The Effect of Slow Motion Videoon Consumer Inference

Yunlu Yin , Jayson S. Jia, and Wanyi Zheng

AbstractVideo advertisements often show actors and influence agents consuming and enjoying products in slow motion. By prolonging

depictions of influence agents’ consumption utility, slow motion cinematographic effects ostensibly enhance social proof and sig-

nal product qualities that are otherwise difficult to infer visually (e.g., pleasant tastes, smells, haptic sensations). In this research,

seven studies, including an eye tracking study, a Facebook Ads field experiment, and lab and online experiments—all using real ads

across diverse contexts—demonstrate that slow motion (vs. natural speed) can backfire and undercut product appeal by making

the influence agent’s behavior seem more intentional and extrinsically motivated. The authors rule out several alternative expla-

nations by showing that the effect attenuates for individuals with lower intentionality bias, is mitigated under cognitive load, and

reverses when ads use nonhuman influence agents. The authors conclude by highlighting the potential for cross-pollination

between visual information processing and social cognition research, particularly in contexts such as persuasion and trust, and

they discuss managerial implications for visual marketing, especially on digital and social platforms.

Keywordsaudiovisual media, eye tracking, intentionality, slow motion video, visual marketing

Online supplement: https://doi.org/10.1177/00222437211025054

Video marketing is the dominant form of advertising.Television aside, over 1.9 billion YouTube users watch onebillion hours of ad-filled videos every day, while mobilevideo ad spend alone accounts for over 72% of digital adspend (Biteable 2021). Despite being more engaging, memora-ble, and popular than other forms of content (Bowman 2017),video advertisements are fundamentally limited to visual (andaudio) information. This makes it necessary for advertisers toemploy actors and influence agents to signal a product’s nonvi-sual attributes (e.g., taste of food, feel of using facial cleanser,motion of riding a bicycle), often by demonstrating and reactingto consumption (e.g., smiling while chewing on food), and toprovide social proof of the product’s utility (Cialdini 2007).

In such marketing contexts, one of the most common andeasy-to-implement cinematographic techniques is to portrayconsumption in slow motion. For example, food advertisementsoften show actors chewing and smiling in slow motion to signalpalatability. Similarly, face wash and shampoo commercialsoften portray actors lathering and rinsing luxuriantly in slowmotion to depict sensations of cleanliness and comfort (forexamples of real ads featuring slow motion cinematographiceffects, see Web Appendix A). Such effects are also pervasiveon social media. Indeed, most social media apps have built-infeatures for recording slow motion video, making it easy to

implement for both influencers and everyday consumers. Alsoreflecting its popularity, all top-five mobile phone manufactur-ers in 2019 (Apple, Samsung, Huawei, OPPO, and Vivo) pro-moted their mobile phones’ ability to “record high-qualityslow motion footage,” with many review websites even includ-ing slow motion video making as a performance metric formobile phones (Singh 2019).

Despite its ubiquity, do slow motion video effects necessar-ily make the advertised product more attractive? As a rough testof the prevalence and efficacy of slow motion in the market-place, we conducted a pilot study analyzing real ads onYouTube, the biggest online video platform worldwide, witha market reach of 90% in the United States (Clement 2019).Specifically, we compiled the top 100 search results (i.e.,videos) for the keyword “food commercial” (search date:June 2020) and coded their usage of slow motion as both

Yunlu Yin (corresponding author) is Assistant Professor of Marketing, School of

Management, Fudan University, Shanghai, China (email: yinyunlu@fudan.edu.cn).

Jayson S. Jia (corresponding author) is Associate Professor of Marketing, HKU

Business School, University of Hong Kong, Hong Kong SAR China (email:

jjia@hku.hk). Wanyi Zheng is a PhD candidate, HKU Business School,

University of Hong Kong, Hong Kong SAR China (email: zwymkt@connect.

hku.hk).

