The Activation Model of Information Exposure

The activation model of information exposure (AMIE) (Donohew, Lorch, & Palmgreen, 1998) provides the basis for explaining the functional mechanism of MSV and PMSV. The AMIE contends that attention is a function primarily of an individual’s level of need for stimulation. Specifically, the activation theory claims that (a) individuals have biologically based optimal levels of activation or arousal that they have the most desire of, and (b) individuals expect information exposure to achieve or maintain this optimal state (Morgan et al., 2003). If a message doesn’t elicit this optimal state, audiences are likely to turn away and seek another source of stimulation to help them achieve their desired state. The match between the individual’s ideal arousal level and sensation value of the messages determines in part the resulting attention and other persuasive outcomes to the stimulus (Harrington et al., 2003).

The AMIE recognizes the importance of messages’ novelty and stimulation; it also focuses on audience characteristics (e.g., differences in sensation seeking) in producing effective media messages (Donohew et al., 1998). The model holds that individuals have an optimal level of arousal pending on their sensation-seeking tendency (Zuckerman, 1994). Sensation-seeking peeks in the late teens and early 20s and declines with age (Zuckerman, 1994); it is closely related to a number of risky behaviors such as substance use and driving under influence (DUI) (Zuckerman, 1994). High sensation seekers prefer messages high in sensation value; as Zuckerman (1988) maintains, they are more receptive to novel stimuli because their “optimal level of stimulation depend on the levels set … by novel stimuli” and that “anything producing lower arousal levels may be considered boring” (p. 182).

Ultimately, the activation theory proposes that exposure is a function of the interaction between an individual’s level of sensation seeking and sensation-enhancing attributes of the particular message (Donohew et al., 1998). The model focuses on the attention of the individual based on the assumption that increased allocation of cognitive resources is pivotal for the following higher-order processing to occur. AMIE-driven research suggested that messages capable of generating an optimal level of affective response resulted in greater arousal regarding the mediated information (Donohew et al., 1998). There is ample evidence that arousing ads can be better recalled (Lang, Dhillon, & Dhong, 1995), a finding in line with other studies showing highest recall rates for ads high in both positive and negative emotions (Biener, Ji, Gilpin, & Albers, 2004). Some advertising research also shows that emotional messages are more likely to promote higher-order cognitive processing (e.g., Keller & Block, 1996). Wei and Zhou (2010) applied the AMIE to understand how the effectiveness of the news stories about the bird flu pandemic. They found that PMSV had main effects on attention (as indicated by lower heart rates), arousal (as indicated by higher skin conductance level), fear, apprehension about the massive impact of a bird flu outbreak, and behavioral intention. Moreover, higher PMSV led to more positive perception of the story in terms of clarity, understandability, credibility, being enjoyable, and informativeness (Wei & Zhou, 2010). Taken as a whole, the AMIE-driven research in mediated communication research has demonstrated that high-sensation related components of messages are likely to generate higher levels of arousal and attention.

Perceived Message Sensation Value

Perceived message sensation value (PMSV) refers to how much an individual perceives a message is of high-sensation-value components (Palmgreen et al., 1991, 2002). Based on AMIE, Everett and Palmgreen (1995) conceptualized and developed a PMSV scale representing four hypothesized dimensions: emotional impact, physiological impact, sensory impact, and novelty. It was then used to classify 13 anti-cocaine PSAs as messages either with high or low sensation value. The sample size was very limited in Everett and Palmgreen’s (1995) study to conduct a factor analysis. Later, Stephenson and Palmgreen (2001) conducted an exploratory factor analysis on the 17-item PMSV scale, analyzing reactions of 368 high school students to six anti-marijuana PSAs. Based on this analysis, three factors emerged: emotional arousal, dramatic impact, and novelty. To further validate the three-factor PMSV scale with a more rigorous empirical evaluation, Palmgreen et al. (2002) gathered data from 444 college students who viewed six of the anti-cocaine PSAs adapted from the Everett and Palmgreen’s (1995) study. Confirmatory factor analyses supported the validity of the three-factor PMSV solution for both high and low sensation seekers from the high school sample and the high sensation seekers from the college sample (a lesser fit was obtained for college low sensation seekers) (Palmgreen et al., 2002; Stephenson & Palmgreen, 2001). Taken together, this line of studies has demonstrated that the 17-item PMSV scale is an empirically rigorous measurement. The PMSV scale attained satisfactory reliability and construct validity. In addition, it has been validated using different samples (e.g., young adults and adolescents) with different health communication objectives (e.g., cocaine and marijuana) (Morgan et al., 2003).

