors (p. 372 – 373).
While this sort of experiment is designed to separate phasic and intrinsic alertness, the exact distinction between these two systems is not entirely clear (Raz & Buhle, 2006, p. 372). Initial brain imaging research has observed some differences between the two systems, suggesting that phasic alertness is motivationally driven and intrinsic alertness is related to arousal. If this is the case, this seems to add legitimacy to the assertion that the study of attention must consider both motivational and emotional factors (Potter & Bolls, 2011).
If alerting is the least studied of the attentional networks, then orienting is probably the most studied (Raz & Buhle, 2006). Once we are alerted to stimuli, the orienting network allows us to select and focus on a specific stimulus. One should start to recognize parallels between the orienting network and our limited capacity to attend to multiple stimuli (Lang, 2000, 2006). The link between the orienting network and selective attention should also become clear (Handy et al., 2001). Oversimplified for ease of discussion, we can either pay a small amount of attention to a large number of things, or a great deal of attention to a small number of cognitively taxing things. The orienting network controls our ability to selectively scan a multitude of stimuli, or selectively focus on a specific stimulus.
When observing the orienting network with an fMRI machine, one typically sees activation in the superior parietal lobe, frontal eye fields (V1), and temporal lobe. Activity is also observable in the temporoparietal junction (TPJ), pulvinar nuclei, and superior colliculus. Like the alerting network, the orienting network can be broken down into two subsystems: endogenous and exogenous orienting (Raz & Buhle,2006). The distinction between the two can be conceptualized as guided vs. unguided. To demonstrate, lets consider a classic visual orienting experiment where participants are asked to attend to a computer monitor and press a button when an “X” appears on the screen. In the endogenous condition, an arrow points the region where the “X” will appear on screen. In the exogenous condition, no arrow is offered. Comparison between the two conditions reveals that those in the endogenous condition have increased RTs and less error than those in the exogenous condition (p. 373).
This experiment also demonstrates the difference between bottom-up and top-down processing. Bottom-up processing is the combined interpretation of multiple senses, or pieces of data, on our understanding of our external surroundings. Top-down processing is when our expectations, or the knowledge we possess impacts (and sometimes overrides) our bottom-up processing. As demonstrated by this visual orientation experiment, the exogenous condition tests our bottom-up processing (Raz & Buhle, 2006). We distinguish where on screen the “X” will appear only after combining multiple inputs. Conversely, the endogenous condition informs us as to where we should focus our attention, giving us additional knowledge beyond what our senses perceive.
The final system to consider is the executive network. If alerting network is the least studied & orienting is the most, the executive networks is sandwiched in-between the two (Raz & Buhle, 2006). This network controls our planning, decision-making, and error detection abilities. As such, a broad view conceptualizes the executive network as any top-down response. A more limited view insists that the executive network is any function that monitors and resolves conflict among bottom-up sensory information. When measured with an fMRI machine, executive network activity is generally observable in the anterior cingulate cortex (ACC).
The executive network is better understood though the use of a conflict experiment known as the Stroop task (Raz & Buhle, 2006). In this experiment participants are exposed to congruent, incongruent, and neutral stimuli. Long reaction times and high error rates are typically associated with incongruent stimuli. Conversely, fast reaction times and low error rates are associated with congruent tasks. The executive network allows the brain to make meaning out of conflicting information.
Relevance for Communication Research
The merits of understanding the neurophysiological structure of attention may not be entirely clear for communication scholars. In fact, much of the discussion around alerting, orienting, and executive control networks seems more focused on elucidating functions of the brain that expanding our understanding of communication. In some instances, this perspective may very well be true, but it is highly dependent on the type of questions we choose to ask. For instance, the Stroop task observes the function of the executive network. But it also demonstrates our ability to make meaning out of incongruent vs. incongruent stimuli. From this task, communication researchers have derived a multitude of experiments designed to test everything from the impact of negatively valenced language on message processing to the interaction between incongruent stereotypes and racial perceptions (for a review of communication studies rooted in a Stroop methodology, see Stroessner & Hamilton, n.d.).