Article

Journal of Marketing Research

1-18

© American Marketing Association 2021

Article reuse guidelines:

sagepub.com/journals-permissions

DOI: 10.1177/00222437211025054

journals.sagepub.com/home/mrj

dichotomic and continuous measurements, in other words,whether the video uses slow motion (no/yes) and how often ituses it (1= “not at all,” and 9= “very often;” rated by two inde-pendent coders; r= .98, p< .001; for details, see Web AppendixB). To explore the marketing efficacy of slow motion, we col-lected several video performance metrics, including eachvideo’s ranking for the keyword as well as number of views,likes, and comments. We found that although 57% of the adsemployed slow motion cinematographic effects, the use of slowmotion did not predict the ranking of the video (Wilcoxonrank-sum test, z=−1.50, p= .13), which suggests that the useof slow motion is not specific to top-ranked (or low-ranked)videos. More importantly, contrary to the notion that slowmotion makes ads more aesthetically attractive, or that its wide-spread adoption is in itself evidence of marketing efficacy, wefound that slow motion video ads have fewer views, likes, andcomments (Table 1). These results remain robust when we usethe continuous measurement of slow motion usage and after con-trolling for video rankings (ps < .01; see Web Appendix B).

The pilot study, which examines real ads in the marketplace,raises the possibility that despite being commonplace, slowmotion video might create undesirable consequences. Indeed,previous social psychology research examining the impact ofslow motion replays in adjudication contexts suggests thatslowed video footage of human actions can unwittingly createunfavorable perceptions. For example, professional soccer ref-erees perceive slow motion replays of fouls as more “willful,deliberate, and pre-meditated” (compared with live footage),and subsequently end up serving more severe penalties (Spitzet al. 2018). Similarly, simulated juries viewing actual surveil-lance footage of violent crimes will perceive the crime as moreintentional and feel that perpetrators are more responsible forharm when the surveillance video is played at a relativelyslower frame rate (Caruso, Burns, and Converse 2016).

Given the results of our pilot study, could analogous meta-inferences occur when slow motion is used in marketing con-texts? Although watching commercials or social media videosis radically different from making decisions in a jury or as areferee (which involve moral adjudication), meta-inferencesarising from perceived intentionality can also occur duringconsumer information processing (Newman, Gorlin, andDhar 2014; Reich, Kupor, and Smith 2017). Perceptions of

intentionality are particularly relevant for advertisements thatrely on behavioral cues from on-screen influence agents (i.e.,actors or social media influencers) to signal product utility. Inthese contexts, whether the actions, reactions, and expressionsof an influence agent are construed as intentional and willfulas opposed to spontaneous and unplanned may directly affectconsumer perceptions and ad efficacy. For example, a viewermay perceive a video of an influence agent joyfully eating choc-olate as a more or less reliable signal of the influence agent’sconsumption utility and thus the predicted utility of the choco-late, depending on whether the viewer construes the influenceagent’s behaviors and reactions as intentional or unintentional.Even the most blissful post-chocolate smile may be discountedif consumers feel that extrinsic motivations (such as the inten-tion to persuade) belie the influence agent’s behavior.

Next, we outline the conceptual background for the pro-posed effect. By circumscribing the conditions under whichslow motion video may hinder versus foster ad persuasiveness,our findings have direct practical implications for how market-ing managers and social influencers design visual marketinginvolving dynamic cinematographic effects. In investigatingthis specific but commonplace effect, we also explore thegeneral question of how consumers make social-cognitiveinferences in dynamic visual marketing environments—atopic of growing importance given the rising importance ofnew media and social commerce, which are inherently video-and social-based platforms, as marketing channels.

