Motivated Processing of Peripheral Advertising Information in Video Games

Abstract

An experiment, guided by the Limited Capacity Model of Motivated Mediated Message Processing (LC4MP), manipulated players’ motivational activation states during play of a custom-built first-person shooter video game and examined memory of billboard advertisements embedded in the game’s peripheral content. In line with the LC4MP, a memory sensitivity test demonstrated that encoding of peripheral billboard advertisements was greatest during moderate-medium aversive activation and worst with moderate-high aversive activation. However, storage, measured using cued recall, was greater with moderate-medium appetitive activation when compared with other motivational states. The findings extend understandings of motivated cognitive processing of secondary advertising information in video game contexts.

Keywords

LC4MP motivated cognition peripheral information processing in-game advertising

In-game advertising expenditures reached US$3.1 billion in 2010 with revenues predicted to top US$7.2 billion by 2016 (DFC Intelligence, 2011). Advertising placements appear across video game genres including first-person shooter (FPS) games (Lewis & Porter, 2010), often in the periphery as information secondary to the games’ central action (Grigorovici & Constantin, 2004; M. Lee & Faber, 2007). Thus, some in-game advertisements may fall outside of players’ attention (M. Lee & Faber, 2007).

Video games possess emotional richness by design. Previous research demonstrates that the emotional tone of a mediated context influences memory of central and peripheral content (A. Lang, 2006a, 2006b; Yegiyan & Lang, 2010). Memory of peripheral in-game advertising information likely depends in part on the emotional tone of the primary video game content during the advertisement’s presentation. However, the influence of emotional game content on peripheral in-game advertisement processing can currently only be inferred based on research in nongaming contexts (e.g., peripheral cognitive processing of still images; Yegiyan & Lang, 2010).

The present study is a first step in investigating general processing propositions from the Limited Capacity Model of Motivated Mediated Message Processing (LC4MP; A. Lang, 2006a, 2006b; Park, 2006) in the specific context of in-game advertising. The effects of emotional tone in central FPS video game content on encoding and storage of information in peripherally placed in-game billboard advertisements are examined in an experiment. The findings offer potentially applicable insight for game producers and advertisers by considering the strategic placement of in-game advertisements in order to maximize advertising recognition and recall.

The LC4MP explains the interaction of mediated messages or content with the human emotional and cognitive systems (A. Lang, 2006a, 2006b; Park, 2006). To summarize the perspective, mediated messages, including video games, convey emotional tone to audiences in the forms of valence and arousal which automatically activate the human motivational systems. Motivational activation, engaged by mediated messages, influences an individual’s affective state and emotional experience. Additionally and importantly, motivational activation has been shown to automatically, adaptively, and predictably influence cognitive resource allocation to processing mediated content (Bradley, Angelini, & Lee, 2007; Gibbons, Lukowski, & Walker, 2005; A. Lang, Borse, Wise, & David, 2002; A. Lang, Chung, Lee, & Zhao, 2005; Miesse & Sparks, 2009; Potter, 2009).

Emotion and the Motivational Systems

In the LC4MP, message valence (pleasant or unpleasant) and arousal (calm or arousing) are the primary dimensions of emotion and codetermine the emotional tone of a mediated message (A. Lang, 2006a). Message emotional tone conveys survival relevance and, thus, serves as signal stimuli that activate the human motivational systems (A. Lang, 2006a, 2006b). Activation in the motivational system is reflexive, unconscious, and outside of human control. Valence of emotional tone determines the system of activation. The appetitive system sustains life through facilitating approach to pleasant stimuli; on the other hand, the aversive system protects life through facilitating avoidance of unpleasant stimuli. Arousal emerges from stimulus intensity or excitement associated with a stimulus and determines the level of system activation. As the arousal level of pleasant or unpleasant stimuli increases, appetitive or aversive activation, respectively, increases (Berntson & Cacioppo, 2000; Cacioppo & Gardner, 1999; A. Lang, 2006a, 2006b).

Motivational activation elicits motivational states that are thought to underlie conscious human emotional experience. That is, increases in appetitive and aversive activation as indicated by psychophysiological indicators are strongly associated with self-reported increases in positive, negative, calm, and arousing feelings (P. J. Lang, Bradley, & Cuthbert, 1997).

The appetitive and aversive systems are independent of one another (Cacioppo & Gardner, 1999). Therefore, activations may be reciprocal with one system active and the other inactive, coactive with both systems active, inactive with neither system active, or uncoupled with both systems active but in an entirely unrelated manner (Berntson & Cacioppo, 2000; Cacioppo & Gardner, 1999; A. Lang, Shin, & Lee, 2005; A. Lang & Yegiyan, 2011).

The motivational systems automatically respond to the motivational relevance in mediated emotional tone (Bradley et al., 2007; A. Lang & Yegiyan, 2011). Humans reflexively respond to environmental opportunities and threats (e.g., food, sex, violence, and death) and to their structural features (e.g., camera changes and pacing). As mediated messages generally possess both pleasant and unpleasant content, most affective occasions during media use are coactive, as opposed to single activation in nature.

Activation in one system may be low-or-moderate-or-high while activation in the other system may be low-or-moderate-or-high. For example, a few possible patterns of coactivation may include high appetitive/moderate aversive, low appetitive/high aversive, or high appetitive/high aversive. The system of greater activation will dominate in contrasting motivational occasions. For example, anger is associated with high aversive and moderate appetitive activation, but more likely to be a negative emotion (S. Lee & Lang, 2009). When appetitive activation is greater, appetitive inclinations are predominate, and vice versa. However, activation levels in the less active system also contribute to the collective motivational state (e.g., S. Lee & Lang, 2009).

To determine which system is dominant during a comparable coactivation occasion, the unique activation functions of the systems are considered (see Figure 1) with attention to the characteristics of positivity offset and negativity bias (Cacioppo & Gardner, 1999). When coactivation is comparable with low appetitive and low aversive activation, the appetitive system activation is primary. Humans’ natural drive to explore the environment in pursuit of sustenance facilitates this appetitive advantage at low levels of coactivation and is referred to as positivity offset. In Figure 1, the greater activation of the appetitive system, when compared with the aversive system at low coactivation, illustrates the positivity-offset characteristic.

Figure 1.

Motivational activation functions and resource allocation suggested by the LC4MP.

The appetitive system has evolved to activate less vigorously in response to pleasant stimuli of increasing intensity, when compared with the aversive system. The less vigorous increase in activation of the appetitive system results in tempered approach, when compared with the aversive system’s more vigorous avoidance response to unpleasant stimuli of similarly increasing intensity. The aversive system’s more vigorous activation response in the face of threats of increasing intensity is called negativity bias (Cacioppo & Gardner, 1999). Negativity bias facilitates adaptive avoidance behavior and results in aversive primacy at comparable medium and high levels of coactivation. In Figure 1, the steeper slope of the aversive activation line illustrates the negativity-bias characteristic.