The impact of PMSV on message processing has been investigated in a series of studies (Niederdeppe, 2005; Noar, Palmgreen, Zimmerman, Lustria, & Lu, 2010; Stephenson, 2002, 2003; Stephenson & Palmgreen, 2001). These studies typically reveal greater levels of favorable message processing with high PMSV ads, and message processing is positively associated with persuasion (Niederdeppe, 2005). Noar et al. (2010) found a strong relationship of PMSV to perceived message effectiveness; in fact, individuals’ perceptions of the PSAs based on PMSV were much more important to perceived message effectiveness than any demographic variables. Fewer studies have explored the relationship between specific PMSV-enhancing features and message processing and persuasive outcomes. Niederdeppe (2005) coded specific anti-tobacco ads for PMSV-enhancing features, merged these coding results with data from a telephone survey among teens, and tested the relationship between message features and processing. The number of unrelated cuts and the use of suspenseful features (intense imagery and a second-half punch) enhanced message processing among older teens. An additive index encompassing all these features was associated with increased message processing among all teens (Niederdeppe, 2005).

SENTAR: The Sensation-Seeking Targeting Approach

The AMIE is readily applied to sensation seeking and PMSV/MSV. The model sheds insights on campaign efforts aimed at developing messages in a variety of persuasive contexts, especially where reaching high sensation seekers about risk-related behaviors is the goal. Compared to their low sensation-seeking counterparts, high sensation seekers exhibit a stronger willingness to engage in risky behaviors such as drug use and unprotected sex (Zuckerman, 1994). These typical behavioral differences regarding risk behaviors make sensation seeking a useful targeting variable for prevention campaigns. Several health communication campaigns have employed HSV messages to target high sensation seekers. Such sensation-seeking targeting approach is called SENTAR. The action theory (Donohew et al., 1998), upon which SENTAR is based, proposes that an individual’s attention to a message is a function of (a) the individual’s need for sensation and (b) the level of stimulation that is provided by the message. Campaign studies targeting high-sensation seekers using a SENTAR approach have produced behavioral changes, especially in reducing marijuana use among adolescents (Palmgreen, Lorch, Stephenson, Hoyle, & Donohew, 2007).

What is especially important from a health campaign perspective is that high sensation seekers have distinct and consistent preferences for particular kinds of messages based on their needs for novel, complex, and emotionally intense stimuli. They need considerably more novel and powerful messages to attract and hold their attention. Such characteristics are found in HSV messages (Donohew et al., 1998). Several earlier experiments have reported that messages high in sensation value generate more attention and desired behavior change in high sensation seekers, whereas messages low in sensation seeking has the same impact among low sensation seekers (Donohew et al., 1998; Lorch et al., 1994; Palmgreen & Donohew, 2003; Palmgreen et al., 1991). Studies have demonstrated messages deemed high in PMSV enhance attention, recall, and comprehension among high sensation-seeking teens and young adults compared to messages with low PMSV (e.g., Harrington et al., 2003; Palmgreen & Donohew, 2003; Stephenson, 2003; Stephenson & Palmgreen, 2001).

