Volume 23 Number 3, 2013
Hedonic Relevance and Outcome Relevant Involvement
Claude H. Miller
Josh M. Averbeck
Message production, reception, and processing studies are considered by many to be at the heart of communication research (Herbst, 2008; Rogers & Chafee, 1993). Researchers focusing in these areas have utilized a variety of theories to examine message variables used in studies employing a seemingly unlimited range of message topics. As a consequence, a considerable amount of journal space has been dedicated to questions concerning how such message-centered research should be conducted, given the multitude of topics addressed, and the potentially limitless range of attitudes about them (Hunter, Hamilton, & Allen, 1989; Jackson & Jacobs, 1983; Jackson, O’Keefe, Jacobs, & Brashers, 1989; Morley, 1988a, 1988b). One generally accepted conclusion has been that message topic matters, and how it bears on the relevant research should be assessed and considered (Burgoon, Hall & Pfau, 1991).
In addressing this issue, many researchers have relied on measuring message receivers’ attitudes and/or involvement levels relative to their message topics. Concerning the latter, Cho and Boster (2005) sought to develop and validate scales designed to assess Johnson and Eagly’s (1989) notion of value-relevant, outcome-relevant, and impression-relevant involvement. Although all three of these distinct types of involvement impact various aspects of social science research, it would seem that communication theory in general, and social and interpersonal influence research in particular, have, for the most part, been focused primarily on outcome-relevant involvement (Crano, 1995; Crano & Prislin, 1995; Johnson & Eagly, 1989).
Examining communication processes, researchers using inoculation theory (Pfau et al., 1997, 2003), psychological reactance theory (Brehm, 1966, 1972), cognitive dissonance theory (Festinger, 1957), dual process models (Johnson & Eagly, 1989; Petty & Cacioppo, 1979a; 1979b), and a number of other approaches (e.g. Brown & Basil, 1995; Roskos-Ewoldsen, Ralstin & St. Pierre, 2002; Slater, 1990) have emphasized the central role of involvement and/or the related construct of importance. However, notwithstanding its focus on involvement, virtually all of this research has paid little if any attention to the more basic and compelling concept of hedonic relevance.
One problem concerning the conceptualization of involvement is that if a topic is considered “important” to receivers, then pertinent messages are assumed to have certain predictable effects on those receivers. Unfortunately, mixed results have cast considerable doubt on this hypothesis (Cho & Boster, 2005; Roser, 1990; Salmon, 1986; Slater, 1997). For example, blue jeans and abortion were reported in one study as both being highly involving, and exhibiting similar profiles regarding attitude extremity (Cho & Boster, 2005); and while both of these topics could have personal consequences for a perceiver, the consequences relevant to the latter topic are likely to be considered far more critical than the former. As it concerns message effects research, the relationship between perceptions of personal consequence and their relative individual significance for a perceiver is central to the concept of hedonic relevance. It is this perception of personal consequentiality that distinguishes perceptions of objective importance or involvement from those of subjective importance and hedonic relevance, the latter of which, at this point, might simply be defined as the sum total potential for pleasure vs. pain inherent within the personal consequences contingent upon engagement with a particular attitude object.
To illuminate this topic and further explore the nature of outcome relevant involvement, we offer a review of the relatively scarce literature on hedonic relevance, and propose some basic definitional considerations drawing distinctions between hedonic relevance and other related constructs, such as arousal, involvement, importance, subjectivity, consequentiality, satisfaction, affinity, and interest. The thesis of this paper is that together, hedonic relevance (HR) and subjective importance (SI) make up the broader concept of outcome relevant involvement (ORI), which may function as perhaps the single most reliable predictor of human behavior. In the pages that follow, we present three studies used in the development of a scale for the measurement of ORI, with special emphasis on the need for further research, particularly as it concerns the nature and measurement of HR. We finish with a discussion of some potential applications for both the ORI scale and the HR subscale.
While discussed in several literatures, there appears to be no clear consensus directed at any single conceptualization of hedonic relevance as it relates to ORI. Based on a review of marketing, psychology, philosophy, communication, and related literatures, we offer two primary components necessary for assessing ORI pertinent to any given topic or attitude object; namely SI, referring to a cognitive assessment of the perceived personal consequentiality of an object for a perceiver, and HR, referring to the perceived pleasantness vs. unpleasantness underlying the motivational forces associated with that object. Because SI has already received considerable attention relative to HR, we train the majority of our focus upon the latter construct.
Foremost, as a motivational force, HR has a pleasure/pain dimension. For any individual, the hedonic relevance of an object is more basic than the simplest affective response associated with that object, since any affective response must itself first be a result of the object’s hedonic relevance for that individual. The basis for this claim is not merely tautological, but physiological: What is hedonically relevant for humans is determined within three pre-emotive hypothalamic centers or systems of the brain —one associated with general arousal, a second with reward, and a third with punishment (Buck, 1984; Lacey, 1967a; Lacey & Lacey, 1970). Hedonically relevant stimuli are fundamentally associated with a wide range of psycho-physiological, stress-related processes influencing mental and physical health (Lacey, 1985). The experience of pleasure is known to increase endorphin levels leading to feelings of euphoria, modulation of appetite, release of growth and sex hormones, lower blood pressure, enhanced immune response, and extended life-span (Olivo, 2009; Stoppler, 2007). In contrast, agitation-related disorders associated with the frustration of pleasure or the experience of pain are directly linked to fear, stress, anger, and hostility, which in turn are associated with a range of negative health outcomes, including depression, anxiety, headaches, sexual dysfunction, obesity, diabetes, impaired immune response, high blood pressure, and coronary heart disease (Chida & Steptoe, 2009; Kemeny, 2003).
The relative weight of this tonic pleasure/pain principle is another important aspect of hedonic relevance. Merely indicating the valence of an attribution is insufficient; the magnitude of its perceived personal consequence is also critical. For instance, message-focused researchers may employ multiple topics to demonstrate the generalizability of their findings, however, differing topics may or may not have personal significance for any given receiver; and only those topics whose consequences are considered personally relevant should be expected to affect pertinent behaviors. A minor issue of negative personal consequence may be considered extremely important in its effect on behavior in one instance, whereas an extremely positive, yet personally irrelevant consequence, may be perceived as wholly insignificant in its effect on behavior in another instance. It is the combination of these two dimensions —hedonic tone and magnitude of personal consequence —that most precisely defines hedonic relevance.
Because the concept of hedonic relevance has been defined by and applied to various lines of research in disparate ways, conventional standards for its measurement have yet to be developed, and the few scales that have emerged have been rudimentary and crude. For example, as described by Hirschman and Holbrook (1982), hedonic consumption refers to those aspects of behavior relating to people’s multisensory (gustatory, auditory, olfactic, tactile, and visual) afferent, fantasy, arousal, and emotive reactions to consumer product experiences. These hedonic responses are most frequently assessed in terms of hedonic tone, sometimes referred to as affective tone. To the extent it has been measured, primarily within the sensory evaluation literature, hedonic tone, as a construct, has been assessed somewhat simplistically. Sensory evaluation studies, applied almost exclusively to food products, typically involve food samples being presented to participants who are asked to decide how much they like or dislike each item, with response options usually offered on a 9-point scale ranging from dislike extremely to like extremely. This type of simple, single-item scale, used with both experts and untrained consumers, is more of an affective measure, which has demonstrated itself to be practical, at least, if not sensitive, when assessing results obtained from untrained participants (Amerine, Pangborn, & Roessler, 1965).