Motivated Cognition

The LC4MP theorizes that motivational systems and levels of activation regulate cognitive processing of motivationally relevant mediated stimuli (A. Lang, 2006a, 2006b; A. Lang, Park, Sanders-Jackson, Wilson, & Wang, 2007). The perspective assumes that humans have a finite and limited pool of cognitive resources for the continuous and simultaneous subprocesses of encoding, storage, and retrieval of information (Baddeley, 2006; A. Lang, 2000, 2006b). During encoding, information intake occurs via the sensory receptors. A fraction of the information is selected and translated into mental representations. Storage is the process of linking encoded information within an associative network of previously stored information. In retrieval, cues reactivate the associational network including the stored information resulting in targeted recall. Information processing requires allocation and availability of resources (A. Lang, 2000).

Resource allocation to encoding, storage, and retrieval is codetermined by controlled and automatic mechanisms (A. Lang, 2000, 2006b; W. Schneider & Chein, 2003). Controlled processes are goal-directed and voluntary. Automatic processes are involuntary and triggered by novel or signal stimuli, such as motivational activation. Information intake is an evolutionary cognitive priority associated with appetitive activation. Remembering the source of life-sustaining opportunities promotes survival. Thus, arousing pleasant stimuli increase the activation level of the appetitive system and subsequently result in increased automatic resource allocation by the cognitive system to encoding and storage (A. Lang, 2006b; A. Lang, Sparks, Bradley, Lee, & Wang, 2004).

On the other hand, the aversive motivational system functions to protect humans from threats to survival (A. Lang, 2006a). With low to moderately threatening stimuli, survival depends on encoding and storing information to promote future avoidance. Between low and moderate aversive activation, automatic resources allocated to encoding and storage rapidly increase. However, when facing life-threatening stimuli, the cognitive imperative shifts to retrieval of escape information. Thus, for high aversive activation, resources allocated to encoding and storage decrease; resources allocated to retrieval increase (A. Lang, 2006a, 2006b; A. Lang et al., 2004).

Evidence for the interrelationships between the human motivational and cognitive systems has been observed across various media including television, radio, photographs, and video games (e.g., Bartsch & Oliver, 2011; A. Lang et al., 2007; Mickley Steinmetz & Kensinger, 2009; Park, 2006). The evidence (see Figure 1) shows that resources allocated to encoding and storage are greater for pleasant than for unpleasant content within a low range of arousal levels. The LC4MP theorizes that the pleasant processing advantage for low arousal stimuli is positivity-offset driven. On the other hand, resources allocated to encoding and storage of content are greater for unpleasant than for pleasant within a moderate range of arousal because of negativity bias. Resources allocated to encoding and storage of highly arousing content are greater for pleasant. However, resources allocated to encoding and storage of unpleasant content dramatically decrease within the high range of arousal.

Motivational Relevance in Video Games

In video games, opportunities and threats to a player’s character’s survival are thought to be emotional events that elicit motivational activation and emotional experiences among the players (A. Lang, 2006a; A. Lang, Bradley, Schneider, Kim, & Mayell, 2012; Ravaja, Saari, Salminen, Laarni, & Kallinen, 2006). For example, a character’s success or failure in game events or missions may be pleasant or unpleasant and elicit both positive and negative emotional experiences of varying intensities for the player (Ravaja et al., 2006). Violently defeating opponents in FPS games has been shown to elicit positive valence indicative of appetitive activation (A. Lang et al., 2012; Wilson & Kerr, 1999), whereas facing hostile opponents elicits negative emotional experience indicative of aversive activation (Vorderer, Klimmt, & Ritterfeld, 2004). Furthermore, violence cues in video games are positively related with arousal (Ivory & Kalyanaraman, 2007; Jeong, Bohil, & Biocca, 2011).

As described, researchers have demonstrated that the motivational systems automatically respond to primary motivational relevance in mediated emotional tone (Bradley et al., 2007; A. Lang & Yegiyan, 2011). However, mediated messages also convey secondary motivational relevance (Detenber & Lang, 2010). Secondary motivational relevance is learned or acquired through experience but similarly signals motivational relevance in the forms of opportunity and threat (e.g., birthday cake, guns, and police sirens; Detenber & Lang, 2010). Viewing of a horror movie scene provides one example. Horror movies present innately unpleasant content and activate the human aversive system. However, an individual who enjoys the entertainment experience associated with viewing horror films may experience appetitive activation and positive emotional experience (A. Lang, 2006a; S. Lee & Lang, 2009).

During video game play, primary and secondary motivational relevance similarly should interactively influence a player’s collective motivational state. As stated, gaming contexts affecting a character’s survival may contain emotional events involving the character’s success or failure that may elicit primary and secondary motivational activation and corresponding emotional experiences. However, positive emotional experience may also occur when events are perceived as challenging (e.g., even the death of the player’s character or the elimination of hostile opponents; A. Lang et al., 2012; Ravaja et al., 2006).

In the present study, opportunities and threats were presented in gaming environments to manipulate a player’s collective primary and secondary motivational state. A purpose of FPS video games is to place a character in a perilous fight for survival. In the context of the FPS game, a player guides the character in dynamic interaction with motivationally relevant stimuli conveying varying levels of valence and arousal (A. Lang et al., 2012; E. F. Schneider, Lang, Shin, & Bradley, 2004). Therefore, the player’s emotional experiences were expected to reflect the collective motivational state emerging from the gaming experience context. Positive emotional experience may depend on appropriate, expected peril during FPS game play.

For example, when a player can too easily eliminate opponents, the level of peril may not meet expectations. When the peril level presents an appropriate challenge in the video game context, secondary motivational relevance will likely be processed as pleasant stimuli, activate the appetitive system, and result in positive emotional experience (e.g., A. Lang et al., 2012). On the other hand, negative emotional experience may result when the aversive system activates in response to unexpected or insurmountable peril—for example, when the player finds it too difficult or impossible to avoid threats from opponents (e.g., Vorderer et al., 2004).

Thus, motivational activation and emotional experience depend on a dynamic interaction of structure, content, player, character, primary, and secondary motivational relevance. As such, the overall emerging motivational state during gaming is expected to be generally moderately arousing and coactive in nature. In addition to influencing emotional experience during video game play, motivational activation levels have been observed to influence allocation of cognitive resources to processing information. Based on these premises, FPS games were constructed to examine the influence of motivational activation on the processing of in-game advertising messages.

Peripheral Processing

The majority of previous motivated cognition studies have focused on processing of primary content of media (e.g., A. Lang, 2006a, 2006b). However, advertisements are often presented as secondary or peripheral information in video games. Researchers have long recognized that resources are differently available for primary and secondary tasks (Grigorovici & Constantin, 2004; Kahneman, 1973; M. Lee & Faber, 2007; Lynch & Srull, 1982). Specifically, if greater resources are allocated to a primary task than required, few resources will remain available for secondary task processing (A. Lang & Basil, 1998). Accordingly, encoding and storage of central and peripheral content may differ in mediated contexts including video games depending on resource availability.