Conceptual Background

Slow Motion and Perception of IntentionalitySlow motion video refers to a cinematographic technique inwhich a video’s frame rate is slowed so that the objectsappear to be moving slower than normal (Zettl 2011).Pioneered in early twentieth century film to create dramaticeffect, slow motion is now used across a variety of contextsoutside of cinema (Zettl 2011). In video advertising, marketershave long used slow motion to signal enjoyment (Areni and Black2015; Bryant and Veroff 2007), particularly for products featuringnonvisual attributes, such as taste, smell, emotional stimulus, andhaptic sensations. Slow motion is also used to aid in adjudication,for example for replays and refereeing in professional sports andvideo surveillance footage as legal evidence in court. Slowmotion effects are also widely used in news footage to highlightor sensationalize events (Barnett and Grabe 2000; Wöllner,Hammerschmidt, and Albrecht 2018). More recently, slowmotion visual effects have become popular on social media, inmemes, and on live streaming platforms, partly because slowmotion effects are accessible and built-in functions on mostsocial media apps (e.g., Instagram, TikTok). This allows users toslow down and highlight specific temporal sequences of a videowith just a few clicks or finger taps.

Prior research on visual cognition has documented a range ofdifferent slow motion video effects (Table 2), many of whichare aesthetic–emotional effects. For example, TV viewers

Table 1. Results of the Effect of Slow Motion on YouTube Video

Ad Metrics.

Views Likes Comments

Slow motion

ads

1,177,015

(2,077,266)

17,790

(34,234)

1,234

(2,854)

Non–slowmotion ads

3,638,711

(7,309,908)

49,914

(90,258)

2,525

(3,734)

P value p= .037

(*p= .028)

p= .031

(*p= .045)

p= .063

(*p= .019)

Notes: Standard deviations are in parentheses. p-values are computed using

Student’s t-test (p) and nonparametric Wilcoxon rank-sum test (*p) forbetween-condition comparisons.

2 Journal of Marketing Research

perceive news stories that feature slow motion video as moresensational and emotional (e.g., more threatening, more stimu-lating) but less informative than news stories that do not employslow motion video (Barnett and Grabe 2000). Relatedly, moviescenes played in slow motion (vs. natural speed film scenes)lead to “higher perceived emotional valence” of the video(Wöllner, Hammerschmidt, and Albrecht 2018). Slow motioncinematography can also change viewers’ psychophysiologicalreactions: when watching slow motion film scenes, viewershave relatively higher gaze dispersion and a smaller centerbias, which suggests they are paying greater attention todetails (Hammerschmidt and Wöllner 2018).

Beyond aesthetic and psychophysiological effects, a smallbody of literature has found that slow motion video can engen-der social cognition inferences. In particular, human actionsplayed out in slow motion look more deliberate and intentional.In turn, this can create perceptions of intentionality and motiva-tion, which can bias decision making during adjudication judg-ments. For example, people judge “helmet-to-helmet” tackles inAmerican football, a move prohibited by the National FootballLeague, as more intentional when viewing the same footage inslow motion rather than at regular speed (Caruso, Burns, andConverse 2016). Such “slow motion intentionality bias” caneven occur for professional referees. In an official FIFAstudy, elite referees from five countries were asked to viewreal foul play situations taken from international soccermatches and provide their professional judgment. Refereesgave harsher penalties when video replays of potential foulswere played in slow motion rather than in real time (e.g., theygave a red instead of a yellow card; Spitz et al. 2018). In crim-inal justice contexts, simulated juries viewing video evidence inslow motion were more likely to presume criminal intentional-ity and guilt (Caruso, Burns, and Converse 2016).

Although such social cognition effects have only been doc-umented in the narrow context of adjudications, we investigate

whether a similar but more general effect (and by extension, aprocess that is not particular to adjudications) may occur in mar-keting contexts and override the aesthetic and visual benefits ofslow motion video. Although consumers do “judge” commer-cials or social media videos to some extent, there are major con-textual differences between information processing for visualmarketing versus adjudication contexts. For example, whereasvideo commercials usually present emotionally positive-valance scenarios such as influence agents consuming andenjoying products, adjudication contexts are mostly concernedwith moral transgressions or human conflict (e.g., criminalbehavior, physical violence). Furthermore, an influenceagent’s action intentionality may be less salient in consumercontexts: Whereas intentionality is highly relevant in adjudica-tions and moral decision making contexts (and is a basic prin-ciple in legal judgments), intentionality is not necessarily topof mind in marketing contexts, in which consumers areexposed to a variety of cognitive, social, and motivationalforces (Macinnis, Rao, and Weiss 2002; Petty and Cacioppo1986).