Although we could find no multi-item scales for hedonic relevance per se, we were able to find reference to other similar, simple, single-item measures for pleasantness/unpleasantness, often presented as proxy measures for stress (e.g., Berntson & Cacioppo, 2004; Cacioppo & Sandman, 1978; Lacey, 1967b), or as a collapsed measure in brain function research (Turner et al., 2007), where correlations among various affective responses are collapsed into pleasant and unpleasant response clusters, with the pleasant cluster typically including responses indicating positive affect associated with joy and amusement, and the unpleasant cluster indicated by negative affect associated with fear, disgust, and anger (but not sadness; Turner et al., 2007). Other researchers have conceptualized pleasantness/unpleasantness as a unidimensional construct measured along a continuum conceptually and empirically distinct from positive/negative mood or affect (Green & Salovey 1999; Tellegan, Watson, & Clark, 1999). However, as mentioned, we could find no multi-item scales assessing hedonic relevance that have been developed or tested to date.
Related Concepts and Proxies for Hedonic Relevance
Numerous related concepts have been utilized to discuss similar issues in message-based research. Below we review several of these concepts and discuss some of their shortfalls. Generally, each of these issues provides a degree of utility; however, they either lack one of the key aspects of hedonic relevance, or they attempt to address the concept in an overly simplistic manner.
Arousal. Arousal has often been used as a proxy for the concept of hedonic relevance, and has been conceptualized as a function of two control mechanisms, both automatic; one characterized as mobilizing, and the other as immobilizing. The former, which Pribram and McGuinness (1975; 1992) term activation, is centered on the basal ganglia involving “go” mechanisms associated with perceptual expectancies and motor readiness. The latter, which they refer to as arousal, “centers on the amygdala, which regulates the monitoring or ‘arousal’ neurons and becomes organized into a ‘stop’ or reequilibrating mechanism” (1975, p. 131). Mobilizing activation precedes effort, and stimulates mating, fight or flight, and appetitive, consummatory, approach behavior. In contrast, immobilizing arousal promotes concealment, and stimulates vigilance and effort directed toward gathering survival-relevant information.
Evidence collected through hundreds of studies has supported Pribram and McGuinness’s (1992) notion of arousal, activation, and effort as three poles of a motivational, dimensional space: “Thus, arousal became paired with familiarization, and activation with readiness; and effort (is) better represented by a comfort-effortful innovation dimension” (1992, p. 68). The points described within this space afford the basis for assessing that which is hedonically relevant for an individual. However, merely accounting for arousal does not go far enough in explicating the nature of approach and avoidance behaviors.
Beyond arousal and affective tone, other related, overlapping, and/or surrogate constructs for hedonic relevance include attention, ego-involvement, outcome-relevant involvement, self-interest, vested interest, subjective self-relevance, attachment, importance, and significance. Most of these terms have been defined, operationalized, correlated and measured in similar or parallel ways, and several have been used as proxy measures for hedonic relevance (Crano, 1995; Crano & Prislin, 1995; Lehman & Crano, 2002).
Given that human interaction in general —and interpersonal influence in particular —most frequently involves themes with consequential outcomes within relatively temporary situational contexts (Houston & Rothschild, 1978), it may be useful for researchers to focus some attention on the hedonic relevance of the issues, topics, and subjects they examine. Johnson and Eagly (1989) make a similar point regarding social influence research, which they suggest has primarily been focused on topics assessed in terms of their potential to stimulate outcome-relevant involvement for the parties concerned.
Vested Interest. Social psychologists have long incorporated the notions of self-relevance (e.g., Jones & De Charms, 1957), involvement (e.g., Johnson & Eagly, 1989), subjective utility (e.g., Edwards, 1954), and personal consequence (e.g., Fishbein & Ajzen, 1975), into their theorizing about human behavior; however, for the most part, they have not focused explicitly on hedonic relevance (although, see Chaikin & Cooper, 1973). One clear exception is Crano and colleagues’ vested interest theory (Crano, 1995; Crano & Prislin, 1995; Sivacek & Crano, 1982), which holds that the hedonic consequences associated with an attitude object provide the central mechanism motivating attitude behavior consistency. Thus, the presence of outcome-relevant, hedonic consequences is what separates vested interest from past conceptualizations of involvement. For example, an attitude object may be considered important in an objective sense, but not necessarily of hedonic consequence for the individual (e.g., civics, or voting in a national election). Conversely, if an attitude object is hedonically relevant to the individual, it will be considered personally and subjectively important (e.g., sex, or finding a mate), and therefore more reliably predictive of behavior (Lehman & Crano, 2002).
Crano contends that the function of hedonic relevance is manifested within five dimensions of vested interest, including: (a) an individual’s stake (i.e., investment) in outcomes associated with an attitude object; (b) the salience of hedonic consequences of those outcomes, given an attitude-relevant action; (c) their certainty; (d) their immediacy; and perhaps most importantly, (e) the individual’s perceived self-efficacy associated with performing or enacting the relevant attitude-implicated behaviors (Crano, 1995). These five dimensions effectively influence the perceived stake an individual associates with the consequences following from engagement (or lack of engagement) with a given attitude object. Thus, they are the cognitive aspects of vestedness, which together indicate the hedonic relevance of the attitude object for an individual —or what might be thought of as an intuitive, gut reaction following upon contact with an attitude object. In other words, vestedness is a direct result of both the cognitive and motivational forces underlying the hedonically relevant consequences perceived by individuals.
Enjoyment. Concerning self-efficacy, there is a large body of research demonstrating that, beyond an individual’s actual ability to enact an attitude implicated behavior, his or her self-perceived ability to enact that behavior has a powerful impact on motivation (Bandura, 1977; 1982; 1994; Brehm, 1966; 1972). Hedonic relevance plays an important role in this process through the manifestation of effectance motivation (White, 1959), or what has more recently been termed, self-expansive, or growth-oriented motives (Greenberg, Pyszczynski, & Solomon, 1995), which serve to organize and integrate cognitive processes and produce pleasure through successful engagement and mastery of challenges found in the environment (Csikszentmihalyi, 1975; 1990, deCharms, 1968; Deci & Ryan, 1980; 1985, Tamborini et al., 2010). Thus, self-expansive motives associated with self-efficacy are hedonically relevant, finding their motivational force directly through the pleasure they are capable of producing.
Similarly, Vorderer and colleagues’ broad notion of entertainment theory (Vorderer 2003; Vorderer, Klimmt, & Ritterfeld, 2004), based on Zillmann’s research on empathy (1991), mood management (1988), excitation transfer (1996), and affective dynamics (2003), sees positive reactions toward media entertainment as a hedonically relevant process in which various forms of “reception phenomena” produce pleasure and tonically satisfying states of enjoyment. Interestingly, the hedonically relevant aspects of both positive and negative situational affect seem to work in concert to produce particularly high levels of enjoyment (Prabu, Horton, & German, 2010).
Causal Attribution. Attribution theorists define hedonic relevance as having significant positive or negative consequences for the individual (Chaikin & Cooper, 1973), and the construct has played a key role in attribution theory because that which is experienced as hedonically relevant tends to be perceived in highly personal terms, which in turn greatly affect the nature of attributional processes. According to Maselli and Altrocchi (1969), the two conditions most central to facilitating an attribution of intent, are (a) power, linked to hedonic relevance, and (b) intimacy, linked to personalism. Thus, the acts of powerful people tend to be perceived as more hedonically relevant, and the intimate relationships people engage in tend to generate the most extreme attributions related to the most acute consequences.