Recent studies, however, suggest that the differences between high and low sensation seekers may not be as pronounced as the activation model (Donohew et al., 1998) predicts. Contrary to the hypotheses by SENTAR, some data do suggest that HSV messages might be effective with low sensation seekers as well (Everett & Palmgreen, 1995; Harrington et al., 2003; Noar et al., 2010; Palmgreen et al., 1991). There has been less evidence for differential effects of PMSV between low and high sensation seekers. High PMSV messages seem to appeal to low sensation seekers in that many studies find a main effect for PMSV but no interaction between PMSV and sensation seeking (e.g., Harrington et al., 2003; Lorch et al., 1994; Palmgreen et al., 2002; Stephenson & Palmgreen, 2001). More recently, Noar et al. (2010) reported that no differences were found in the ability of PMSV in predicting perceived message effectiveness in low- compared to high-sensation seekers. It should be pointed out that the reverse of the case does not hold—low sensation value messages tend not to be effective with high sensation seekers (Noar et al., 2010). Because the focus of health campaigns is often on high sensation seekers, whether HSV messages work well with low sensation seekers is often of limited theoretical or practical interest (Noar et al., 2010).

MSV vs. PMSV: Message Attributes vs. Subjective Reactions

MSV is conceptualized as a set of message features that can function independently or in combination to attract attention (Morgan et al., 2003). Messages with HSV tend to be “novel, creative, exciting, intense, dramatic, or fast-paced” (Morgan et al., 2003, p. 513). As aforementioned, the concept MSV has initially been operationalized as perceived message sensation value (PMSV). PMSV is a useful tool to ascertain that messages contain the features that are considered as sensory and arousing. Because it is a measure of how message features are perceived, it presents as a valuable scale as a manipulation check. Besides high reliability, the PMSV scale is easy to administer and does not require excessive effort to complete (Morgan et al., 2003). The downside is that “the PMSV relies on an individual’s subjective perception of what is high and low in sensation value, requiring extensive screening of message features with target audience members before PSAs are produced” (Morgan et al., 2003, p. 516). Therefore, it may be long and cumbersome if study incorporates many messages. Time and monetary investments are demanding, which is therefore impractical for researchers who are understaffed and underfunded (Stephenson, 1999).

Morgan et al. (2003) pointed out several benefits to the conceptualization and development of the objective measure of MSV. Instead of asking audience members to subjectively evaluate a PSA, campaign practitioners could assess an existing PSA more objectively on formal and content features relevant to sensation value. Further, if campaign designers decide to use focus groups to evaluate campaign messages, this instrument could be used to reduce the number of messages by eliminating those low in MSV, which are very likely to be low in PMSV. Third, a more objective measure might allow the formulation of a more prescriptive/proactive approach to create new PSAs. Consequently, it is valuable to develop a coding scheme (before focus group testing is conducted) that retains the reliability, validity, and effectiveness of the PMSV scale and is highly correlated with PMSV, but is not associated with a large investment of resources (Morgan et al., 2003).

With the benefits and goals in mind, Morgan et al. (2003) sought to determine which message features lead to perceptions of message sensation value. This endeavor should shed light on how theoretical concepts within activation theory can be operationalized and manipulated. It is also conducive for crafting better and more efficient health communication campaigns. Morgan et al. (2003) coded a series of anti-drug PSAs, based on which they proposed three primary types of message features that contribute to MSV. The first dimension is on formal video features, including the number of cuts, visual special effects, and the use of unusual colors, graphic images, and slow motion. The second dimension points to formal audio features, consisting of the use of auditory special effects, saturation of sound, and unusual music. The third is the content dimension, focusing on the use of narrative, whether the action is shown or simply described, the violation of norms (i.e., message presented in a novel way), and the use of a surprise twist at the end. The number of cuts are converted to low (0–6), moderate (7–14), and high (15 and more) and coded as 0, 1, or 2, all other features are coded as absent (0) and present (1), which yields to a range of 0 to 13 in MSV scoring (Morgan et al., 2003). They also pointed out that variables in the content dimension “do not have to do with specific consequences or arguments, but are more concerned with how these more specific features are arranged … or portrayed” (Morgan et al., 2003, p. 523). Therefore, MSV is largely considered to be a format feature (Kang & Cappella, 2008).