Central and peripheral content has been distinguished in terms of the contents’ relevance to media users’ goals (e.g., Brewer, 1986; Christianson & Loftus, 1991). The rule of thirds has been used to structurally distinguish central from peripheral content in visual media (Shook, Larson, & DeTarsio, 2008; Yegiyan & Lang, 2010). Alternatively, the interaction of medium, message, and placement has been considered (Grigorovici & Constantin, 2004). Paid advertisements typically occupy an exclusively defined space and/or time separate from the sponsored message. In such cases, the advertising content is primary during its presentation.

However, advertising content may be considered peripheral to the sponsored content. When integrated within sponsored media (e.g., Williams, Petrosky, Hernandez, & Page, 2011), advertising most frequently serves a supporting role in the peripheral context of the primary content (e.g., Grigorovici & Constantin, 2004; M. Lee & Faber, 2007). The primary content of video games consists of the tasks central to advancement within the context of the game. Advertisements are not typically central to advancement but provide peripheral context secondary to players’ aims (Dardis, Schmierbach, & Limperos, 2012; Grigorovici & Constantin, 2004; M. Lee & Faber, 2007). For example, billboard advertisements in FPS games are usually not central to overcoming opponents.

A recent study (Yegiyan & Lang, 2010) provides a starting point for investigating encoding of peripheral information. Consistent with previous studies (e.g., Grigorovici & Constantin, 2004; Kahneman, 1973; M. Lee & Faber, 2007; Lynch & Srull, 1982), Yegiyan and Lang (2010) reported an encoding advantage for central over peripheral information from still photographs. Specifically related to this study, encoding patterns for central content were consistent with the resource allocation predictions of Figure 1 (with encoding of peripheral information greater during aversive than appetitive activation within the moderate-low to moderate-medium range of arousal). Also, encoding of peripheral information during moderate-medium to moderate-high aversive activation abruptly decreased. Figure 2 summarizes the findings of Yegiyan and Lang (2010) and the LC4MP’s motivated cognitive predictions.

Figure 2.

The resource allocation pattern of peripheral information.

Hypotheses

In FPS games, players remain vigilantly poised for fight or flight in the face of targets of opportunities and threats to life. As such, the appetitive and aversive systems activate during FPS game play (A. Lang et al., 2012). Approaching enemies and ripe targets increase intensity of the FPS game environment and elicit the players’ arousal (A. Lang et al., 2012). FPS game contexts of varying emotional tone were expected to elicit appropriate motivational system activation patterns (e.g., Cacioppo & Gardner, 1999). The present study examined the influence of moderate-medium and moderate-high motivational activations on encoding and storage of peripheral billboard advertisements. The following sections offer specific measures and hypotheses to examine the resource allocation predictions offered in Figure 2.

Resources Allocated to Encoding: Recognition Sensitivity

Resources allocated to encoding were indicated using signal detection. Signal-detection tests have been widely used to measure encoding by examining recognition memory strength or sensitivity (e.g., Bradley et al., 2007; Fox, 2004; Fox, Park, & Lang, 2007; Shapiro & Fox, 2002; Sparks, Matthews, & Chung, 2011). The approach is rooted in radar studies of the 1950s and measures an individual’s ability to accurately distinguish actual memories from false memories through the use of both target signal and foil background noise (Fox, 2004; Fox et al., 2007; Macmillan & Creelman, 2005). In signal-detection tests, participants are presented with both a target, which was actually presented in experimental content, and a foil, which was not presented but is similar to the signal. Participants are asked whether they have seen the target and the foil, respectively. When a target is correctly identified, a hit occurs. When a foil is incorrectly identified as signal, a false alarm occurs. Sensitivity increases when participants more accurately distinguish targets from foils (Fox, 2004; Fox et al., 2007; Macmillan & Creelman, 2005). In the present study, the sensitivity measure called d′ (d prime) was used to evaluate encoding of peripheral billboard advertisements. Resources allocated to encoding should correspond with the predictions of Figure 2. Thus, we hypothesize the following:

Hypothesis 1: There will be an interaction between motivational system and activation level such that increasing activation level will increase sensitivity (d′) during appetitive and decrease sensitivity (d′) during aversive activation. At the same time, sensitivity (d′) will be better for aversive compared with appetitive activation at moderate-medium levels of activation and better for appetitive compared with aversive activation at moderate-high levels of activation.

Evaluation of the sensitivity hypothesis would be incomplete without consideration of participants’ criterion for judging signal from noise stimuli in the signal-detection task. An individual’s criterion bias depends on whether the participant’s recognition judgments in distinguishing signal from noise are nonbiased, conservative, or liberal. An unbiased participant is equally likely to say “yes” as say “no.” A conservatively biased participant is less likely to say “yes,” which decreases the number of false alarms. A liberally biased participant is more likely to say “yes,” which increases the number of false alarms (Fox, 2004; Fox et al., 2007; Macmillan & Creelman, 2005). Criterion bias results will be reported in relation to Hypothesis 1 and provide additional context for interpreting the sensitivity results.

Resources Allocated to Storage: Cued-Recall Memory

Storage has been frequently measured using cued-recall memory tests (e.g., Bolls, Lang, & Potter, 2001; Dardis et al., 2012; A. Lang et al., 2007; Sparks et al., 2011). In the cued-recall tests, participants are presented with a cue linked to a targeted bit of information. Storage is assessed by asking the participants to write down what they remember based on the cue. If the targeted information is successfully recalled, the cue is considered to have activated the associative memory network including the stored target (A. Lang, 2000, 2006b; Sparks et al., 2011). As connections increase within the associative memory network, storage increases with the likelihood of cued recall. Thus, cued recall was used to indicate storage of peripheral billboard advertisements. Resources allocated to storage should correspond with the predictions of Figure 2. Therefore, we hypothesize the following:

Hypothesis 2: There will be an interaction between motivational system and activation level such that increasing activation level will increase storage (cued recall) during appetitive and decrease storage (cued recall) during aversive activation. At the same time, storage (cued recall) will be better for aversive compared with appetitive activation at moderate-medium levels of activation and better for appetitive compared with aversive activation at moderate-high levels of activation.

Method

A mixed Motivational System (2: appetitive, aversive) × Activation Level (2: moderate-medium, moderate-high) × Presentation Order (6) experimental design was used to test the hypotheses. FPS games, one of the most popular video game genres (Lewis & Porter, 2010) and rich in motivationally relevant content (A. Lang et al., 2012), served as experimental stimuli. Motivational System and Activation Level were within-subjects factors and manipulated with 16 virtual rooms in a FPS video game built using PC game creation software, FPS Creator (Bamber & Vanner, 2012). Four rooms were assigned to each of the four following treatment categories in the fully crossed design: (a) moderate-medium appetitive, (b) moderate-high appetitive, (c) moderate-medium aversive, and (d) moderate-high aversive activation (see Figure 3). Presentation Order was a between-subjects factor to control possible order effects (see Table 1).