Jones and Davis (1965) contend that hedonic relevance is seen as having “personal implications” for the perceiver, such that, “perceiving a behavior to have hedonic relevance tends to infer a correspondence between that behavior and the actor’s internal dispositions” (p. 237), resulting in a greater attribution of causal responsibility to the actor for that behavior and its consequences. Thus, attributions of intent tend to be clear when extreme behavior with positive or negative effects is directed personally toward the subject (Maselli & Altrocchi, 1969). Similarly, many clinical and literary observations suggest that attributions of intent become particularly salient, and with markedly critical effect, in contexts of intimacy. Intimacy is related to both personalism and hedonic relevance. Thus, the acts of intimates are more likely to be conditioned by one's physical presence and to have powerful psychological effects.
Similarly, an individual is more likely to associate the powerful love-hate emotions involved in intimate relationships with attributions of intent focusing on the acts of intimates. Thus, such relationships tend to generate strong attributions of intent related to issues of trust or mistrust, and in the cases of both love and hate, such attributions tend to result in the most extreme consequences for the individuals involved (Maselli & Altrocchi, 1969).
Involvement. Measuring involvement is essential for an understanding of operant decision-making processes. The concept has been conceptualized and operationalized in various ways across an array of contexts, and has been identified as an important variable within a wide range of social phenomena, including social and interpersonal influence, message framing (Maheswaran & Meyers-Levy, 1990), message processing (Park, Lee, & Han, 2007), resistance to influence in general (Pfau et al., 1997), and in relation to accessibility in particular (Pfau et al., 2003; Roskos-Ewoldsen, Ralstin, & St. Pierre, 2002).
Two general approaches to the study of involvement have recently surfaced within the interpersonal influence and marketing domains in response to many of the overly specific measures used in the past. Recognizing the study of involvement across a wide variety of objects and processes has produced a multitude of results with differing outcomes related to each. Zaichkowsky (1985; 1994) developed the personal involvement inventory (PII) to capture a target’s overall involvement with a given concept. Assuming personal, physical, or situational concerns may affect one’s involvement (Bloch & Richens, 1983), the PII focuses primarily on what is deemed personally relevant, which is defined as the perceived significance of an object based on the intrinsic needs, values, and interests of the perceiver (Zaichkowsky, 1985). The widely used scale consists of 20 semantic differential items such as “important/not important,” “valuable/worthless,” and “mundane/fascinating.”
Two other generalist approaches conceive of and measure involvement simply in terms of general physiological arousal (Greenwald & Leavitt, 1984), or by assessing cognitive and affective aspects of physical arousal separately (Cameron, 1993; Park & Mittal, 1985). Although these approaches have become more sophisticated relative to earlier simple measures of arousal, they are primarily about the amount or magnitude of arousal, not about any specific associated goals or desired outcomes, conscious or unconscious, an individual might have (Slater, 1997).
In response to the generalist approach within the social influence domain, Johnson and Eagly (1989) identified three distinct types of involvement: value-relevant, outcome-relevant, and impression-relevant. Though the overall approach to measuring the amount of personal relevancy remains, these specific approaches differ in two key aspects. First, the ways in which objects may have personal relevancy are identified and conceptually distinguished; and second, each type of involvement is associated with a unique set of outcomes.
Value-relevant involvement, also referred to as ego-involvement (Sherif & Cantril, 1947; Sherif & Hovland, 1961; Sherif, Sherif, & Nebergall, 1965), is defined as the psychological state created by the activation of attitudes linked to important values (Johnson & Eagly, 1989). Because the notion of value-relevant involvement originated within social judgment research, it is conceptualized in terms of latitudes of acceptance, noncommitment, and rejection. Specifically, high ego-involvement is associated with wide latitudes of rejection and narrow latitudes of noncommitment, and low ego-involvement is associated with narrower latitudes of rejection and wider latitudes of noncommitment (Sherif, et al., 1965).
In contrast, outcome-relevant involvement is somewhat of an umbrella term covering so-called issue involvement (Petty & Cacioppo, 1979a; 1979b), and response involvement (Zimbardo, 1960). Research associated with these forms of involvement is generally modeled after Petty and Cacioppo’s (1981; 1984) program of study utilizing the elaboration likelihood model (ELM) of persuasion. Johnson and Eagly (1989) contend ELM research is characterized by the use of outcome-relevant involvement because the manipulations employed make salient the outcome of an issue on an individual’s own goals and desires. Thus, this type of manipulation is about whether the issue will have consequential outcomes for the individual, and if so, how much. Because of the focus on personal consequences, outcome-relevant involvement is associated with information seeking behaviors (McQuarrie & Munson, 1992), greater elaboration on relevant information (Parse, 1990), and more objective processing (Hubbell, Mitchell, & Gee, 2001).
Impression-relevant involvement, on the other hand, differs from the other two forms of involvement in that it is concerned with public perceptions of the self (Zimbardo, 1960). It differs as it is neither about one’s concern with an issue (value relevancy), nor with what direct consequences an issue might hold in terms of some resource (outcome relevancy). Leippe and Elkin (1987) suggest impression relevancy concerns whether one holds opinions deemed socially acceptable by some potential evaluator. As a result, those motivated by impression relevant involvement tend to be more readily manipulated by others’ expectations (Leippe & Elkin, 1987); thus, this form of involvement is generally associated with other-directedness (Dillard & Hunter, 1989).
Goal of Present Operationalization
Although various forms of involvement may be useful in theorizing, they tend to suffer from measurement issues. In their review, Cho and Boster (2005) found a number of distinct and valid measures of involvement. However, they noted that most scales tend to focus on quantity of arousal rather than direction of arousal relevant to specific goals. As an alternative, Slater (1997) has argued for the development of goal-directed models capable of indicating directionality for motivation as well as magnitude. For example, lacking an indication of directionality, if an individual were to exhibit high outcome-relevant involvement regarding some activity, the only useful piece of information provided would be that outcomes relevant to that activity matter in some way, but with no indication as to why they matter, or what consequence they may hold for the individual.
While there is a plethora of measures available to capture involvement, arousal, and the like, we contend these measures do not address the more basic notion of HR, nor the broader notion of ORI. We sought to develop a bidimensional model of ORI containing both a measure of SI as well as a measure of HR, assessing both the importance/consequential and pleasure/pain dimensions in a parsimonious fashion. Such an ORI measure should be able to demonstrate perceived differences in topics commonly utilized in communication research that have heretofore remained undifferentiated.
Specifically, the goal was to develop a measure of ORI capable of assessing and discriminating the two aspects of HR and SI that were formerly unexamined and/or indistinguishable using previous measures of importance, involvement, and preference. Thus, we predict:
Thus the HR and SI subscales should be capable of discriminating between the two factors within the broader ORI construct reliably, regardless of the magnitude, intensity, and/or valance of what is or isn’t perceived as hedonically relevant for any given topic or attitude object.
To test these hypotheses we developed a bidimensional ORI scale comprised of two subscales assessing HR and SI, and tested their reliability and discriminant validity across five different topics within three studies, each using a different medium for delivering the treatment materials (text, video, and internet based html markup).
Because there are numerous scales assessing importance, our primary focus was on gathering items pertinent to hedonic relevance. Thus, to devise the HR and SI subscales, we used a combination of three methods to generate items pertinent to assessing HR within the ORI construct. First, through a logical criterion keying approach (Craig, 2007), we developed a list of relevant HR terms based on what has generally been written on the subject. Our literature review sought to cover a cross-section of scholarly articles, chapters and books on the concept published over the past 250 years. This primarily included philosophical, psychological, physiological, and practical research related approaches (the latter being mostly product development and advertising) examining or dealing with the concept of hedonic relevance (sometimes referred to in the various literatures as hedonistic relevance, hedonic tone, affective tone, emotional tone, fundamental tone, affectivity, or motivational significance).