Nevertheless, there are also several features of human socialcognition that may trigger intentionality judgment biases inresponse to slow motion advertisements. First, humans havean inherent tendency to infer other people’s intentions fromtheir actions, and such tendency has a biological basis(Baldwin and Baird 2001; Blakemore and Decety 2001). Forexample, research using functional imaging has shown thatthe temporal and parietal lobe (i.e., temporoparietal junction)is involved in reasoning about other people’s beliefs, intentions,and desires (Saxe and Kanwisher 2003). Also, lesions to the lefttemporoparietal junction can impair cognitive processes specif-ically involved in reasoning about others’ intentionality(Blakemore and Decety 2001). Second, humans have aninnate sense of normal temporal speed against which theycompare events and behaviors and thus are able to

Table 2. Literature Review on The Effect of Slow Motion on Human Perceptions.

Study Context/Stimuli Boundary Conditions Key Findings

Barnett and Grabe

(2000)

TV news stories News stories using slow motion are evaluated as more

sensational but less informative.

Caruso et al. (2016) Footage of murder and

violent contact in American

football

Salient time reminder Slow motion videos cause viewers to perceive actions

as more intentional, resulting in harsher punishment.

Spitz et al. (2017) Replays of foul play situations

in soccer

Referee expertise; type of

foul

Slow motion footage leads to higher decision accuracy

for technical fouls only.

Hammerschmidt and

Wöllner (2018)

Movies Individuals attend to more details (more fixations/

saccades) in slowed film scenes.

Wöllner et al. (2018) Movie, ballet, and sports clips Slow motion video leads to higher perceived

emotional valence, lower respiration rates, and

smaller pupillary diameters.

Spitz et al. (2018) Replays of foul play situations

in soccer

Referees penalize players more severely when foul

replays are played in slow motion.

Present research Real ads and live streaming

clips

Intentionality bias;

cognitive resource

availability;

humanness of agents

Slow motion ads make advertised products less

attractive and appealing.

Yin et al. 3

automatically detect when slow motion is occurring (Arstila2012; Wöllner, Hammerschmidt, and Albrecht 2018). Puttogether, social-cognitive inferences arising from temporalspeed may be difficult to de-bias. For example, in jury contexts,even explicitly highlighting the objective passage of time (byallowing viewers to see both natural speed and slow motionvideos or by prominently displaying a clock on the video thatshows the actual passage of time) cannot eliminate the inten-tionality bias triggered by the slow motion video (Caruso,Burns, and Converse 2016).

Next, we discuss psychological processes that may causeslow motion effects to backfire and generate negative consumerinferences in marketing contexts. In particular, we considerwhether slow motion effects make influence agents’ behaviorsseem more intentional and willful and make their reactions toproducts seem less spontaneous and genuine, thus underminingthe persuasiveness of video ads.

Intentionality and Consumer InferencePerception of intentionality plays a significant role not only inshaping consumer evaluations of products (Newman, Gorlin,and Dhar 2014; Reich, Kupor, and Smith 2017) but also inmaking social cognition attributions in general (Gilbert andMalone 1995; Heider 1958; Rosset 2008). We postulate thatstronger perceptions of intentionality of an influence agent’sbehavior in slow motion video will lead consumers to attributethe influence agent’s consumption reaction (e.g., how muchthey smile) more to factors extrinsic (rather than intrinsic) tothe product. This in turn can generate worse attitudes towardthe advertised product relative to when the product is presentedat natural speed.