In addition, Morgan et al. (2003) examined the association between the presence or absence of MSV features and PMSV. They concluded PMSV is roughly characterized by the use of intense images, sound saturation, music, scenes acted out, unexpected format, and a surprise ending (Morgan et al., 2003). As more research on MSV emerged, scholars have realized that the crucial distinction between subjective reactions to the message and the formal and content features contributing to these reactions (Kang, Cappella, & Fishbein, 2006; Stephenson, 2002; Stephenson & Palmgreen, 2001). MSV refers to a set of formal and content features of televised messages that are designed by message producers. It is an attribute of the message that leads to PMSV, which refers to the arousing, sensory, and affective responses to these message features. MSV can be more easily controlled in the development phase, whereas PMSV is determined only through after-viewing tests (Morgan et al., 2003). Moreover, as a measure of one’s cognitive assessment of MSV, PMSV is measured at the same time as attention; it is not clear whether PMSV is a causal predictor of attention or a post hoc validation (Kang et al., 2006).

The Impact of MSV on Message Processing and Persuasion

Some individual features of MSV, such as edits and cuts (Lang, Zhou, Schwartz, Bolls, & Potter, 2000; Niederdeppe, 2005), movement (Simons, Detenber, Cuthbert, Schwartz, & Reiss, 2003), visual graphics (Thorson & Lang, 1992), the use of suspenseful features (Niederdeppe, 2005), pace (Bolls, Muehling, & Yoon, 2003), special audio effects (Potter & Callison, 2000), and emotionally arousing messages (Lang, Newhagen, & Reeves, 1996) have been demonstrated to affect arousal and cognitive outcomes among teenagers and young adults in a variety of areas such as reducing drug use and tobacco use. Besides self-reported measures, empirical studies have also supported the impact of certain MSV features on arousal and attention using psychophysiological measures such as heart rate and skin conductance (e.g., Xu, 2015a).

Although messages of high relative to low in sensation value have been found to elicit greater message processing and more favorable evaluations across a range of outcome variables in health communication (Kang & Cappella, 2008; Xu, 2015a, 2015b), in many cases, the influence of MSV on message effectiveness seems to be very similar for high sensation seekers and low sensation seekers. High MSV messages also seem to appeal to low sensation respondents (Xu, 2015a, 2015b). In other words, the two groups process messages very similarly, with more desirable outcomes when MSV increases. The lack of meaningful differential or target effects on the attitudes of high and low sensation seekers may be partially due to the rather subtle distinctions on MSV between the two versions of PSAs applied in some studies (Xu, 2015a).

Limited Capacity Model of Motivated Mediated Message Processing

Another theoretical perspective of importance to the line of scholarly inquiry on MSV is the Limited Capacity Model of Motivated Mediated Message Processing (LC4MP) (Lang, 2006). LC4MP proposes that humans are information processors with a limited capacity of cognitive resources (Lang, 2006). The information processing of mediated messages is composed of two basic mechanisms (the orienting response and mental resource allocation) and three basic subprocesses (encoding, storage, and retrieval). Allocated resources can be used to encode (into perceptual or working memory), store (into long-term memory), and retrieve (activate from long-term memory) information (Lang, 2006). Attention to mediate messages such as televised PSAs is a function of viewers’ goals and of automatic processes evoked by content and message features. According to LC4MP, novel and significant structural and content aspects of messages will elicit orienting responses in message recipients, which subsequently result in the automatic allocation of resources for processing. Thus, this model predicts that stimuli that elicit either an orienting response or emotional or physiological arousal will receive more mental resources and be more thoroughly processed than those that do not (Lang, 2009).

Formal features have been shown to affect attention to the message (e.g., Lang et al., 2000), induce changes in physiological arousal such as heart rate and skin conductance (e.g., Lang, 2009), sway emotional arousal (e.g., Lang, 2009), alter message memory (e.g., Lang, 2000), affect the cognitive capacity required to process the message (e.g., Lang, 2000), and influence the favorability of viewers’ evaluations or the persuasiveness of the message (e.g., Lang, 2009). Thus, message features related to MSV such as cuts, edits, pacing, sound, use of narrative, and incorporation of emotionally intense materials are linked to greater orienting, attention, and liking of the message, all of which are arguably associated with message effectiveness and persuasiveness (Morgan et al., 2003).