Next we used a rational-theoretical approach (Craig, 2007) to develop a selection of items from among the gathered terms conforming to theoretical conceptions of the construct. The theoretical frameworks used were: from philosophy, Jeremy Bentham’s (1781/1948) hedonic calculus, and Henry Sidgwick’s (1874/2002) empirical hedonism; from economics and marketing theory dealing with self-interest and hedonic consumption (Hirschman, 1982; Hirschman & Holbrook, 1982; Holbrook & Huber, 1979); from attribution theory, Jones and colleagues’ theory of correspondent inferences (Jones & Davis, 1965; Jones & De Charms, 1957); from brain function research, Paradiso and colleagues’ work on cerebellar activity and emotional experience (Paradiso et al., 1997; 1999; Turner et al., 2007); and from social psychological approaches, Chaikin & Cooper’s work on evaluation and involvement (1973), and Crano and colleagues work on vested interest (Crano, 1995; Sivacek & Crano, 1982).
Finally, we incorporated a factor analytic assessment to ascertain the most basic, bidimensional structure of ORI, differentiating the theoretic motivational construct of HR from the related cognitive construct of SI, comprised of perceptions about concepts such as involvement, importance, significance, and consequentiality, etc. The relevant terms identified by the above process were converted into adjective pairs used as opposing anchors for a series of items measured along a 7-point (0-6) semantic differential continuum. These items (listed below) were used as repeated measures within Studies 1 and 2, and as a single measure within Study 3, to assess the perceived characteristics of five different topics —two each per the repeated measures design studies (eating and study habits, in Study 1, sex and the environment, in Study 2) —and one (camera equipment) in Study 3, all five of which were intended to be approximately of equal objective importance. All scale items were entered into a series of confirmatory factor analyses (CFAs) organized by topic across the three studies.
Participants were instructed as follows: “Based on the message you just read [video you just watched; webpage you just viewed], please indicate how you feel the following items apply to the topic of [fill in the topic].” The 15 adjective pairs included to assess ORI were: appealing/unappealing; boring/not boring; consequential/inconsequential; exciting/unexciting; important/unimportant; interesting/uninteresting; painful/not painful; personal/impersonal; pleasant/unpleasant; pleasurable/unpleasurable; punishing/not punishing; relevant/irrelevant; rewarding/not rewarding; satisfying/unsatisfying; and significant/insignificant.
Ten additional items were included within all three studies to assess the perceived quality and persuasiveness of the messages used to test the items of theoretical interest. These message quality items, measured along a 7-point (0-6) semantic differential continuum, were: coherent/incoherent; compelling/not compelling; convincing/unconvincing; credible/not credible; effective/ineffective; influential/not influential; persuasive/unpersuasive; polished/ unpolished; professional/unprofessional; and refined/unrefined. The desire was to have the five different target messages vary as little as possible on these dimensions. Because the predictor and criterion variables for all three studies were identical, the combined results are reported below after brief overviews of the procedures for each study.
Participants, Study 1
In exchange for course credit in their communication classes at the University of Oklahoma, a total of 106 undergraduates, mean age 20.1 (SD = 2.82) participated in Study 1. Of these, roughly 57% were female, 79% White, 6% Black, 6% Asian American, 2% Hispanic, 4% Native American, and 3% other. About half of the sample were communication majors. Participants were recruited from introductory communication classes (mostly general education students) as well as advanced classes (mostly communication majors). The demographic profile of this sample is representative of the university as a whole.
In Study 1, participants evaluated four separate, 200-word, text messages presented on a computer monitor (using MediaLab software), each designed to take approximately 30 seconds to read. Of the four messages, two were on the topic of food and eating habits, and were intended to represent an issue of high hedonic relevance; and two others were on the topic of scholarship and study habits, and were intended to represent an issue of relatively low hedonic relevance (across topics, message conditions varied in the use of detailed, concrete language, or more general, abstract language, and in promotion versus prevention framed appeals). This was a repeated measure design with topic treated as a within group variable and language treated as a between group variable (i.e., each participant saw messages in both topics, but in only one language condition), however, for the present analysis, the data were collapsed across language and appeal conditions. Sample messages within the two topics are presented in Appendix 1.
Participants, Study 2
A total of 121 undergraduates, mean age 20.2 (SD = 2.87) participated in Study 2. Of these, roughly 62% were female, 74% White, 5% Black, 8% Asian American, 6% Hispanic, 5% Native American, and 2% other. The makeup of this sample is comparable to that of Study 1, and its demographic profile is similarly representative of the university as a whole.
Study 2 was similar in concept to Study 1; however, instead of four text messages, eight 30-second, video messages were presented via computer (again, using MediaLab software). The two intended high hedonic relevance videos were on the topic of sex and contraception (one focusing on condom use [low demand], and one on abstinence [high demand]); and the two intended low hedonic relevance videos were on the topic of the environment and conservation (one focusing on recycling [low demand], and one on volunteering for an anti-litter campaign [high demand]). Messages versions also varied in the use of high vs. low controlling language. As in Study 1, this was a repeated measure design with topic treated as a within group variable and behavior (high/low demand) and language (high/low controllingness) treated as a between group variable (i.e., each participant saw messages in both topics, but in only one demand or language condition). As in Study 1, the data were collapsed across language and demand conditions, leaving only message topic as the within group variable of interest. Also, as mentioned, across topics, all messages were designed with the intention of being roughly equivalent in terms of their quality and objective importance. Voiceover transcripts of two sample video messages —one from each topic —are presented in Appendix 2.
Participants, Study 3
In exchange for course credit in their communication and management information science courses, a total of 277 undergraduates, mean age 19.9 (SD = 1.63) participated in Study 3. Roughly 65% were female, 74% White, 8% Asian American, 7% Native American, 5 % Hispanic, 4 % Black, and 2 % other. Participants were recruited from communication and MIS courses.
Study 3 presented participants with mockup product review webpages of competing digital camera brands (Sony and Nikon), which included equipment specifications, pricing, image illustrations, and product reviews. These pages were viewed on a computer (via MediaLab), where the product review text was manipulated for lexical complexity, level of affect, and one vs. two sided arguments. The hedonic relevance items pertained specifically to the topics of digital camera use and ownership, and were assessed prior to participants reading the product review text. An example screenshot of the product review webpage is included in Appendix 3.
A mentioned, since the predictor and criterion variables were identical for all three studies, the combined results are reported below. The ten items assessing the quality and persuasiveness of the message were entered into exploratory factor analyses (EFAs) organized by topics, which clearly identified two factors across all three studies. These factors formed the basis for two scales demonstrating excellent internal consistency across all five topics. The first was a message quality index, which indicated how refined, polished, professional, and coherent the messages were (4-item α: eating = .88; study = .83; sex = .87; environment = .88; camera = .87); and the second was a message persuasiveness index, which indicated how influential, persuasive, effective, convincing, and compelling the messages were (5-item α: eating = .90; study = .90; sex = .92; environment = .96; camera = .90).
Four paired-samples t-tests (two for each of the repeated measures studies) were conducted to compare perceived message quality and persuasiveness between topics within each study; ideally, there should be no differences in these attributes. In Study 1, although there was a significant difference in the scores for perceived message quality between the eating habits and study habits messages, it was relatively small, and there was no significant difference in perceived persuasiveness between the two topics, suggesting the four text messages were relatively equivalent in persuasiveness, if not perceived quality (see planned comparisons 1a-2b in Table 1). In Study 2, there were no significant differences between the environment and sex messages for either perceived message quality or persuasiveness, suggesting the four video messages were relatively equivalent in both these message attributes.