Figure 3.

Brand logo placement in four emotional conditions.

Table 1.

A Semi-Random Presentation Order of Four Emotional Conditions.

Six presentation orders
Order	Repetitions
Order	1	2	3	4
1	A1 B1 C1 D1	B2 D2 A2 C2	C3 A3 D3 B3	D4 C4 B4 A4
2	B2 C2 A2 D2	C3 D3 B3 A3	D4 A4 C4 B4	A1 B1 D1 C1
3	C3 D3 A3 B3	D4 B4 C4 A4	A1 C1 B1 D1	B2 A2 D2 C2
4	D4 A4 B4 C4	A1 C1 D1 B1	B2 D2 C2 A2	C3 B3 A3 D3
5	A1 D1 B1 C1	C3 B3 D3 A3	D2 C2 A2 B2	B4 A4 C4 D4
6	B2 C2 D2 A2	C1 A1 B1 D1	D4 B4 A4 C4	A3 D3 C3 B3

Note. A = moderate-medium appetitive activation; B = moderate-high appetitive activation; C = moderate-medium aversive activation; D = moderate-high aversive activation.

In addition, in each of the 16 virtual rooms, a brand logo was embedded in a billboard advertisement frame and positioned in the room’s periphery (see Figure 3). Four product categories were utilized: beer (e.g., Heineken), soft drink (e.g., Mountain Dew), fast food (e.g., Chipotle), and retail (e.g., Home Depot). Four different brands were used in each of the four categories. For the placement of brand logos in the billboards, each of the 16 brands was randomly assigned to each virtual room using a Microsoft Excel’s random formula.

Independent Variables

Motivational system

Motivational System of activation was manipulated by varying the number of opponents (2: appetitive; 4: aversive) and speed of opponents (half: appetitive; full: aversive). Resulting emotional experience scores of positive and negative valence were used to indicate appetitive and aversive activation, respectively. After a character exited each virtual room, the participants responded to the following items, using a 1- to 9-point scale, adopted from previous studies (S. Lee & Lang, 2009; Yegiyan & Lang, 2010): (a) “Please rate your emotional experiences in the virtual room that you just exited,” ranging from 1 (not at all positive, happy, or pleased) to 9 (extremely positive, happy, or pleased); and (b) “Please rate how bad you felt,” ranging from 1 (not at all negative, unhappy, or annoyed) to 9 (extremely negative, unhappy, or annoyed).

Activation level

Activation level was manipulated by varying nonbloodied (low activation) and bloodied-items (high activation). Resulting arousal emotional experience scores were used to indicate activation levels. After a character exited each virtual room, the participants responded to the following item on a 1- to 9-point scale, used in the previous studies (S. Lee & Lang, 2009; Yegiyan & Lang, 2010): “Please rate how aroused you felt,” ranging from 1 (not at all aroused, excited, or awake) to 9 (extremely aroused, excited, or awake).

Dependent Variables

Sensitivity (d’)

In order to measure d′, participants completed 32 forced-choice visual detection tasks (e.g., S. Lee & Lang, 2009; Yegiyan & Lang, 2010). Each 1,024 × 768 logo appeared for two-frames (i.e., 66.7 milliseconds) within an approximately 5-second movie clip. The targets included the 16 brand logos presented in billboards actually seen in the periphery of the virtual rooms. For each target logo, a matched foil for the same brand that differed in design from the target logo actually seen was presented. Recognition measures may yield ceiling effects or indiscriminately high scores when participants are too easily able to discern targets. Increasing the period of memory decay using visual masks following the stimulus presentation helps to control for ceiling effects (Rothschild, Qualheim, Deith, & Hyun, 1990). The masks may also focus participants’ attention on the embedded content (Rothschild et al., 1990). Thus, to prevent ceiling performance, a 2.5-second visual noise mask both proceeded and followed the logo.

The signal-detection targets and foils were randomly presented on a desktop computer using MediaLab software (Jarvis, 2006). Participants were instructed to watch each clip and respond to the item “Did you see this logo image while playing the game?” by clicking “yes” or “no.” A target correctly identified as “yes” was scored as a hit. When a foil was incorrectly identified by “yes,” it was scored as a false alarm. Sensitivity (d′) means were calculated using the following formula: sensitivity (d′) = Z (p (hits)) − Z (p (false alarms)). Criterion bias rates, denoted by c, were calculated using the formula: (−.5) (Z (p (hits)) + Z (p (false alarms))). When a respondent is equally likely to say “yes” as “no,” hit and miss rates are equal and c equals zero indicating an unbiased criterion for familiarity judgments. When a respondent is more likely to say “no” than “yes,” miss rates exceed false alarm rates and c is positive (greater than zero) indicating a conservative bias. When a respondent is more likely to say “yes” than “no,” false alarm rates exceed miss rates and c is negative (less than zero) indicating a liberal bias (Fox, 2004).

Cued-recall memory

Cued-recall memory was used to measure storage (e.g., M. Lee & Faber, 2007). Participants were asked to write down the brand names for the logos viewed in the billboard advertisements in the game based on provided cues. The four category names (i.e., beer, soft drink, fast food, and retail) were presented as cues—for example, “Please write down ALL the “BEER” brand names which you have seen while playing the game.” The participants were given unlimited time to respond or could respond “no” and skip the specific question if they could not remember brand names. When a brand name was correctly reported, the response was scored as one point. As there were four virtual rooms in each emotional condition, the total cued-recall scores ranged from 0 to 4 for each condition. For the calculation of cued-recall means, total scores were divided by four. The questions for cued recall were randomly presented using MediaLab software (Jarvis, 2006).

Participants and Procedure

Participants (N = 71, 20 males, 51 females; M age = 20.85 years, SD = 2.15) were recruited from mass communication undergraduate courses at a large southwestern university in exchange for extra credit in a course. The sample size exceeded the estimated requirement of 66 participants for small-to-medium effect sizes (f = .20) for the repeated-measures F test for two within-subjects factors and one between-subjects factor (Cohen, 1992), according to a priori power analysis using G*Power software (Faul, Erdfelder, Buchner, & Lang, 2009).

Participants were randomly assigned to one of six presentation orders (Order 1: n = 12, Order 2: n = 13, Order 3: n = 13, Order 4: n = 12, Order 5: n = 12, Order 6: n = 9) and completed the entire protocol at a private computer workstation in a 19-seat research laboratory (see Table 1). The computer keyboard’s arrow keys (left hand) were used to maneuver the on-screen character and the computer’s mouse (right hand) for changing direction and shooting (left click). After exiting each of the 16 virtual rooms, participants rated their emotional experience during play in that room on a paper questionnaire. After completing the entire game, participants completed computer-based measures in the following order: (a) cued-recall test, (b) signal-detection test, and (c) demographic questionnaire.