Syntactic Indeterminacy

Messaris’s (1997) construct of syntactic indeterminacy (SI) provides another plausible explanation for why messages featured with HSV might generate enhanced processing and more favorable changes in attitudinal and behavioral outcomes (Niederdeppe, 2005). SI proposes that, to process juxtaposed images, individuals retrieve relevant information from memory to encode the meaning of images. Thus, the memory retrieval prompted by SI increases resources allocated to message encoding. While SI was originally presented to investigate the impact of montage-style features (Messaris, 1997), the logic can be reasonably extended to visual imagery in general, including televised messages (Morgan et al., 2003).

Visual images usually do not have an explicit syntax, which is a set of pre-established rules for interpreting pictures (Messaris, 1997). Consequently, viewers decode the meanings of montage style edits (changes from one visual image to another seemingly unrelated image) based on the context of the message and references to information stored in memory. “Because a visual argument cannot be entirely explicit, making sense of it may require of the viewer a greater degree of mental participation than would otherwise be the case. In a way, therefore, the viewer’s interpretation of a visual argument is more a product of her or his own mind than it would be if the argument were completely explicit to begin with” (Messaris, 1997, p. xviii). Because viewers have already made efforts to connect two disparate images, Messaris (1997) argues that montage-style edits facilitate message processing even when they do not explicitly contribute to the message’s central argument.

Although definitive support for SI as the explanatory mechanism for the relationship between message features and processing cannot be provided by survey data alone, Niederdeppe (2005) incorporated SI into the context of anti-tobacco media campaign design and examined the relationship between message processing and persuasion and stylistic features. He found that the number of unrelated cuts and the use of suspenseful features increased message processing among teenagers (Niederdeppe, 2005). SI provides some theoretical underpinnings for explaining a relationship between unrelated cuts, edits, and message processing. Mental participation demands viewers to actively process the message by retrieving relevant information from memory to construct meaning. The research on SI and MSV is very limited; more research endeavor would certainly be a welcome addition to the literature.

The Tension Between AMIE and ELM

One important issue when thinking about 30-second PSAs high in sensation value is that they are likely to grab the viewers’ attention, but whether they are able to comprehend and evaluate the message may be another vital consideration. Some researchers (e.g., Kang et al., 2006; Lang, 2000; Stephenson & Southwell, 2006) have raised concern on whether the attention demanded by format features, particularly in HSV messages, “steals” cognitive resources that otherwise might be devoted to the main content of the message, which consequently, would make MSV a distraction rather than an attention grabbing feature conducive to message processing and persuasion.

A notable theoretical critique of AMIE is that, by focusing only on MSV and sensation seeking, the model largely does not explain the majority of the variance in attention, and less is explained in terms of the subsequent processing to a message. Two out of the three dimensions of MSV focus exclusively on format features, with limited attention to message content. Scholars have attempted to situate AMIE in a broader context of persuasion, for example, some suggest that AMIE has certain connection with theories of behavioral decision making (Stephenson & Southwell, 2006) and other message processing theories such as Elaboration Likelihood Model (ELM) (Harrington, Lane, Donohew, & Zimmerman, 2006).

Theoretically, Harrington et al. (2006) made the initial attempt to integrate some core AMIE ideas with ELM, the major dual-process model of persuasion (Petty & Cacioppo, 1986). Whereas AMIE focuses more on explaining why individuals stay attuned to certain messages (Donohew et al., 1998), ELM (Petty & Cacioppo, 1986) deals more with how messages are processed after exposure. The integrative model attempts to consider both affect and cognition, highlighting the conditions under which affect or cognition may occur simultaneously, and the ways to manipulate specific message factors, aiming to enhance persuasive effects.