The 15 hedonic relevance-related items were used to specify a series of confirmatory factor analysis (CFA) models based on theoretical considerations, and those models were refined and simplified based on fit, parsimony, and unity (i.e., the desire to specify a single model that could fit the data from all five topics). Maximum likelihood (ML) estimation was used.
A multi-group analysis was performed to confirm the structure of the model across all five topics: sex, environment, study habits, eating habits, and digital camera ownership, and all fit indices were found to be satisfactory, including relative chi square (χ²/df) = 2.4, comparative fit index (CFI) = .94, normed fit index (NFI) = .91, root mean square error of approximation (RMSEA) = .04, and probability of poor fit (PPF) = .91 (i.e., nonsignificant). Cut-off points used were that the NFI should be greater than 0.9, the CFI should be greater than 0.9, the RMSEA should be less than 0.08, and the χ²/df should be less than 3 (Bagozzi & Yi, 1988; Bentler & Bonett, 1980; Byrne, 2001; Hair et al., 1998). All fit indices exceed these cut-off points.
The final multi-group model offers a parsimonious conception we refer to as outcome-relevant involvement (ORI), which is made up of two latent factors we have labeled subjective importance (SI) and hedonic relevance (HR). The model is shown fitted with the data from each of the five message topics in Figures 1- 5. As can be seen, four items were retained in the final HR factor: pleasant/unpleasant; pleasurable/unpleasurable; punishing/not punishing; and satisfying/unsatisfying, and four items were retained in the final SI factor: important/unimportant; relevant/irrelevant; significant/insignificant; and consequential/inconsequential. Table 2 provides the final ORI scale divided into HR and SI subscales.
The squared multiple correlation (SMC) for an item represents its reliability, i.e., the percentage of the total variance of the item accounted for by the factor. It might also be thought of as an indication of measurement error in terms of the item’s adequacy in describing the latent factor. Thus, higher SMC values indicate less error. For the validity and integrity of the latent constructs, we set the minimum acceptable SMC at .10 because it represents the conventional minimum value for a small effect size in communication research (Cohen & Cohen, 1983). The mean SMC value across the five topics measured was .54 for the hedonic relevance items, and .50 for the importance items, indicating the eight items together adequately describe these two latent constructs underlying ORI.
The correlations between HR and SI for each topic are also included in each model. The degree to which the correlations differ is of importance. The relationship between HR and SI for eating (r = .12, p < .001), studying (r = .11, p < .001), sex (r = .74, p < .001), environment (r = .57, p < .001), and cameras (r = .79, p < .001) differ dramatically (see Table 3).
Based on the CFA results, two scales were computed to represent the two latent factors, each of which demonstrated good internal consistency across all five topics: HR (4-item α: eating = .84; study = .85; sex = .83; environment = .76; camera = .74), and SI (4-item α: eating = .74; study = .74; sex = .78; environment = .86; camera = .76). Four paired-samples t-tests (two for each repeated measure study) were conducted to compare the perceived message importance and perceived hedonic relevance between topics within Studies 1 and 2. Ideally, in assessing the discriminant validity of the two scales, there should be little or no differences in importance, however, given the nature of the message topics, we would expect to see significant differences in perceptions of hedonic relevance.
In Study 1, although there was a significant difference in the scores for SI between the eating habits and study habits messages, it was relatively small, and in the opposite direction as the difference in HR between these two topics, where, eating habits was perceived to be significantly more hedonically relevant than study habits (see planned comparisons 3a and 4a in Table 1). This suggests that although the topic of eating was perceived to be significantly and substantially more hedonically relevant than the topic of study (in support of Hypothesis 1), it was nonetheless perceived as less subjectively important (in support of Hypothesis 2).
Similar results were found for the message topics assessed in Study 2. Although the means were in the predicted direction and approached statistical significance, there was no significant difference in the scores for SI between the sex and environment messages (thus, Hypothesis 2 did not receive additional support). However, there was a significant difference in HR between these two topics, whereby sex was perceived to be significantly more hedonically relevant than the environment (see planned comparisons 3b and 4b in Table 1), suggesting that, although the topics of sex (i.e., safe sex and abstinence) and the environment (i.e., recycling and volunteering) may have been perceived as being roughly similar in subjective importance, sex was nonetheless perceived to be a significantly more hedonically relevant message topic than the environment (providing further support for Hypothesis 1).
Three studies were conducted to construct and validate a measure of ORI comprised to two underlying dimensions, HR and SI. Confirmatory factor analyses indicate the proposed measures have utility for multiple topics across three different message media. Indeed, HR and SI represent critical concepts for all forms of message reception research involving issues or topics likely to be influenced by the effects of outcome relevant involvement.
The differences observed between topics for HR and SI provide evidence for the notion that HR is a unique construct, conceptually distinct from SI within the broader ORI construct. Through multiple messages and media, there were significant differences in HR where there were no significant differences in SI, or where SI levels were in the opposite direction of HR levels. Thus, it would appear in many instances, merely inquiring how important an individual considers a topic to be may not be sufficient, since, for example, topics such as the environment and study habits may be considered as objectively important, or even more subjectively important than topics such as sex and eating habits, yet they would appear to imply lesser personal consequences as a result of their being assessed as less hedonically relevant.
Therefore, it may be wise to assess both the HR of a topic as well as its perceived SI in cases where an individual’s outcome relevant involvement may have a substantial influence in the conclusions being drawn, the decisions being made, and the actions being taken.
Including an assessment of ORI, and particularly HR, would be useful in a variety of research contexts. For instance, interpersonal studies examining relational satisfaction could point to unique issues with varying outcome relevant consequences depending on whether they are considered high in SI but low in HR or vice versa. Social influence research might especially benefit from a measurement of HR. In particular, persuasive messages emphasizing the objective, or even subjective, importance of an attitude object may produce substantially less behavioral change relative to comparable messages targeting the same attitude object, but emphasizing its hedonic relevancy for the target. Likewise, framing an issue to have great SI without also noting its potential for hedonically relevant personal consequences may result in impressive self-reported attitude-relevant intentions but inconsistent or wholly absent attitude-relevant behaviors.
Similarly, health communication is an area ripe for the assessment of HR with regard to a wide variety of health concerns that may be considered subjectively important (i.e., potentially involving severely threatening outcomes) but not necessarily framed as having immediate nor certain hedonically relevant personal consequences for an individual. A fuller consideration of and focus on HR would have the effect of boosting the personal stakes individuals perceive themselves to have in various topics and issues central to a wide range of health risk campaigns, greatly enhancing message design and targeting.
Finally, marketing and product testing research would gain from the interaction between SI and HR. A product such as a digital camera may not seem important to individuals or appear to have personal consequences for potential consumers, but, as the results of Study 3 indicate, the outcome-relevant involvement of various topics, products, and services may not be intuitive. Understanding concepts of brand preference and product choice requires a clearer conceptualization of involvement. In fact, one of the most common measures of involvement was developed to examine brand and product preference (Zaichkowsky, 1985). Rather than the overall assessment of whether a particular product is interesting, the evaluation of competing products or brands surely relies upon the relative importance of the particular service or need the product can fulfill, and the extent to which the product has a bearing on consequences personally and hedonically relevant for the consumer. If a brand is considered important, but not hedonically relevant, then consumers are likely to give a lesser amount of consideration to purchasing it. Knowledge of consumers’ outcome-relevant involvement in the product, and particularly their assessment of its hedonic relevance, will allow marketers the ability to focus advertisements and other marketing campaigns more accurately on deficiencies in consumer interest.