Results

Manipulation Check

As a manipulation check, Motivational System (2) × Activation Level (2) × Presentation Order (6) repeated-measures ANOVAs were performed on positive valence, negative valence, and arousal ratings. First, Motivational System had a significant main effect on positive valence, F(1, 65) = 38.29, p < .001, η² = .37. On the positive rating scale, pleasant content (virtual rooms) eliciting appetitive activation (M = 4.94, SD = 1.33) was rated greater than unpleasant content eliciting aversive activation (M = 4.37, SD = 1.12). No significant effect of Presentation Order was found, F(5, 65) = .12, p = .988. Second, a significant main effect of Motivational System was also found for negative valence, F(1, 65) = 30.54, p < .001, η² = .32. On the negative rating scale, unpleasant content (M = 4.79, SD = 1.40) was rated greater than pleasant content (M = 4.17, SD = 1.46). Presentation Order did not have a significant effect, F(5, 65) = 1.77, p = .131. Third, Activation Level had a significant main effect on arousal ratings, F(1, 65) = 37.67, p < .001, η² = .37. The arousal ratings were higher for high arousing content (M = 5.91, SD = 1.63) than low arousing content (M = 4.89, SD = 1.62). Presentation Order had no significant effect, F(5, 65) = .51, p = .771. The Activation Level manipulation was determined within the moderate range of arousal ratings (see Figure 2).

Hypothesis 1: Recognition Sensitivity (d′)

Hypothesis 1 predicted an interaction between motivational system and activation level. A repeated-measures ANOVA was performed on the recognition sensitivity (d′) means (see Table 2) and revealed a significant Motivational System × Activation Level interaction, F(1, 65) = 16.83, p < .001, η² = .21.

Table 2.

Repeated-Measures ANOVA Analysis for Recognition Sensitivity (d′).

Source	Type III SS	df	MS	F	p	η²	Obs. Power
Within-subjects factors
Motivational System	.40	1	.40	1.10	.299	.02	.18
Motivational System × Order	1.55	5	.31	.86	.514	.06	.29
Error	23.45	65	.36
Activation Level	2.76	1	2.76	6.20	.015*	.09	.69
Activation Level × Order	1.69	5	.34	.76	.582	.06	.26
Error	28.91	65	.45
Motivational System × Level	5.86	1	5.86	16.83	.000***	.21	.98
System × Level × Order	2.71	5	.54	1.56	.184	.11	.51
Error	22.63	65	.35
Between-subjects factors
Presentation Order	3.47	5	.69	1.51	.200	.10	.50
Error	29.87	65	.46

Note. SS = sum of squares; MS = mean square; Obs. Power = observed power.

p < .05. **p < .01. ***p < .001.

Hypothesis 1 predicted that increasing the activation level would increase sensitivity during appetitive and decrease sensitivity during aversive activation. Sensitivity (d′) means were greater for moderate-medium aversive (M = .30, SD = .60) than moderate-high aversive activation (M = −.22, SD = .72), t(70) = 4.78, p < .001, two-tailed, supporting Hypothesis 1. Although sensitivity means increased between moderate-medium appetitive (M = .07, SD = .57) and moderate-high appetitive activation (M = .16, SD = .66), the difference was not significant, t(70) = −.86, p = .392, two-tailed.

At the same time, sensitivity was predicted to be better for aversive compared with appetitive activation at moderate-medium levels of activation. Sensitivity (d′) means were greater for moderate-medium aversive than moderate-medium appetitive activation, t(70) = −2.48, p < .05, two-tailed, supporting Hypothesis 1.

Furthermore, sensitivity was predicted to be better for appetitive compared with aversive activation at moderate-high levels of activation. Sensitivity (d′) means were greater for moderate-high appetitive than moderate-high aversive activation, t(70) = 3.47, p < .01, two-tailed, also supporting Hypothesis 1. Table 3 shows t-tests results and Figure 4 illustrates the mean comparisons.

Table 3.

Paired-Samples t Tests for Recognition Sensitivity (d′).

Pair	Paired differences					t	df	p (two-tailed)
	M	SD	SEM	95% confidence interval of the difference
	M	SD	SEM	Lower	Upper
M-medium appetitive and M-medium aversive	−.23	.78	.09	−.42	−.05	−2.48	70	.015*
M-high appetitive and M-high aversive	.37	.91	.11	.16	.59	3.47	70	.001**
M-medium aversive and M-high aversive	.51	.91	.11	.30	.73	4.78	70	.000***
M-medium appetitive and M-high appetitive	−.09	.88	.10	−.30	.12	−.86	70	.392

Note. SEM = Standard Error of the Mean.

p < .05. **p < .01. ***p < .001.

Figure 4.

Mean comparisons for recognition sensitivity (d′).

For criterion (c) means, there was also a significant Motivational System × Activation Level interaction, F(1, 65) = 12.20, p < .001, η² = .16. The observed criterion means did not differ significantly among the conservatively biased moderate-high appetitive (M = .15, SD = .55), moderate-medium aversive (M = .17, SD = .59), and moderate-high aversive (M = .23, SD = .55). However, a significantly greater liberal bias existed for moderate-medium appetitive (M = −.32, SD = .57), when compared with the moderate-high appetitive (t(70) = −5.22, p < .01, two-tailed), moderate-medium aversive (t(70) = 5.71, p < .01, two-tailed), and moderate-high aversive (t(70) = −6.22, p < .01, two-tailed).

Hypothesis 2: Cued Recall

Hypothesis 2 predicted an interaction between motivational system and activation level. A repeated-measures ANOVA was performed on the cued-recall means to test Hypothesis 2 (see Table 4). As predicted, the Motivational System × Activation Level interaction effect for cued-recall means was significant, F(1, 65) = 16.46, p < .001, η² = .20.

Table 4.

Repeated-Measures ANOVA Analysis for Cued-Recall Memory.

Source	Type III SS	df	MS	F	p	η²	Obs. Power
Within-subjects factors
Motivational System	1.39	1	1.39	34.05	.000***	.34	1.00
Motivational System × Order	.03	5	.01	.12	.987	.01	.08
Error	2.66	65	.04
Activation Level	1.87	1	1.87	55.18	.000***	.46	1.00
Activation Level × Order	.15	5	.03	.91	.483	.07	.30
Error	2.20	65	.03
Motivational System × Level	.62	1	.62	16.46	.000***	.20	.98
System × Level × Order	.29	5	.06	1.56	.185	.11	.51
Error	2.46	65	.04
Between-subjects factors
Presentation Order	.21	5	.04	.54	.742	.04	.19
Error	5.12	65	.08

Note. SS = sum of squares; MS = mean square; Obs. Power = observed power.

p < .05. **p < .01. ***p < .001.