The integrative model proposed a new construct: message cognition value (MCV), referring to a set of content characteristics capable of enhancing the cognitive processing of persuasive messages (Harrington et al., 2006; Lane, Harrington, Donohew, & Zimmerman, 2006). Previous studies have investigated several message features that are relevant to MCV, such as the complexity of syntax (Lowery, 1998), the frames (Salovey, Schneider, & Apanovitch, 2002), order of arguments in a message (Petty, Tormala, Hawkins, & Wegener, 2001), and the argument strength (Kang et al., 2006). An objective list to comprehensively code a message’s cognition value has yet to be developed (Lane et al., 2006); however, several descriptors can help us to identify these attributes (Harrington et al., 2006). Messages of high cognition value tend to have multiple and complex (rational/logical) arguments; they present problems to be solved, and usually have open endings (Kim & Kardes, 1992). Messages with low cognition value, in contrast, tend to present straightforward information with simple or limited arguments. They have clear-cut presentation of information, and usually finish with closed endings (Kim & Kardes, 1992). Most of the message strategies have been supported by formative research using focus-group interviews (Harrington et al., 2003). According to the extended model of AMIE and ELM, an optimal level of cognitive engagement would engender greater message processing.

On the other hand, there is tension between AMIE and ELM, such that the two theories offer competing predictions in terms of the role of message structure features (e.g., MSV) in persuasion (Kang et al., 2006; Xu, 2015a). AMIE stresses a straightforward attention-getting effect of MSV. It posits that the increasing allocation of people’s cognitive resources, generated by HSV features, leads viewers to pay more attention to the arguments embedded in the message, which is conducive to the higher-order persuasion effect (e.g., message memorability, perception, and persuasiveness). ELM, however, questions this attention-enhancing argument. It posits that the presence of distraction, as a situational factor, can hinder one’s ability and motivation to process an incoming message (Kang et al., 2006). Although external distractions can induce relevant results, distraction may also come from the message itself. Some message format features (e.g., special visual effects) can attract people’s attention and generate a similar distraction effect (Kang et al., 2006). Thus, theoretically, an HSV message has the potential to act as a distractor, which may interfere with the reception of the central message and dampen message memorability and persuasiveness (Kang et al., 2006), and consequently shrink or even reverse the persuasive gap between strong and weak arguments (Kang et al., 2006). Taken as a whole, the key idea is that AMIE addresses an attracting effect of MSV, whereas ELM predicts a distracting effect of this formal feature interfering with message’s content. The competing predictions of the effect of formal features of AMIE and ELM propose the probability of an interaction between MSV and MCV.

Based on a secondary data analysis of evaluations by adolescents of 60 anti-marijuana PSAs, Kang et al. (2006) found that, for at-risk adolescents, the HSV high argument quality ads were among the least effective ones for perceived ad effectiveness and thought valence. The results were in line with the ELM that MSV could distract attention to reduce ad persuasiveness when the argument quality was high and could enhance ad persuasiveness when the argument quality was low (Kang et al., 2006). Furthermore, this interaction was evident only for adolescents with greater risk for marijuana use, suggesting that high MSV messages were especially distracting for the high-risk adolescents. Kang et al. (2006) acknowledged that the underlying causal mechanisms were explored rather than confirmed, because the study was based on a secondary data analysis.

More recently, Xu (2015a) conducted a study based on the integrative model of Harrington et al. (2006), involving MSV and MCV. The results indicated that MSV was not a distraction, but a facilitator of message persuasiveness. HSV messages were effective for increasing perceived ad effectiveness and ad liking especially when paired with high cognition value. These findings contributed to the AMIE. A number of studies (e.g., Anderson, Revelle, & Lynch, 1989; Kang, 2007) have supported the curvilinear argument between attention and arousal. Since MSV is positively linked with arousal, it is possible that the relationship between MSV and attention resembles that between arousal and attention. Although it is unclear what the optimal point for MSV is, some findings might fall into the right side of the curvilinear model, suggesting a distraction effect of MSV (e.g., Kang et al., 2006), while others fall into the left half of the figure stressing attention-enhancing effects (e.g., Xu, 2015a). Obviously, more research using primary data is warranted to examine the role of MSV.