The alerting and orienting networks are equally useful for communication researchers. There are additional cognitively framed questions that media scholars might also consider. For instance, we know that the human brain continues to develop as we age. As such, we might ask what role cognitive development plays in our ability to attend to, interpret, and make meaning of mediated messages. Research exploring young children and TV programming demonstrates that children less than two years old do not learn much by watching television (D. Anderson & Pempek, 2005). Instead, the authors advocate activities such as playing with toys as a far more effective learning tool for young children. Interestingly, studies on more interactive television shows, such as Blue’s Clues, have demonstrated a mild link between watching the show and cognitive development (D. Anderson et al., 2000). One wonders what impact more interactive media, such as videogames, might have on early child learning and development. An understanding of the executive system’s role in negotiating conflicting stimuli might provide a meaningful foundation for such research.
We cannot design cognitively motivated studies to expand our understanding of communication theory without a firm understanding of the neurophysiological underpinnings of attentional networks. Designing these studies will take creativity, and the ability to understand that which neuroscientists observe, and how it relates to media research.
The next section of this paper will consider three methods for measuring attention: eye-tracking, secondary task response time, and cardiac activity. Each is commonly used to test one, or several of the three attentional networks: alerting, orienting, and executive. There are two primary goals of this section. The first is to describe each method while considering its relative strengths and weaknesses. The second is to offer methodological insights as to how one might go about conducting communication research within a cognitive framework.
One of the oldest methods associated with attention research is eye tracking. Simply stated, eye tracking is the process of measuring looks (D. Anderson & Kirkorian, 2006; Hawkins et al., 2005). Typically, this method is used to test research questions associated with the alerting network. Specific to media research, eye tracking can be broken down into two key areas of empirical study: what triggers looks, and what impacts the length of a look? In much of early media research, eye tracking was measured as a binary observation; is the participant looking at the TV, yes or no. This observation made perfect sense in a media landscape where the average household passively enjoyed media on a relatively small TV screen.
However, changes in the media landscape have expanded the scope of eye tracking. Today, it is increasingly common for households to consume media on a variety of screen sizes (both large and small). Interactive media also challenge and extend the nature of eye tracking research. Instead of passively attending to messages on a TV screen, interactive media (such as videogames) require nearly constant visual attention. As such, it is no longer sufficient to simply measure looks at a screen. Instead, modern media research utilizing eye tracking should observe where on the screen participants are looking (D. Anderson & Kirkorian, 2006). Thankfully, research tools and techniques have adapted to this new challenge. Today it is possible to precisely measure where a participant is looking at any given time (see figures 8-9).
While eye tracking is useful for understanding what causes individuals to attend to media, it does little to explain how much attention is devoted to a given stimulus. While some studies have tried to equate looks with depth of attention (see Hawkins et al., 2005), this is an indirect measure at best. Just because one stares at a TV does not mean that they are actively paying attention to the message. Similarly, one might never look at a TV screen but still devote considerable resources to listening to the message. Different research methods are required to measure intensity of attention.
Secondary Task Response Time (STRT)
Secondary task response time (STRT) measures the intensity of attention (D. Anderson & Kirkorian, 2006). More precisely, STRT is theoretically grounded in a limited capacity framework that assumes we have a finite amount of resources available for attending to stimuli (Lang, 2000, 2006; Potter & Bolls, 2011). While this concept was mentioned briefly earlier in the paper, it is worth further consideration now. For any given stimulus, there is a fixed amount of resources required to adequately attend to the stimulus. And, each of us has a limited amount of available resources that can be devoted to attending to stimuli. Finally, there is a specific ammount of resources we allocate to attending to any given stimulus. What STRT measures is the amount of resources available at any given moment (Lang, Bradley, Park, Shin, & Chung, 2006). With this in mind, we can begin to consider how STRTs are measured, and why we might apply this technique to attention research.
Secondary tasks response times were briefly mentioned in this paper when discussing experiments for testing orienting and executive attentional networks. But what exactly are STRTs, how are they measured, and why are they useful for media research? To answer these questions, lets imagine an experiment where a researcher wants to observe how much attention a participant is paying to an anti-smoking advertisement. In this fictional experiment, the researcher hypothesizes that participants who pay more attention will have better recall of the anti-smoking message.