Hypothesis 2 predicted that increasing the activation level would increase cued recall during appetitive and decrease cued recall during aversive activation. Cued-recall means were greater for moderate-medium aversive (M = .21, SD = .22) than moderate-high aversive activation (M = .14, SD = .17), t(70) = 2.30, p < .05, two-tailed, supporting Hypothesis 2. A significant difference was found but unexpectedly cued-recall means were greater for moderate-medium appetitive (M = .44, SD = .27) than moderate-high appetitive activation (M = .19, SD = .18), t(70) = 7.58, p < .001, two-tailed.

At the same time, cued recall was predicted to be better for aversive compared with appetitive activation at moderate-medium levels of activation. Although a significant difference was found, cued-recall means were significantly greater for moderate-medium appetitive than moderate-medium aversive activation, t(70) = 6.02, p < .001, two-tailed.

Furthermore, cued recall was predicted to be better for appetitive compared with aversive activation at moderate-high levels of activation. Cued-recall means were greater for moderate-high appetitive than moderate-high aversive activation, t(70) = 1.87, p < .05, one-tailed (p = .066, two-tailed). Table 5 shows t-tests results and Figure 5 displays the mean comparisons.

Table 5.

Paired-Samples t Tests for Cued-Recall Memory.

Pair	Paired differences					t	df	p (two-tailed)
	M	SD	SEM	95% confidence interval of the difference
	M	SD	SEM	Lower	Upper
M-medium appetitive and M-medium aversive	.23	.33	.04	.16	.31	6.02	70	.000***
M-high appetitive and M-high aversive	.05	.22	.03	−.00	.10	1.87	70	.033*, one-tailed
M-medium aversive and M-high aversive	.07	.26	.03	.01	.13	2.30	70	.024*
M-medium appetitive and M-high appetitive	.25	.28	.03	.19	.32	7.58	70	.000***

Note. SEM = Standard Error of the Mean.

p < .05. **p < .01. ***p < .001.

Figure 5.

Mean comparisons for cued-recall memory.

Discussion

The study was a first attempt to investigate motivated cognitive processing of peripheral information in video games using the LC4MP. The findings demonstrate that encoding and storage of peripheral billboard advertisements differ according to the player’s motivational activation state and level. The encoding results provide insight for in-game advertisers whose goal is to increase brand awareness, which is usually assessed by brand recognition measures (M. Lee & Faber, 2007).

The recognition sensitivity measure indicated encoding patterns consistent with the predictions based on the LC4MP. Increasing motivational activation level increased sensitivity during appetitive and decreased sensitivity during aversive activation. At the same time, sensitivity was better for aversive compared with appetitive activation at moderate-medium levels of activation and better for appetitive compared with aversive activation at moderate-high levels of activation. Criterion judgments were comparable and conservative for moderate-high appetitive, moderate-medium aversive, and moderate-high aversive. However, a significantly greater liberal bias existed for moderate-medium appetitive.

As currently written, the LC4MP tends to explicate the cognitive processing of the specific stimulus eliciting the specific motivational activation. In as much as the billboards used in the study were part of the overall emotional tone of the video game rooms, the LC4MP is logically applied to encoding of the billboards as demonstrated by the results. Furthermore, a function of the encoding subprocess is to scan the world and encode novel objects. The billboards were novel objects within the periphery of the virtual world and, thus, within the encoding subsystem’s purview.

The storage of peripheral information in the current context differed from encoding. The beer, retail, fast food, and soft drink billboard content were not directly relevant in the situational context of the game. Storage depends on associational connections and simultaneously active information in the brain’s network. When active encoded information is irrelevant, disassociated knowledge structures are simultaneously active. Lacking relevant active associated information for linkage, the encoded information quickly deactivates and storage suffers.

For example, “Home Depot” may be encoded, but storage is less likely because the FPS context is not likely to activate “home improvement” associations. However, the FPS context is likely to activate associations with “shooting.” A billboard advertisement for “Glock” handguns in the FPS context would likely be better stored due to its contextual relevance. Because “Glock” and “shooting” are both activated, “Glock” is more likely to be linked to the relevant “shooting” associational knowledge structure.

In addition, although the storage predictions of Hypothesis 2 were plausible based on the LC4MP, the hypothesis may not have been supported because the motivational system manipulation was intertwined with message complexity. Beyond influencing motivational activation, varying the number and speed of opponents also varied the content’s structural complexity. The content that elicited aversive system activation and a high activation level across systems was also more complex. Thus, resources allocated to encoding and storage likely increased with message complexity. Past LC4MP studies have demonstrated that differences in structural complexity influence automatic resources allocated and, thus, resources available for processing information in mediated messages (A. Lang et al., 2007). The difference between resources allocated and resources required determines resources available. Resources available may be positive, negative, or zero. When resources available are negative, cognitive overload is thought to exist and processing suffers (A. Lang et al., 2007). In the present study, the aversive and high activation level content likely elicited greater resources allocated because of their complexity—more opponents with faster speeds. Also, the aversive and high activation level content likely required greater resources to fend off the opponents. Thus, those stimuli likely overloaded the cognitive processor, leaving insufficient resources available to store the peripheral information in the billboards. As evidence, moderate-high aversive sensitivity means were lowest and criterion means were most conservative. Therefore, Hypothesis 2’s results may demonstrate the greater susceptibility of the storage subprocess to the influences of cognitive overload, when compared with encoding.

For example, the seemingly paradoxical pattern of results for the moderate-medium appetitive condition demonstrated the potential influence of cognitive overload on storage. The condition had poor sensitivity scores and a liberal criterion bias, which indicates guessing due to poor encoding. The sensitivity results likely stem from a lack of sufficient resource allocation, as the condition was least structurally complex and elicited the fewest automatic encoding resources. However, resource requirements for encoding the information were comparable with other categories and, thus, resources available for encoding became negative resulting in cognitive overload. At the point of cognitive overload, all resources were shifted from encoding to the more secondary task of storing the peripheral information, resulting in the greatest cued-recall scores of the four conditions (see A. Lang & Basil, 1998).

Another important issue when considering the results relates to the brands and products used in the billboard advertisements. Both risky (beer) and nonrisky (retail) products were used in the peripheral billboards in this study. A. Lang, Chung, et al. (2005) reported that risky and nonrisky products uniquely influence motivated cognition. For example, based on A. Lang, Chung, et al.’s (2005) results, “Heineken” beer would be expected to convey a pleasant arousing emotional tone to the college student participants in the study and activate the appetitive system, but “Sears” retail would not. In addition, Kissler and Keil (2008) demonstrated that the emotional tone of visual images presented in the periphery modulated saccadic eye movement. According to their research, voluntary and involuntary eye movements toward peripheral images depend on the valence, arousal, and the visual field of placement (left vs. right).