MSV, Appraisal Theory, and the Excitation Transfer Theory

Although MSV does not have a uniform impact on message effectiveness, it may moderate the effect of other message features. Situating MSV with appraisal theory, Kang and Cappella (2008) revealed that MSV may work as an elicitor of arousal to intensify the impact of discrete emotions on perceived message effectiveness. Specifically, intensified feelings of anger and sadness increased the perceived message persuasiveness for messages with HSV. A similar pattern was found with other emotions, with no statistical significance. The impact of discrete emotions on perceived message persuasiveness was stronger for messages with HSV compared to those of low sensation value. This moderating effect of MSV on the relationship between discrete emotions and message effectiveness may be explained by restricted attention, excitation transfer, its close relationship with arousal, and the impact of arousal on cognitive processing (Kang & Cappella, 2008).

Appraisal theory explains how cognitive interpretations of message content lead to discrete emotions and can be used to locate message content features that elicit specific discrete emotions (Kang & Cappella, 2008). As a reliable inducer of arousal and affective responses, MSV should exert some equivalent impact as arousal. High MSV messages should be more likely to bring one’s attention on affective reactions toward the message than low MSV messages. This intensified focus on emotional responses should render certain emotions more accessible in memory and hence more salient in message evaluation. “Because the relational themes and situational context of the PSA as the basis for the appraisal process and the audiovisual features of MSV as the basis for arousal elicitation are indispensable message features of PSAs, the appraisal theory and MSV effects can help pinpoint message features that contribute to message effectiveness” (Kang & Cappella, 2008, p. 53).

On a relevant note, Zillmann’s (1971) excitation-transfer theory can also be used to explain the intertwined impact of MSV and discrete emotions on message processing. The theory purports that residual excitation from one message amplifies the excitation to another following message, if these two sources are presented close enough to each other in time (Zillmann, 1971). Arousal elicited by high MSV of the message (i.e., first source of arousal) may be combined with arousal generated by the emotional appeals of the message (i.e., second source of arousal). Therefore, the MSV-elicited arousal will be added to the content-elicited emotional arousal, which then increases one’s reliance upon emotion during message evaluation. In this case, both the strength of one’s emotional experience and the impact of these emotions on persuasion will be reinforced (Kang & Cappella, 2008).

Psychological Reactance Theory: Could HSV Messages Backfire?

Despite the support existing data have afforded to AMIE in evaluating health campaigns (Quick & Stephenson, 2008), it is worth pointing out that health campaign designers are faced with practical challenges when using messages with HSV. For example, television stations are reluctant to air safer sex PSAs if they are too sexually explicit or arousing, which limits the upper limit for MSV (Noar et al., 2010). In addition, there are theoretical challenges with implementing HSV messages, given the fact that they could possibly backfire among the target audiences. Recently, Miller and Quick (2010) called for a more cautious approach implementing HSV messages, particularly among young adults, who are likely high-reactance-prone, with an increasing resistance to any restrictions placed upon their free choice options. Health campaign designers can use HSV messages to attract the target audiences, especially those with high sensation-seeking propensity. However, if the messages disregard their progressively more reactant temperaments, even well-intentioned health promotion messages are likely to engender message rejection, boomerang effects, and source derogation. Miller and Quick (2010) particularly advised designers to bolster HSV messages with content cautiously crafted to avoid alienating those who are most at risk. “A worst-case scenario would be to produce powerful health message campaigns successfully engaging the attention of … individuals, only to provoke them into frenzied efforts to take up the contra-advocated behaviors in anxious attempts at restoring newly threatened freedoms” (Miller & Quick, 2010, p. 20).