We next draw on attribution theory literature, which explainshow individuals arrive at causal explanations for specific events(Harvey andWeary 1984; Kelley and Michela 1980), to explainthe impact of perceived intentionality on consumer evaluations.There are two main types of causal attributions that consumersmay generate in response to influence agents’ behavior: (1)attributions to the intrinsic characteristics of a focal object or(2) attributions to forces that are extrinsic to the focal object.Thus, when viewing a typical consumption skit in a video(e.g., a person eating pizza with a smile), viewers may attributethe observed behavioral outcome (i.e., eating with a smile) totwo possible underlying causes: (1) intrinsic product-relatedfactors (e.g., taste and quality), which reflect experienced plea-sure and consumption utility, or (2) extrinsic factors unrelatedto the product (e.g., hunger, intentional expression of certainreactions or emotions for social signaling purposes [Settle andGolden 1974; Sparkman and Locander 1980]), which are lessreflective of the product’s consumption utility. As a result,when slow motion video enhances the perceived intentionalityof an influence agent, we expect viewers to attribute the agent’s(re)actions in the video (e.g., favorable reactions to consump-tion) more to extrinsic forces (e.g., agent’s intention to signalquality) rather than intrinsic factors (e.g., how a pizza actuallytastes). This in turn may undermine the relative efficacy of

the advertisement (i.e., how persuaded consumers are by thead), which may lead to relatively worse attitudes toward theadvertised product or brand and lower purchase intentions(Newman, Gorlin, and Dhar 2014; Reich, Kupor, and Smith2017).

In summary, we predict that slow motion can backfire andgenerate negative consumer inferences in marketing contextsby triggering perception of intentionality in consumers(Figure 1). We hypothesize that:

H1: A slow motion (vs. natural speed) ad depicting con-sumption leads to relatively more negative evaluations ofthe advertised product.H2: The effect of slow motion on product evaluation ismediated by heightened perceptions of intentionality inthe influence agent’s observed behaviors.

Notably, our proposed mechanism based on perceivedintentionality differs from the Persuasion Knowledge Model(PKM; Friestad and Wright 1994), a theory of how consum-ers use their knowledge of persuasion motives and tactics tointerpret, evaluate, and respond to persuasion attempts. ThePKM posits that consumers may develop dynamic knowledgestructures about persuasion and draw on this knowledgesystem to identify marketers’ attempts to persuade them. Inthis framework, persuasion knowledge refers to a group ofintuitive theories or beliefs about persuasion, including iden-tifying which persuasion tactics marketers use, how thesemarketing tactics may trigger psychological consequences,which tactics are effective or appropriate under different con-texts, and what firms’ goals and motives are. These processesdiffer in important ways from our proposed mechanism andstudy context: we study viewers’ perceptions of and attitudestoward the influence agent (i.e., the actor or social influencer)who engages in consumption behavior in a video, rather thanthe marketer or advertiser of the product (who is usuallyunseen). Our proposed intentionality-based mechanism isthus a more direct and parsimonious explanation for the pro-posed effect than the PKM (which requires consumers tocounterargue the unseen marketer, which is an additionallevel of inference beyond thinking about the influenceagent). Furthermore, our mechanism, which relates to con-sumers’ social cognitions of a visible agent, is a better fitfor new media contexts (e.g., influencer marketing, livestreaming, social commerce), in which the distinctionbetween influence agents and marketers can be blurred ornonexistent (as compared to traditional TV advertisingcontexts).

Our proposed mechanism also has a visual perceptionelement that the PKM does not have. If the effect of slowmotion video is indeed driven by heightened perceptions ofaction intentionality, we also expect consumers to engage ingreater visual social monitoring (of the influence agent). Wepredict that this effect can be captured by the degree to whichviewers’ visually attend to (i.e., gaze at) the influence agents’eyes in the slow motion condition (vs. natural speed condition).