Application to Theory-Based Research
Parallel Process Models. Several theory-based programs of research have focused on the concepts of involvement and attitude strength across a variety of topics, and most if not all of these research areas would benefit by an assessment of the hedonic relevancy of the attitudes and topics concerned. Attitude research, in general, would gain from an evaluation of attitude objects that vary in HR. Researchers using both the elaboration likelihood and heuristic systematic models of persuasion contend issues with which individuals are highly involved result in more elaborate and systematic message scrutiny. However, issue involvement may be considered in terms of objective importance and/or perceived personal consequentiality. Although dual process research tends to focus on outcome-relevant involvement (Johnson & Eagly, 1989), the narrower concept of HR may capture an aspect of personal consequentiality that is more powerfully focused and predictive of behavior. Assessing HR as a conative complement to the cognitive and affective aspects of message processing and attitude formation and modification would allow theorists to evaluate a more elemental aspect of motivation, thereby optimizing predicted attitude-behavior consistency (Crano, 1995), and improving the effectiveness of message designs to bring about actual behavioral change.
Attribution Theory. Research examining causal attribution processes could also benefit from an assessment of ORI or HR. Because extreme attributions tend to result in outcomes having great personal consequences for all concerned, the HR of a potential outcome or conclusion could function as a reliable predictor of such extreme attributions. Likewise, those attributions held with great conviction are likely to be correlated with assessments of high HR, and be perceived as personally consequential, even if they may not be considered as objectively important. Indeed, the emotional reactions associated with attributions are likely a direct function of the hedonic relevancy of the consequences involved (Maselli & Altrocchi, 1969). Thus, the greater the consequence, the greater the likelihood of a strong emotional response, and the more hedonically relevant the experience, the stronger the attributions formed.
Likewise, the concept of HR has been used by management researchers to theorize about how certain attributions concerning pleasant and unpleasant motivational responses are made, and why they tend to create such an influential impact on organizations when they occur (Martinko & Gardner, 1987), especially as they apply to utilitarian vs. hedonic concerns expressed across diverse disciplines such as sociology, psychology, economics, and the study of consumer behavior (Christians, 2007; Voss, Spangenberg, & Grohmann, 2003). More specifically, the emotional aspects of trust may be strongly related to HR; for instance, if a behavior has previously been indicative of harm, then individuals are likely to make emotionally charged attributions about the personal consequences implicated, and the resulting impact on trust may be in direct proportion to the hedonic relevance of the actions and outcomes involved. Indeed, the more an object of trust is hedonically relevant or experiential in nature, rather than utilitarian or functional, the more important emotions relative to rational expectations will be in determining trusting attitudes (Huff & Lane, 1999).
Attachment Theory. Hedonic relevance may also be central to the process of establishing secure attachments. Human attachments are based on affective relationships resulting from the responsiveness and availability of caregivers (Bowlby, 1982, 1988). Such relationships clearly hold tremendous personal consequences for individuals, since a lack of a security tends to lead to uncertainty and anxiety (Grossman, Grossman, & Zimmermann, 1999). On the other hand, secure attachments tend to allow individuals to focus on more pleasurable experiences, and such hedonically relevant interactions with caregivers are likely to be closely associated with nurturance and the types of positive experiences individuals will seek out and expect from their partners in the future.
Psychological Reactance Theory. Brehm (1966), in his theory of psychological reactance, defined reactance as a motivational state occurring when an important perceived freedom is threatened with elimination. The motivation to restore such a freedom is posited to vary in magnitude in accordance with three factors: the severity of the threat, the depth of conviction with which the threatened freedom is held, and the importance of that freedom. Hedonic relevance has direct implications for these latter two factors, since hedonically relevant freedoms are likely to be held with great conviction, and be considered even more vital than merely objectively important freedoms.
This may have three critical implications for reactance processes: First, threats to hedonically relevant freedoms may be perceived as more severe relative to other perceived threats; second, hedonically relevant freedoms may seem more critically connected to serious personal consequences relative to other freedoms. Thus HR may be potentially associated with particularly high levels of psychological reactance, and its resultant corollaries, such as message rejection, boomerang effects, and source derogation may be particularly sensitive to hedonically relevant threats (Miller et al., 2013). Moreover, it may be that the mixed findings and equivocal results reported in the reactance literature stem in part from the possibility that reactance is more likely to be found when importance functions as a valid proxy for HR; such that those cases where merely objectively important freedoms are threatened may not reliably stimulate characteristic reactance effects.
A third implication has to do with the nature of the restoration of freedom processes that may be most useful in alleviating reactance. Restoring a threatened choice may be most effective when the restoration method has direct, identifiable implications for HR, or when the hedonic relevancy of a restoration method is made salient. This might require pointing out not only how a particular choice option or freedom might be reinstated, but also how its reinstatement should have additional personal, hedonically relevant consequences for the individual (e.g., “the choice is yours, enjoy it.”) (Miller et al., 2007).
Regulatory Focus Theory. When considering regulatory focus (Higgins, 1996; 1998; 2002), the hedonic relevancy of a topic is central to the need for and benefits of regulatory fit (i.e., the value of matching a message to one’s motivational orientation; Higgins, 1998), given that the operant differences between the effectiveness of prevention versus promotion orientations are based on personally consequential outcome-relevant motivations associated with minimizing losses or maximizing gains. The HR of a particular gain or loss consequence is vital because when an outcome is not hedonically relevant, regulatory focus (and thus regulatory fit) will not matter, or will be less critical. The hedonic relevancy of the consequences to be gained or lost is what drives the persuasiveness of messages relying on the motivational value of good regulatory fit, rather than the objective importance of those consequences.
Vested Interest Theory. As noted above, involvement-driven, dual process models could benefit from a more nuanced assessment of HR. In making his case for vested interest theory as a better predictor of attitude-behavior consistency, Crano (1995) contends if an attitude is hedonically relevant, it will be highly vested, and thus act as a powerful predictor of outcome-relevant behavior. Moreover, if not highly vested, behaviors and action tendencies are predicted to have little or no relation to their focal attitudes. Thus, vested interest concerns the HR of a particular attitude-object in its capacity to have meaningful personal consequences for an attitude holder, and thereby reliably predict attitude-consistent behaviors. Likewise, the HR of dissonant or consonant cognitions may have bearing on the ability of those cognitions to stimulate or alleviate cognitive dissonance (Festinger, 1957).
Crano and colleagues have identified five key dimensions affecting the level of vestedness associated with a given attitude and its associated behaviors based primarily on one’s stake in relevant outcomes (Crano & Prislin, 1995). The concept of stake in a particular issue can be thought of as the invested level of self-interest, or perception of potential gain or loss related to that issue. Research has verified that when individuals perceive direct hedonically relevant personal consequences, they are more likely to behave in anticipation of those consequences (Crano & Prislin, 1995). Additional research has shown that when individuals are asked about attitudes of high HR, they generate more explicitly detailed, issue-relevant thoughts, and tend to use those thoughts and details as a basis for their decision making (Lehman & Crano, 2002). Beyond stake, which is essentially a function of HR, Crano asserts vestedness is concerned with four other constructs, including the salience, perceived certainty, and perceived immediacy of potential consequences following from behaviors relevant to a given attitude, as well as one’s pertinent level of self-efficacy regarding behaviors relevant to that attitude. Hedonic relevance, then, could be measured simultaneously with these five components of vested interest to fully capture the likelihood that a particular attitude will lead to a desired behavioral outcome (e.g., attitudes about disaster preparedness predicting actual preparedness behaviors; Miller, Adame, & Moore, 2013).