Although the study did not directly address the emotional tone of the products used in the billboards, the design accounted for the contribution of the peripheral information to the overall emotional tone of an in-game virtual world’s context. Holistic emotional experience ratings were reported immediately after the participant’s character exited each virtual room. Therefore, a billboard’s emotional tone was included in a participant’s emotional experience evaluation of the virtual context. Furthermore, the product categories were distributed across treatment levels. With respect to the left-versus-right visual field, placement was controlled in the current study, as it was constant across virtual rooms.

A potential initial step toward addressing these limitations may include a study that controls ongoing motivational activation level and examines the independent influence of the motivational relevance of the brand/product in the billboard on motivational activation, encoding, and storage. Also, future studies should examine the influence of emotional tone and peripheral billboard placement on eye movement and attention (Kissler & Keil, 2008). Next, a study should investigate the interaction of these important variables. Such a study should address whether left-versus-right visual field placement of peripheral advertising information of varying emotional tone presented during varying motivational states elicited via central information (e.g., video games) differentially impacts recognition and recall of the peripheral advertising information.

In addition, the influence of the action being undertaken in the game on motivational activation and memory for peripheral billboards should be considered. For example, billboards appearing when players are scanning the room for opponents may be better processed than those appearing during active engagement with an opponent. During engagement with opponents, the player may experience an aversively driven narrowing of attention scope to the central fighting task. A widening of attention scope related to hunting should occur during scanning phases.

There were other minor limitations. First, only widely familiar, popular brands were chosen for the advertisements and memory tests in this study. Brand familiarity (e.g., Kent & Allen, 1994) may have influenced storage patterns. However, as all advertisements were for comparably familiar national brands, the effect was assumed constant across brands. Second, this study examined a narrow range of motivational activation levels (i.e., moderate-medium to moderate-high) with emotional experience means near the midpoint of the 9-point scale. Although the differences in the manipulation check were significant, the manipulation may not have been successful. To advance understandings of motivated cognitive processing of secondary information, an increased range of motivational activation should be examined. Third, the global self-report ratings employed in this study may not have adequately captured moment-to-moment emotional responses reflecting the dynamic processing of these mediated environments. A suggestion for future work is to pair the self-reported ratings with physiological measures, such as facial Electromyography (EMG) activity, heart rate, and skin conductance, during play.

Practical Implications

The present study advances understandings related to processing of in-game peripheral advertising information for game producers and advertisers. For example, contextual placement should depend on an advertiser’s goal. Whether a brand is recognized versus recalled appears to depend on contextual placement. To enhance recall, advertisements should be placed in the periphery of contexts similar to the moderately appetitive conditions of the present study. Furthermore, brand recognition increased within moderately aversive and highly appetitive contexts, but was poor in the highly aversive category. Brand memory—recognition and recall—is necessary, but not sufficient to explain purchase behavior. Thus, future studies should explore the relationships among emotion, memory, attitudes, and purchase intent.

Footnotes

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Author Biographies

Sungwon Chung (PhD in mass communications, Texas Tech University, 2012) is a doctoral candidate and graduate instructor in the Instructional Technology Program in the College of Education at Texas Tech University.

Johnny V. Sparks (PhD, Indiana University, 2006) is an associate professor of advertising in the Department of Journalism at Central Michigan University.

References

Baddeley

A. D.

(2006). Working memory: An overview. In Pickering

S. J.

(Ed.), Working memory and education (pp. 1-31). New York, NY: Academic Press.

Bamber

Vanner

(2012). FPS creator (Version 1.18). Lancashire, UK: The Game Creators.

Bartsch

Oliver

M. B.

(2011). Making sense of entertainment: On the interplay of emotion and cognition in entertainment experience. Journal of Media Psychology: Theories, Methods, and Applications, 23, 12-17.

Berntson

G. G.

Cacioppo

J. T.

(2000). Psychobiology and social psychology: Past, present, and future. Personality and Social Psychology Review, 4, 3-15.

Bolls

P. D.

Lang

Potter

R. F.

(2001). The effects of message valence and listener arousal on attention, memory, and facial muscular responses to radio advertisements. Communication Research, 28, 627-651.

Bradley

S. D.

Angelini

J. R.

Lee

(2007). Psychophysiological and memory effects of negative political ads: Aversive, arousing, and well remembered. Journal of Advertising, 36, 115-127.

Brewer

W. F.

(1986). What is autobiographical memory? In Rubin

(Ed.), Autobiographical memory (pp. 25-49). New York, NY: CambridgeUniversity Press.

Cacioppo

J. T.

Gardner

W. L.

(1999). Emotion. Annual Review of Psychology, 50, 191-214.

Christianson

S.-Å.

Loftus

E. F.

(1991). Remembering emotional events: The fate of detailed information. Cognition and Emotion, 5, 81-108.

10.

Cohen

(1992). A power primer. Psychological Bulletin, 112, 155-159.

11.

Dardis

F. E.

Schmierbach

Limperos

A. M.

(2012). The impact of game customization and control mechanisms on recall of integral and peripheral brand placements in video games. Journal of Interactive Advertising, 12(2), 1-12.

12.

Detenber

B. H.

Lang

(2010). The influence of form and presentation attributes of media on emotion. In Doveling

von Scheve

Konijn

E. A.

(Eds.), The Routledge handbook of emotions and mass media (pp. 275-293). New York, NY: Routledge.

13.

DFC Intelligence. (2011). DFC intelligence forecasts global advertising in video games to reach $7.2 billion in 2016. Retrieved from http://www.dfcint.com/wp/?p=315

14.

Faul

Erdfelder

Buchner

Lang

A.-G.

(2009). Statistical power analyses using G*Power 3.1: Tests for correlation and regression analyses. Behavior Research Methods, 41, 1149-1160.

15.

Fox

J. R.

(2004). A signal detection analysis of audio/video redundancy effects in television news video. Communication Research, 31, 524-536.

16.

Fox

J. R.

Park

Lang

(2007). When available resources become negative resources: The effects of cognitive overload on memory sensitivity and criterion bias. Communication Research, 34, 277-296.

17.

Gibbons

J. A.

Lukowski

A. F.

Walker

W. R.

(2005). Exposure increases the believability of unbelievable news headlines via elaborate cognitive processing. Media Psychology, 7, 273-300.

18.

Grigorovici

D. M.

Constantin

C. D.

(2004). Experiencing interactive advertising beyond rich media: Impacts of ad type and presence on brand effectiveness in 3D gaming immersive virtual environments. Journal of Interactive Advertising, 5(1), 1-26.

19.

Ivory

J. D.

Kalyanaraman

(2007). The effects of technological advancement and violent content in video games on players’ feelings of presence, involvement, physiological arousal, and aggression. Journal of Communication, 57, 532-555.

20.

Jarvis

B. G.

(2006). MediaLab Research Software (Version 2006). New York, NY: Empirisoft.

21.