To answer this question, the researcher needs to measure two things, memory for the message, and level of attention paid to the message. Recall is relatively straightforward and can be answered with a questionnaire. But how would the researcher measure level of attention? One method would utilize STRT. In this case, while participants watch the anti-smoking advertisement (the primary task), they are also asked to press a button every time they see an “X” appear on a second monitor (the secondary task). The researcher knows that level of attention is positively related to the amount of resources the participant allocates to the anti-smoking commercial. As this fixed pool of available resources decreases, the researcher would expect a greater delay in completion of the secondary task. Conversely, if the participant pays little attention to the anti-smoking advertisement, they will have a greater pool of available resources to devote to the secondary task. In this case, response times should decrease.
Secondary task research is a widely utilized technique for measuring the intensity of attention to mediated messages (D. Anderson & Kirkorian, 2006; Potter & Bolls, 2011). The appeal of STRT is that research protocols can be rapidly developed and deployed. Moreover, there are a handful of freely available STRT software titles that can be run on most modern computers. This makes STRT a fast, cost effective, way to test research questions associated with the orienting and executive networks. However, STRT is limited in that it only measures cognitive load as related to attention (Lang et al., 2006). There are other physiological responses associated with attention that are unobservable with STRT.
The connection between cardiac activity and attention is not intuitively clear. Potter and Bolls (2011) rightly point out that references to the heart often conjure up thoughts of everything from emotions, to pulse (p. 74). But cardiac activity is actually a strong indicator of how much attention we pay to a given stimulus. As such, it is of considerable utility to media researchers interested studying the orienting and executive attentional networks. A oversimplified explanation of changes in heart rate will help clarify this point.
The heart is impacted by both the sympathetic and parasympathetic branches of the nervous system (Potter & Bolls, 2011). But what are these two systems, and why do they matter to attention researchers? The sympathetic system is often referred to as the “fight or flight system” (p. 74). This system tells our heart to beat faster when exposed to arousing stimuli. The parasympathetic system is associated with orienting responses. Moreover, the parasympathetic and tells our heart to slow down when exposed to new stimuli, even as our sympathetic system tells our heart rate to increase. The two systems operate simultaneously, but one will exert more influence over heart rate than the other. Accordingly, we can interpret increases in heart rate as a defensive response, one that rejects environmental stimuli, and decreases in heart rate as indicative of increased attention. Observations of changes in heart rate allow researchers to consider the impact of emotions and arousal on attention.
There are several ways to measure cardiac activity, the simplest being an observation of heartbeats per minute, or heart rate. This measure observes the mechanical functions of the heart, that is the excitation and contraction of the heart (Potter & Bolls, 2011, p. 77-78). However, cardiac activity is actually an electromechanical process; contractions and excitations are the result of the electrical depolarization and repolarization of the heart. The electrocardiogram (ECG), observes changes in voltage during cardiac activity. Where as pulse is a measure of heart rate, ECG allows for a more precise measure known as heart period. Without getting too technical, heart period represents the time, in milliseconds, between “R-spikes in the QRS complex of the cardiac cycle” as measured by ECG (p. 78-79). Of importance to media researchers, heart rate can be mathematically converted to heart period, and visa versa (Potter & Bolls, 2011). While Potter and Bolls advocate ECG measures of cardiac activity, they admit that measuring heart rate is often sufficient for media research.
In fact, there are three points to keep in mind when considering a research program that utilizes ECG (Potter & Bolls, 2011). The first is cost. ECG requires special equipment as well as highly trained laboratory staff. A second limitation of ECG research is time. The task of hooking research participants up to ECG equipment can be laborious and time consuming. Finally, experiments utilizing ECG are prone to capturing noise in the data. Since voltage changes in the cardiac cycle are rather small, they must be amplified by laboratory equipment. This amplification can add noise to the data. Additionally, ECG requires research participants to be relatively still during measurement as movement also causes noise in the data. These limitations mean that special care needs to be taken in both experimental design, and back-end data cleaning.