Expectancy Violations. Two other related theories make predictions based on issues of HR: expectancy violations theory (EVT; J. Burgoon, 1978; J. Burgoon & Hale, 1988; J. Burgoon & Jones, 1976), and language expectancy theory (LET; M. Burgoon, 1995; M. Burgoon & Miller, 1985). Simply stated, these theories are based on the notion that people form normative expectations about other’s nonverbal and verbal behaviors, and violations of those expectations (termed, expectancy violations) can have positive or negative effects on interpersonal interaction and influence. Expectancies are based in part on social and cultural norms, but also on biological influences and unique individual behavioral patterns. Hedonic relevance has particularly important implications for these latter two categories, given that such expectancies —similar to attributions —occurring in intimate or power-relevant relationships, when violated, tend to generate the most extreme and acute consequences. Like other novel or vivid acts, expectancy violations are part of an innate orientating response, often accompanied by arousal, drawing a receiver’s attention away from the message and toward the violation itself, as well as the person committing it. Thus, violations produce a wide range of responses depending on the reward value (EVT) or credibility (LET) of the communicator —which is likely to be determined in large part by HR. Therefore, the consequences of expectancy violations should center in large part on the contexts within which they occur, the normative expectations they involve, the salient attributes of the communicators, and the hedonic relevancy of the potential outcomes they imply.
Summary and Limitations
The present studies sought to develop valid subscales for HR and SI within the ORI construct. The CFA results indicate a bidimensional, latent construct, we refer to as outcome-relevant involvement, composed of two factors; one representing the motivational aspects of HR and another representing cognitive perceptions of SI. The parsimonious subscales designed to measure these two factors were found to be valid and reliable across several topics of relatively similar importance; and the structure of the scales was internally consistent across three studies examining messages using a range of differing language attributes within three different channels —text, video, and webpage.
Although we have argued for the utility of this measure of ORI in general, and HR in particular, there are some limitations to the present studies in need of mention. First, we have only examined the HR and SI of five topics, and although they all are quite different from each other, examination of additional topics is required to ensure these scales are valid across a variety of contexts and topics. Second, we have yet to examine ORI or HR simultaneously in conjunction with other previously validated measures of involvement (e.g., Greenwald & Leavitt, 1984, 1985; Laurent & Kapferer, 1985; Petty & Cacioppo, 1979b; Zaichkowsky, 1985; 1987). Although we contend HR taps into a unique aspect of ORI, more evidence is needed to better substantiate this claim.
A third concern involves the notion that perceptions of HR may vary over the lifespan, and college students, as emerging adults, many of whom are recently removed from familial influences, may be more prone to focus on more hedonically relevant aspects of life as they develop their personal preferences and outlooks relatively unconstrained by family or career responsibilities. Thus, drawing from a college population may not provide an unbiased assessment of the variables involved in developing a measure of perceived hedonic relevance. Until topics and issues that have been assessed using previous measures of involvement can be comparatively evaluated for their hedonic relevance by various populations across the lifespan, the validity and generalizability of these measures will remain in question.
We deliberately chose topics relevant to an undergrad population; thus, their responses should be appropriate for testing these measures. Moreover, the nature of the scale items (i.e., their simple and basic focus on ORI) should make them appropriate for all populations; and the fact that responses might differ from one population to another does not mean the scales themselves would be ineffective or invalid with different populations, only that they would perhaps indicate differing scores —i.e., higher or lower levels of HR and/or SI depending on the particular objects assessed. Nonetheless, further research is needed within differing populations to clarify important distinctions between the two dimensions of the ORI construct. In distinguishing between SI and HR, this research provides discriminant validity only relative to the subscales measuring these two dimensions. Although we have not tested the discriminant validity of the HR subscale against other motivation-related constructs, such as arousal, anxiety, or simple affect, or against other more cognitive constructs, such as consequentiality or subjectivity, we acknowledge the need for demonstrating both the discriminant and convergent validities of both the HR subscale and the full ORI scale.
We would encourage future research across a range of theoretical approaches to include ORI as both a predictor and/or a criterion variable. The scales presented here are relatively short (each subscale comprised of 4 items) and could easily be added to social science studies assessing perceptions about attitudes, behaviors, topics, communicators, and outcomes. Such an inclusion could help researchers more fully understand various elements and processes critical to the study of message production, processing, and reception, as well as a host of issues relevant to social and interpersonal interaction, and attitude behavior consistency.
Within the communication discipline, message-centered research, in particular, might benefit greatly from the measurement of hedonic relevance. Jackson, O’Keefe, Jacobs, and Brashers (1989) contend research designs employing single messages, absent replication, tend to suffer from random effects introducing unnecessary error into the data. Although we essentially agree with their position, we believe some of the problems associated with single-message designs may be ameliorated in part by a more nuanced understanding of message processing derived from the assessment of ORI associated with various message topics, since, even across replications, a population may demonstrate a broad variety of disparate beliefs and motivations concerning a given topic. Thus, we propose the measurement of HR and ORI as especially useful in helping to assure consistency across population samples and treatments involving a wide range of approaches to message-centered research.
 Given there is measurement error in the sample covariance matrix, ML reweights the discrepancy function when model-fit improves. Thus, unlike other estimation methods, ML relies on a fitted covariance matrix that accounts for such error (Jöreskog, 1967; Muthén & Kaplan, 1985).
 Items dropped from the CFA models on the basis of fit, parsimony, and unity, included: painful/not painful; exciting/unexciting; rewarding/unrewarding; interesting/uninteresting; appealing/unappealing: boring/not boring; and personal/impersonal.
Berntson, G. G., & Cacioppo, J. T. (2004). Heart rate variability: Stress and psychiatric conditions. In Malik, M. & Camm, A. J. (Eds.), Dynamic Electrocardiography (pp. 57-64). New York: Blackwell/Futura.
Brown, W. J., & Basil, M. D. (1995). Media celebrities and public health: Responses to “Magic” Johnson’s HIV disclosure and its impact on AIDS risk and high risk behaviors. Health Communication, 7, 345-370.
Burgoon, M. (1995). Language expectancy theory: Elaboration, explication, and extension. In C. R. Berger & M. Burgoon (Eds.), Communication and social influence processes (pp. 29-52). East Lansing, MI: Michigan State University Press.
Burgoon, M., & Miller, G. R. (1985). An expectancy interpretation of language and persuasion. In H. Giles & R. N. St. Clair (Eds.), Recent advances in language communication and social psychology (pp. 199-229). London: Lawrence Erlbaum.
Cacioppo, J. T., & Sandman, C. A. (1978). Physiological differentiation of sensory and cognitive tasks as a function of warning, processing demands, and reported unpleasantness. Biological Psychology, 6, 181-192.
Chida, Y., & Steptoe, A. (2009). The association of anger and hostility with future coronary heart disease: A meta-analytic review of prospective evidence. Journal of the American College of Cardiology, 53, 936-946.
Deci, E. L., & Ryan, R. M. (1980). The empirical exploration of intrinsic motivational processes. In L. Berkowitz (ed.), Advances in experimental social psychology (Vol. 13, pp. 39-80). New York: Academic Press.
Dillard, J. P., & Hunter, J. E. (1989). On the use and interpretation of the emotional empathy scale, the self-consciousness scales, and the self-monitoring scale. Communication Research, 16, 104-129.
Greenwald, A. G. & Leavitt, C. (1985). Cognitive theory and audience involvement. In L. F. Alwitt & A. A. Mitchell (Eds.), Psychological process and advertising effects: Theory, research, and applications (pp. 221-240). Hillsdale, NJ: Erlbaum.