Jeong

E. J.

Bohil

C. J.

Biocca

F. A.

(2011). Brand logo placements in violent games: Effects of violence cues on memory and attitude through arousal and presence. Journal of Advertising, 40, 59-72.

22.

Kahneman

(1973). Attention and effort. Englewood Cliffs, NJ: Prentice Hall.

23.

Kent

R. J.

Allen

C. T.

(1994). Competitive interference effects in consumer memory for advertising: The role of brand familiarity. Journal of Marketing, 58, 97-105.

24.

Kissler

Keil

(2008). Look–don’t look! How emotional pictures affect pro- and anti-saccades. Experimental Brain Research, 188, 215-222.

25.

Lang

(2000). The limited capacity model of mediated message processing. Journal of Communication, 50, 46-70.

26.

Lang

(2006a). Motivated cognition (LC4MP): The influence of appetitive and aversive activation on processing of video games. In Messaris

Humphreys

(Eds.), Digital media: Transformations in human communication (pp. 237-256). New York, NY: Peter Lang.

27.

Lang

(2006b). Using the limited capacity model of motivated mediated message processing (LC4MP) to design effective cancer communication messages. Journal of Communication, 56(Suppl. S1), S57-S80.

28.

Lang

Basil

(1998). Attention, resource allocation, and communication research: What do secondary task reaction times measure anyway? In Rolloff

M. E.

(Ed.), Communication yearbook (Vol. 21, pp. 443-474). Thousand Oaks, CA: Sage.

29.

Lang

Borse

Wise

David

(2002). Captured by the World Wide Web: Orienting to structural and content features of computer-presented information. Communication Research, 29, 215-245.

30.

Lang

Bradley

S. D.

Schneider

E. F.

Kim

S. C.

Mayell

(2012). Killing is positive! Intra-game responses meet the necessary (but not sufficient) theoretical conditions for influencing aggressive behavior. Journal of Media Psychology: Theories, Methods, and Applications, 24, 154-165.

31.

Lang

Chung

Lee

Zhao

(2005). It’s the product: Do risky products compel attention and elicit arousal in media users? Health Communication, 17, 283-300.

32.

Lang

Park

Sanders-Jackson

A. N.

Wilson

B. D.

Wang

(2007). Cognition and emotion in TV message processing: How valence, arousing content, structural complexity, and information density affect the availability of cognitive resources. Media Psychology, 10, 317-338.

33.

Lang

Shin

Lee

(2005). Sensation seeking, motivation, and substance use: A dual system approach. Media Psychology, 7, 1-29.

34.

Lang

Sparks

J. V.

Bradley

S. D.

Lee

Wang

(2004). Processing arousing information: Psychophysiological predictors of motivated attention. Psychophysiology, 41(Suppl. S1), S61.

35.

Lang

Yegiyan

(2011). Individual differences in motivational activation influence responses to pictures of taboo products. Journal of Health Communication: International Perspectives, 16, 1072-1087.

36.

Lang

P. J.

Bradley

M. M.

Cuthbert

B. N.

(1997). Motivated attention: Affect, activation, and action. In Lang

P. J.

Simons

R. F.

Balaban

M. T.

(Eds.), Attention and orienting: Sensory and motivational processes (pp. 97-135). Hillsdale, NJ: Lawrence Erlbaum.

37.

Lee

Faber

R. J.

(2007). Effects of product placement in on-line games on brand memory: A perspective of the limited-capacity model of attention. Journal of Advertising, 36, 75-90.

38.

Lee

Lang

(2009). Discrete emotion and motivation: Relative activation in the appetitive and aversive motivational systems as a function of anger, sadness, fear, and joy during televised information campaigns. Media Psychology, 12, 148-170.

39.

Lewis

Porter

(2010). In-game advertising effects: Examining player perceptions of advertising schema congruity in a massively multiplayer online role-playing game. Journal of Interactive Advertising, 10(2), 46-60.

40.

Lynch

J. G.

Jr. Srull

T. K.

(1982). Memory and attentional factors in consumer choice: Concepts and research methods. Journal of Consumer Research, 9, 18-37.

41.

Macmillan

N. A.

Creelman

C. D.

(2005). Detection theory: A user’s guide (2nd ed.). Mahwah, NJ: Lawrence Erlbaum.

42.

Mickley Steinmetz

K. R.

Kensinger

E. A.

(2009). The effects of valence and arousal on the neural activity leading to subsequent memory. Psychophysiology, 46, 1190-1199.

43.

Miesse

S. E.

Sparks

J. V.

(2009, May). Emotional responding to violent and risky video game content. Paper presented at the annual meeting of the International Communication Association, Chicago, IL.

44.

Park

(2006). Video game play and motivation: Variation in appetitive and aversive motivational system activation as a function of virtual threat level (Unpublished doctoral dissertation). Indiana University, Bloomington.

45.

Potter

R. F.

(2009). Double the units: How increasing the number of advertisements while keeping the overall duration of commercial breaks constant affects radio listeners. Journal of Broadcasting & Electronic Media, 53, 584-598.

46.

Ravaja

Saari

Salminen

Laarni

Kallinen

(2006). Phasic emotional reactions to video game events: A psychophysiological investigation. Media Psychology, 8, 343-367.

47.

Rothschild

M. L.

Qualheim

Deith

Hyun

Y. J.

(1990). Constructing a more difficult recognition test for television commercial scenes. Advances in Consumer Research, 17(1), 785-791.

48.

Schneider

E. F.

Lang

Shin

Bradley

S. D.

(2004). Death with a story: How story impacts emotional, motivational, and physiological responses to first-person shooter video games. Human Communication Research, 30, 361-375.

49.

Schneider

Chein

J. M.

(2003). Controlled & automatic processing: Behavior, theory, and biological mechanisms. Cognitive Science, 27, 525-559.

50.

Shapiro

M. A.

Fox

J. R.

(2002). The role of typical and atypical events in story memory. Human Communication Research, 28, 109-135.

51.

Shook

Larson

DeTarsio

(2008). Television field production and reporting (5th ed.). New York, NY: Pearson.

52.

Sparks

J. V.

Matthews

Chung

(2011). The influence of arousal on encoding and storage of information in pleasant television messages. Southwestern Mass Communication Journal, 27(1), 45-56.

53.

Vorderer

Klimmt

Ritterfeld

(2004). Enjoyment: At the heart of media entertainment. Communication Theory, 14, 388-408.

54.

Williams

Petrosky

Hernandez

Page

(2011). Product placement effectiveness: Revisited and renewed. Journal of Management and Marketing Research, 7, 1-24.

55.

Wilson

Kerr

J. H.

(1999). Affective response to success and failure: A study of winning and losing in competitive rugby. Personality and Individual Differences, 27, 85-99.

56.

Yegiyan

N. S.

Lang

(2010). Processing central and peripheral detail: How content arousal and emotional tone influence encoding. Media Psychology, 13, 77-99.