4 Journal of Marketing Research

Indeed, eyes are often metaphorically referred to as the“windows of soul,” and during face-to-face social interaction,eyes often serve as the most important source of informationregarding another person’s mental state (Emery 2000;Hamilton 2016; Khalid, Deska, and Hugenberg 2016; Senjuand Johnson 2009). In addition, humans intuitively and activelyengage in eye contact and follow the gaze of another personwhen trying to detect their intentions and beliefs (Colombatto,Chen, and Scholl 2020; Grossmann 2017; Kuhn, Tatler, andCole 2009). Drawing on these findings, we postulate thatvisual attention to influence agents’ eyes may serve as aproxy of perception of intentionality and mediate the effect ofslow motion on consumer attitudes toward the advertisedproduct.

Boundary ConditionsTo provide more process insight on the proposed intentionalitymechanism, we also explore the conditions under which slowmotion video may hinder versus foster the ad’s persuasiveness.If consumer inferences of intentionality indeed mediate theeffect, the effect should be mitigated when the (cognitive) infer-ence process is inhibited. One circumstance in which this mayoccur is when mental resources are impaired, such as when con-sumers are cognitively busy when viewing the commercial(e.g., if the ad has an information-intensive soundtrack).Indeed, prior research has shown that inferring and construingothers’ intentionality and motives requires higher-order inferen-tial processing, and that cognitive load can attenuate consum-ers’ ability to infer others’ motives and intentions acrossvarious contexts (Campbell and Kirmani 2000; Waytz et al.2010; Williams, Fitzsimons, and Block 2004). Consequently,we predict that when consumers view slowed commercialsunder cognitive load, they will be unable to counterargue theinfluence agents’ behavior and are less likely to make infer-ences about intentions, which should mitigate the negativeimpact of slow motion on ad persuasiveness. We thus hypoth-esize that:

H3: The effect of slow motion on product evaluationattenuates when consumers are under cognitive load.

As a second qualification, we expect the effect to be contingenton whether influence agents are human. People are generallyless likely to attribute intentionality and mental states to nonhu-man entities, which share less morphological similarity withreal humans (Epley, Waytz, and Cacioppo 2007). If the effectof slow motion video on product evaluation is driven by theway consumers perceive and interpret the motivation behindinfluence agents’ consumption behavior, the effect may bediminished when ads employ nonhuman agents to demonstrateproduct utility (compared with when human influence agentsare used). We thus postulate:

H4: The effect of slow motion on product evaluationattenuates when influence agents are nonhuman entities.

Empirical OverviewWe present seven studies (and two additional studies reported inthe Web Appendix), including a field experiment and an eyetracking study, to test these hypotheses. Using real commercialsacross multiple product categories (including both ads that wereoriginally filmed in slow motion and ads that we slowed downfrom natural speed; Web Appendix A), we documented theimpact of slow motion effects on visual attention, attitudes,preferences, and click-throughs (of real online consumers).Study 1 provides initial evidence of the main effect in a fieldsetting. Studies 2a and 2b explore the underlying process byexamining the mediating role of perceived intentionalityusing eye tracking as well as behavioral measurements.Studies 3–5 offer additional process insights by testing theoret-ically relevant moderators: The effect weakens among individ-uals with lower intentionality bias (Study 3) and whenconsumers’ cognitive capacity is constrained (Study 4), andthe effect can reverse when the ad uses nonhuman influence

Figure 1. Conceptual framework.

Yin et al. 5

agents (Study 5). The moderators we tested have importantmanagerial implications because they define conditions underwhich slow motion video advertisements can be effective (orat least less ineffective). Finally, we conducted a within-papermeta-analysis to demonstrate the robustness of our findingsacross experiments.