Grossmann, K. E., Grossmann, K. & Zimmermann, P. (1999). A wider view of attachment and exploration: Stability and change during the years of immaturity. In J. Cassidy & P.R. Shaver (Eds.), Handbook of attachment: Theory, research, and clinical applications (pp. 760-786). New York: Guilford.
Higgins, E. T. (1996). Emotional experiences: The pains and pleasures of distinct regulatory systems. In R. D. Kavanaugh, B. Zimmerberg, & S. Fein (Eds.), Emotion: Interdisciplinary perspectives (pp. 203- 241). Mahwah, NJ: Erlbaum.
Higgins, E. T. (1998). Promotion and prevention: Regulatory focus as a motivational principle. In M. P. Zanna (Ed.), Advances in experimental social psychology: Vol. 30 (pp. 1-46). New York: Academic Press.
Houston, M. J., & Rothschild, M. L. (1978). Conceptual and methodological perspectives on involvement. In S. C. Jain (Ed.), 1978 educators’ proceedings (pp. 184-187). Chicago, IL: American Marketing Association.
Huff, L. & Lane, K. (1999). Trust formation in collectivist and individualist societies. Proceedings of the Seventh Cross-Cultural Consumer and Business Studies Research Conference, Cancun, Mexico, December 12-15, 1999. Retrieved May 27, 2010, from the Institute of Marketing Website, Marriott School of Management, Brigham Young University: http://marketing.byu.edu/htmlpages/ccrs/ccs.htm
Jones, E. E., & Davis, K. E. (1965). From acts to dispositions: The attribution process in person perception. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 2, pp. 219-266). New York: Academic Press.
Lacey, J. I. (1967b). Somatic response patterning and stress: Some revisions of activation theory. In M. H. Appley & R. Trumbell (Eds.), Psychological Stress: Issues in research, (pp. 14-36). New York: Appleton-Century-Croft.
McQuarrie, E. F., & Munson, J. M. (1992). A revised product involvement inventory: Improved usability and validity. In J. F. Sherry & B. Sternthal (Eds.), Advances in consumer research (Vol. 19, pp. 108-115). Provo, UT: Association for Consumer Research.
Miller, C. H., Ivanov, B, Compton, J., Averbeck, J. Robertson, K., Sims, J., Parker, K., & Parker, J.L. (2013). Boosting the Potency of Resistance: Combining the Motivational Forces of Inoculation and Psychological Reactance. Human Communication Research, 39, 127-155
Miller, C. H., Lane, L. T., Deatrick, L. M., Young, A. M. & Potts, K. A. (2007). Psychological reactance and promotional health messages: The effects of controlling language, lexical concreteness, and the restoration of freedom. Human Communication Research, 33, 219-240.
Paradiso, S., Johnson, D. L., Andreasen, N. C., O’Leary, D. S., Watkins, G. L., Ponto, L. L., et al. (1999). Cerebral blood flow changes associated with attribution of emotional valence to pleasant, unpleasant, and neutral visual stimuli in a PET study of normal subjects. American Journal of Psychiatry, 156(10), 1618-1629.
Paradiso, S., Robinson, R. G., Andreasen, N. C., Downhill, J. E., Davidson, R. J., Kirchner, P. T., et al. (1997). Emotional activation of limbic circuitry in elderly normal subjects in a PET study. American Journal of Psychiatry, 154(3), 384-389.
Park, D., Lee, J., & Han, I. (2007). The effect of on-line consumer reviews on consumer purchasing intention: The moderating role of involvement. International Journal of Electronic Commerce, 11, 125-148.
Petty, R. E. & Cacioppo, J. T. (1979b). Issue involvement can increase or decrease persuasion by enhancing message-relevant cognitive responses. Journal of Personality and Social Psychology, 37, 1915-1926.
Petty, R. E., & Cacioppo, J. T. (1984). The effects of involvement on responses to argument quantity and quality: Central and peripheral routes to persuasion. Journal of Personality and Social Psychology, 46, 69-81.
Pfau, M., Roskos-Ewoldsen, D., Wood, M., Yin, S., Cho, J., Kerr-Hsin, L., & Shen, L. (2003). Attitude accessibility as an alternative explanation for how inoculation confers resistance. Communication Monographs, 70, 39-51.
Pfau, M., Tusing, K. J., Lee, W., Godbold, L. C., Koerner, A., Penaloza, L. J., Hong, Y., & Yang, V. S. (1997). Nuances in inoculation: The role of inoculation approach, ego-involvement, and message processing disposition in resistance. Communication Quarterly, 45, 461-481.
Pribram, K. H., & McGuinness, D. (1992). Attention and para-attentional processing: Event-related potentials as tests of a model. In D. Friedman & G. Bruder (Eds.), Psychophysiology and experimental psychopathology: A tribute to Samuel Sutton (pp. 65-92). New York: The New York Academy of Sciences.
Roskos-Ewoldsen, D. R., Ralstin, L. A., & St. Pierre, J. (2002). The quick and the strong: Implications of attitude accessibility for persuasion. In J. P. Dillard & M. Pfau (Eds.), Persuasion: Developments in theory and practice (pp. 1-45). Thousand Oaks, CA: Sage.
Sidgwick, H. (1874/2002). The methods of ethics. London: Macmillan. Retrieved June 16, 2013 from the Classical Utilitarianism Website: http://www.laits.utexas.edu/poltheory/sidgwick/me/index.html.
Stoppler, M. C. (2007). Endorphins: Natural pain and stress fighters. Doctor’s Views, MedicineNet.com W. C. Shiel (Ed.). Retrieved September, 2009 from the MedicineNet.com Website: http://www.medicinenet.com/script/main/art.asp?articlekey=55001#
Zaichkowsky, J. L. (1987). The emotional affect of product involvement. In M. Wallendorf & P. Anderson (Eds.), Advances in consumer research, 14 (pp. 32-35). Provo, UT: Association for Consumer Research.
Zillmann, D. (1991). Empathy: Affect from bearing witness to the emotions of others. In J. Bryant & D. Zillmann (Eds.), Responding to the screen: Reception and reaction processes (pp. 135-168). Hillsdale, NJ: Erlbaum.
Zillmann, D. (1996). Sequential dependencies in emotional experience and behavior. In R. D.Kavanaugh, B. Zimmerberg, & S. Fein (Eds.), Emotion: Interdisciplinary perspectives (pp. 243-272). Mahwah, NJ: Erlbaum.
Zillmann, D. (2003). Theory of affective dynamics. In J. Bryant, D. Roskos-Ewoldsen , & J. Cantor (Eds.), Communication and emotion: Essays in honor of Dolf Zillmann (pp. 533-567). Mahwah, NJ: Lawrence Erlbaum.
Concrete, Promotional, Good Food Message (213 words)
Concrete, Promotional, Studying Message (219 words)
[Note: These two messages are presented as examples. They were designed to use concrete language within a promotion-based appeal. Other versions based on the same two topics (eating and studying), used various combinations of concrete/abstract language and promotion/prevention appeal formats. For this analysis, however, the data were collapsed across language and appeal conditions.]
Voiceover: High Demand Sex Message (with Low Demand Restoration)
Voiceover: Low Demand Environment Message (with High Demand Restoration)
[Note: These two messages are presented as examples. They were designed to use low controlling language within appeals emphasizing high or low demand solutions, while offering restoration postscripts of low or high demand solutions. Other versions based on the same two topics (sex and the environment), used various combinations of high/low controlling language and order of demand emphasis. For the present analysis, however, the data were collapsed across language and demand conditions.]
Product review mockup with outcome relevant involvement topics concerning digital camera use, satisfaction, and ownership
Copyright 2013 Communication Institute for Online Scholarship, Inc.
This file may not be publicly distributed or reproduced without written permission of