Still, measures of cardiac activity are appealing for media researchers. Measuring changes in heart rate allow for observing the impact of arousal on level of attentional processing. This has important implications for both passive and interactive media research. For instance, a recent study measuring heart rate while playing videogames found that we are more aroused by and pay more attention to avatars (characters controlled by a human) than by characters controlled by a computer (Lim & Reeves, 2010).
Cognitive Measures – Concluding Thoughts
By now, it should be clear that media scholars have a well-developed toolkit for researching attention. The next question is, what will we choose to study, and how will we go about studying it? The final section of this paper will consider potential avenues for future research.
Future Directions for Media Scholars
It should be clear by now that there are promising opportunities for media scholars interested in attention research. Until recently, media scholars have lacked the many of the methodological tools necessary to directly observe the impact of media on the brain. As such, many of our theories are built on an assumption that the brain is an unknowable “black box” unworthy of investigation or consideration (Geiger & Newhagen, 1993). But recent advancements in cognitive research methods offer new hope. In fact, a neurophysiological perspective of communication research advocates the use of cognitive frameworks for critiquing existing theory (see Weber, Tamborini, Westcott-Baker, & Kantor, 2009). Here, the goal is fewer, but more powerful, theories.
A New Direction for Interactive Media Research
As mentioned earlier in the paper, much of the eye tracking research focuses on relatively passive media use (such as watching television). This presupposes that the meaningful unit of analysis is whether or not someone is actually looking at the screen. This is a reasonable assumption for observing television viewers, but makes little to no sense when considering interactive media. For instance, success in videogames often requires long periods of sustained looking. As such, observing the onset and offset of looks at a screen is no longer useful.
To demonstrate, consider what World of Warcraft (a highly successful Massively Multiplayer Online Roleplaying Game with over 10 million subscribers) players exerience during a play session (figure 11). Here we see a screen filled with more information than one can possibly observe simultaneously. For those unfamiliar with the game, this screenshot captures what is otherwise real-time (dynamic) information about friendly players, enemies, personal resources, personal rankings, complex mathematical calculations, as well as simultaneous interpersonal/group/intergroup/organizational communicative messages. This is all framed within a narrative offered by the videogame publisher (Blizzard entertainment) that is then shaped by individual player and group level player motivations.
Quickly we realize that the conceptualizations of attention as whether or not one is looking at the screen are no longer useful. Even more complex understandings of look strategies, as considered by Hawkins et al. (2005) start to lose their explanatory meaningfulness. In fact, this one screenshot offers multiple opportunities to expand our understanding of attention in interactive media. We could use eye tracking to understand when a player will look a specific piece of information on the screen, where exactly they will look, why they will look at this piece of information (is the look motivationally or emotionally based), and what will cause the player to attend to a new piece of information. Similarly, one might ask questions related to level of attention. At what point does the complexity of the game overload available resources? How does this impact player enjoyment, learning, or video game design? How is this moderated by cognitive ability? Secondary task measures are well suited to answer these questions. Finally, one might consider the interaction between arousal and attention by measuring the cardiac activity of game players. For instance, how does negatively valenced narrative impact attention? Or, might pseudonymous players be seen as more arousing than known players, and thus receive more attention?
There is considerable opportunity to improve our understanding of existing theory. The above-mentioned questions represent just a small handful of the types of problems we might consider when conducting interactive media and attention research. Attention research is well suited to explaining these, and many more types of questions. It is our responsibility to peek into the black box, thereby utilizing our understanding of cognitive methods to frame inquiry into media research.
Anderson, B. (2011). There is no Such Thing as Attention. Frontiers in Psychology, 2(246), 1-8. doi:10.3389/fpsyg.2011.00246
Anderson, D., & Kirkorian, H. (2006). Attention and Television. In J. Bryant & P. Vorderer (Eds.), Psychology of Entertainment (pp. 35-54). Mahwah, NJ: Lawrence Erlbaum Associates.