Study 1Study 1 aims to provide initial evidence that applying slowmotion to a video advertisement can affect its marketing effi-cacy. To enhance managerial relevance, we conducted a fieldstudy that measured actual consumer choice on Facebook.Moreover, we intentionally selected a product that was locallyunfamiliar at the time of the study to mitigate the possibilitythat prior exposure to the product and its advertisements mayhave driven the effect.

ProcedureWe ran ads for Soylent, a beverage brand, on Facebook’s adver-tising platform with the objective of increasing click-throughrates (CTRs) for the target commercials. CTR is a commonlyused ad performance metric that is a measurement of the per-centage of clicks an ad receives out of all unique individualsreached. We utilized Facebook’s split-test function, which ran-domly assigned users to one of two advertisement conditions(i.e., slow motion vs. natural speed), both of which used thesame ad (but played at different speeds). The video commercial,once clicked by viewers, led to the official website of the bev-erage brand. Our campaign targeted users age 18 or above whowere currently living in Hong Kong, and it ran for 36 consecu-tive hours (Final N= 26,972; i.e., unique impressions). At thetime of the study, Soylent was only available in the NorthAmerican market and had not officially launched (or been pro-moted) in the Hong Kong market.

We played the original commercial for Soylent, which wasdownloaded from the company’s official website. The commer-cial, which showed different people consuming and enjoyingSoylent, was originally in slow motion (approximately .5Xnatural speed, rated by two independent judges). We sped thevideo’s frame rate up 2X in the natural speed condition. Thispreempted the possibility that the effect was caused by ourmanipulation of slow motion negatively affecting the advertise-ment’s intended design and aesthetics. As Facebook recom-mends that advertisers limit video length to around15 seconds, we edited the original commercial to meet thevideo length requirement. The videos are in high resolution(720p or above) throughout all studies (see Web AppendixA). Finally, to isolate the pure visual processing effect and side-step potential audiovisual mismatch problems, we muted thesound of the commercial. Viewers should not have perceivedthis to be unusual, as video ads on Facebook (and Instagram)are played on mute by default. We used Apple’s iMovie soft-ware for video editing and playback speed adjustments.

Pretest. A total of 106 undergraduate students from a majorHong Kong university (45 women, Mage= 20.62 years)viewed the commercial in either slow motion or natural speedand indicated (1) “to what extent do you think the person inthe video was expressing her emotional reaction in slowmotion,” (2) “to what extent do you think the person in thevideo was moving in slow motion,” and (3) “to what extentdid the video feel like it was playing in slow motion,” on a nine-point scale (1= “not at all,” and 9= “to a great extent”).Responses to these items were aggregated to create a slowmotion index (α= .88). As expected, the slowed ad scoredhigher on the slow motion index (p < .001; for pretests ofslow motion video across studies, see Web Appendix C).

Results and DiscussionAltogether, 565 Facebook users clicked on the video, yielding a2.09% total click-through rate (as a benchmark, the averageCTR on Facebook is .90%; Irvine 2019). More importantly, theCTR in the slow motion condition (1.77%) was significantlylower than in the natural speed condition (2.17%; χ2(1)= 4.69, p= .03, odds ratio= 1.23). Overall, Study 1 provides initial evidencefrom a field study that applying slow motion in commercials for anew product can backfire by making the product less attractive andappealing. It is worth noting that product unfamiliarity or noveltydid not impede the effect, nor did the fact that the commercial wasoriginally in slowmotion. The latter point shows the effect was notmerely the result of us creating a subpar advertisement (indeed, thevideo we created performed better in terms of real consumerresponse). In the following studies, we extend the study paradigmto a range of real ads that varied in visual properties, product cat-egories, original speed, and country of origin to enhance externalvalidity and establish robustness.