Providing a related theoretical perspective, PRT has been cited frequently to explain why certain health messages sometimes have null or even counterproductive effects (e.g., Miller & Quick, 2010; Rains, 2013). PRT states that individuals experience an aversive state known as psychological reactance following the exposure to a freedom threat (Brehm & Brehm, 1981). The literature has identified a few reactance-contributing message features. For example, the association between high-controlling language and greater levels of reactance has been well documented, as revealed in reference to general health campaigns (Miller et al., 2013), safe sex (Quick & Stephenson, 2008), exercising (Quick & Considine, 2008), and environmental campaigns (Kronrod, Grinstein, & Wathieu, 2012). High-controlling messages explicitly pressure audiences to conform to a message with powerful and directive language. In contrast, low-controlling messages are more polite with less forceful and persuasive tactics (Miller, Lane, Deatrick, Young, & Potts, 2007).

Based on AMIE and PRT, Xu (2015b) examined the individual and combined effects of MSV and controlling language on young adults’ perception of anti-DUI and anti-smoking messages. A consistent interaction was revealed across two experimental studies, such that participants responded positively to the HSV messages, but only when they used low controlling language. Basically, the study suggests that increasing MSV coupled with high controlling language can backfire, especially when targeting young adults. Xu (2015b) advocates for the continued refinement of message strategies designed not only to maximize sensation value for high sensation seekers, but also to minimize the need for source derogation and message rejection by young adults who are high-reactance-prone. High sensation may garner better perceived effectiveness when the messages tone down the controlling flair without threatening viewers’ perceived freedom. It would be wise for health practitioners combatting risky behaviors to adopt the AMIE-based approach in conjunction with a strategy fully considering the critical effects of psychological reactance on young people (Xu, 2015b).

Future Research

There are several promising lines of inquiry that might yield fruitful results for future research with regard to MSV. First, although numerous studies have shown that individual MSV features affect message processing and persuasion, direct evidence about the impact of MSV as a composite feature on message effectiveness has just begun to accumulate (Kang & Cappella, 2008). Different individual format features under the construct of MSV may not produce the same effects, with the same strength or in the same direction. Future studies should investigate the relationships between specific MSV-enhancing features and message processing. Clearly, we need more research to specify the optimal mix of audio, video, and format features that maximize the various persuasive outcomes.

Second, a great deal of research has been conducted to investigate the impact of MSV on outcome measures such as attention, arousal, and behavior intention, but relatively little is known about the relations between MSV and viewers’ perception of media messages, especially those that are outside of the health communication context (Wei & Zhou, 2010). Therefore, future research can expand the literature by studying the impact of MSV in a variety of media message contexts (e.g., broadcast journalism). Future experiments might also incorporate pychophysiological measures (e.g., skin response and heart-rate deceleration) to complement self-reported measures in order to provide a broader understanding of the relationship between MSV and different outcomes.

Third, it is still unsettled, whether MSV is a distraction-contributing or an attention-enhancing feature. A deeper concern lies in the fact that, despite its intuitive appeal and conceptual soundness, there is still unclear mapping regarding the mechanism through which MSV affects attention, which often documents a curvilinear relationship with arousal. The curvilinear relationship between arousal and attention suggests that attention increases with arousal to a tipping point, beyond which attention will be decreasing (Kang, 2007). Since MSV is positively linked with arousal, it is possible that the relationship between MSV and attention resembles that between arousal and attention. It is unclear what the optimal point is that elicits the highest level of attention during message exposure, beyond which more sensational stimulation is too overwhelming or more than necessary. This explains the discrepancy in the literature: while some research has suggested a distraction effect of MSV (Kang et al., 2006), which falls into the right side of the curvilinear model; other studies (Xu, 2015a) fall into the left half of the figure, stressing attention-enhancing effects. On a related note, future studies should continue to explore other features that might affect attention and message processing jointly with MSV. For example, HSV ads that are also high in visual-verbal redundancy may not be distracting because, for example, when attention is drawn to the visual for an ad with high visual-verbal redundancy and high sensation elements, it is essentially drawing attention to the central message content (Cappella, Leader, White, & Kang, 2007).

Finally, existent MSV research has thus far focused primarily on one single individual difference factor: sensation seeking. SENTAR has received limited support. What is the optimal level of sensation for different audience members? Future studies should assess other individual difference variables, such as need for cognition, trait reactance, locus of control, etc.