Anderson, D., & Pempek, T. (2005). Television and Very Young Children. American Behavioral Scientist, 48(5), 505-522. doi:10.1177/0002764204271506
Anderson, D., Bryant, J., Wilder, A., Santomero, A., Williams, M., & Crawley, A. M. (2000). Researching Blue’s Clues: Viewing Behavior and Impact. Media Psychology, 2(2), 179-194. doi:10.1207/S1532785XMEP0202_4
Csikszentmihalyi, M. (1988). The Flow Experience and Human Psychology. In M. Csikszentmihalyi & I. S. Csikszentmihalyi (Eds.), Optimal Experience: Psychological Studies of Flow in Consciousness (pp. 15-35). New York, NY: Cambridge University Press.
Csikszentmihalyi, M. (1990). Flow: The Psychology of Optimal Experience. New York, NY: HarperCollins Publishers.
Geiger, S., & Newhagen, J. (1993). Revealing the Black Box: Information Processing and Media Effects. Journal of Communication, 43(4), 42-50. doi:10.1111/j.1460-2466.1993.tb01303.x
Handy, T. C., Hopfinger, J. B., & Mangun, G. R. (2001). Functional Neuroimaging of Attention. In R. Cabeza & A. Kingstone (Eds.), Handbook of Functional Neuroimaging of Cognition (pp. 75-108). Cambridge, MA: The MIT Press.
Hawkins, R. P., Pingree, S., Hitchon, J., Radler, B., Gorham, B. W., Kahlor, L., Gilligan, E., et al. (2005). What Produces Television Attention and Attention Style? Human Communication Research, 31(1), 162-187. doi:10.1111/j.1468-2958.2005.tb00868.x
James, W. (1890). The Principles of Psychology, Volume 1. Mineola, NY: Dover Publications.
Lang, A. (2000). The limited capacity model of mediated message processing. Journal of Communication, 50(1), 46-70. doi:10.1111/j.1460-2466.2000.tb02833.x
Lang, A. (2006). Using the Limited Capacity Model of Motivated Mediated Message Processing to Design Effective Cancer Communication Messages. Journal of Communication, 56(s1), S57-S80. doi:10.1111/j.1460-2466.2006.00283.x
Lang, A., Bradley, S. D., Park, B., Shin, M., & Chung, Y. (2006). Parsing the Resource Pie: Using STRTs to Measure Attention to Mediated Messages. Media Psychology, 8(4), 369-394. doi:10.1207/s1532785xmep0804_3
Lim, S., & Reeves, B. (2010). Computer agents versus avatars: Responses to interactive game characters controlled by a computer or other player. International Journal of Human-Computer Studies, 68(1-2), 57-68. Elsevier. doi:10.1016/j.ijhcs.2009.09.008
Potter, R. F., & Bolls, P. (2011). Psychophysiological Measurement and Meaning: Cognitive and Emotional Processing of Media. New York, NY: Routledge.
Raz, A., & Buhle, J. (2006). Typologies of attentional networks. Nature reviews. Neuroscience, 7, 367-379. doi:10.1038/nrn1903
Reeves, B., & Nass, C. (1996). The Media Equation: How People Treat computers, Television, and New Media Like Real People and Places. Stanford, CA: CSLI Publication.
Reeves, B., & Read, J. L. (2009). Total Engagement: Using Games and Virtual Worlds to Change the Way People Work and Businesses Compete. Boston, MA: Harvard Business Press.
Reeves, B., Lang, A., Kim, E. Y., & Tatar, D. (1999). The Effects of Screen Size and Message Content on Attention and Arousal. Media Psychology, 1(1), 49-67. doi:10.1207/s1532785xmep0101_4
Sherry, J. (2004). Flow and Media Enjoyment. Communication Theory, 14(4), 328-347. doi:10.1111/j.1468-2885.2004.tb00318.x
Stroessner, S., & Hamilton, D. (n.d.). Attention: The Mind’s Eye Meets the Social World. Social Cognition. Thousand Oaks, CA: Sage Publications, Ltd.
Weber, R., Tamborini, R., Westcott-Baker, A., & Kantor, B. (2009). Theorizing Flow and Media Enjoyment as Cognitive Synchronization of Attentional and Reward Networks. Communication Theory, 19(4), 397-422. doi:10.1111/j.1468-2885.2009.01352.x
Timeline | A sketch of selected landmarks in the study of attention (Raz & Buhle, 2006, p. 368-369