Study 2Study 2 explores the process underlying the effect by using botheye tracking and scale-based measurements. Prior research onthe effects of slow motion video on adjudication judgmentssuggests that human actions viewed in slow motion are associ-ated with greater perceived intentionality (Caruso, Burns, andConverse 2016; Spitz et al. 2018). We therefore theorize thatslowing a video advertisement is more likely to lead consumersto perceive that influence agents in commercials are purposelysignaling consumption enjoyment (and that influence agents areintentionally behaving in a manner caused by exogenous moti-vations, e.g., the intention to persuade). This in turn may reducethe degree to which viewers attribute the influence agents’ pos-itive reactions to intrinsic product features such as productquality (as opposed to extrinsic factors like motivation to per-suade) and lead to relatively worse attitudes toward the product.

Study 2aStudy 2a offers initial process evidence by measuring consumereye movement as a behavioral mediator. Past literature has

6 Journal of Marketing Research

shown that during face-to-face social interactions, people oftengaze at the eyes of their interaction partner to discern their psy-chological state (e.g., emotional state, focus of visual attention)or infer their intentions (Birmingham, Bischof, and Kingstone2008; Grossmann 2017). Conversely, insufficient fixation onpartners’ eyes may lead to difficulties in recognizing, or evendistort perceptions of, their emotional and cognitive states.Indeed, persistent avoidance of eye contact is also a behavioralmaker of social interaction and communication difficulties anddisorders such as autism (Baron-Cohen et al. 1995; Guillonet al. 2014). Moreover, humans have an innate tendency toactively follow the gaze of interactive partners when trying tofigure out their intentions and beliefs (Colombatto, Chen, andScholl 2020; Kuhn, Tatler, and Cole 2009). Thus, if slowmotion video does indeed heighten perceptions of intentionalityof the agent’s consumption behavior, we should expect consum-ers’ visual attention to focus relatively more on influence agents’eyes when a commercial is played in slow motion (vs. in naturalspeed). Furthermore, if consumers paying greater visual attentionto the influence agent’s eyes does indeed reflect their attempts todiscern intentionality, we also expect visual attention to mediatethe impact of slow motion video on product evaluation. Forrobustness, we tested the effect on two new real video advertise-ments that vary in product category, video length, agent, and con-sumption behavior.

Method

Procedure. A total of 199 undergraduate students (128women, Mage= 20.72 years) from a major university in HongKong participated in this in-lab study for HK$80 (∼US$10). Allparticipants had normal or corrected-to-normal vision.Participants were randomly assigned to one of two conditions(slow motion vs. natural speed) and viewed two advertisementvideos (all from the same condition, in random order). Thestimuli were real video commercials for Sanyo instant noodlesand Neutrogena facial cleanser. We edited the videos in thesame manner as described in Study 1 (Web Appendix A). Afterviewing each video, participants were asked to report their atti-tudes on an 11-point scale (i.e., “how much do you like theproduct in the video”; 1= “not at all,” and 11= “very much”).

Eye tracking measures. Participants’ eye movements wererecorded using the Xinsight Gazelab system (http://www.xinsight.cn/html/EN/) at a sampling rate of 30 Hz with aspatial resolution of .1° of visual angle. All stimuli were pre-sented on a 15-inch LCD screen with a resolution of 1680×1050 in full-color bitmaps. To analyze the gaze pattern duringcommercial viewing and test the proposed hypotheses, twoareas on the video clips were drawn as areas of interest(AOIs): the influence agents’ eyes and nose regions (WebAppendix E). The latter served as the control area to rule outthe possibility that slow motion may invite general attentionto the face rather than to the eyes in particular (Auyeunget al. 2015; Yu, Duan, and Zhou 2017). We selected the nosearea as a control because it has a similar surface area, is animportant facial component, and is also close to the eyes,

which makes for a fair comparison. Our analysis focuses on fix-ation duration, a common eye tracking measurement defined asthe duration of all the fixations made on an AOI during thewhole video-viewing period. To facilitate between conditionand across commercial comparisons, we created a normalizedmeasure of fixation duration by dividing the fixation durationby the length of each video clip (Auyeung et al. 2015; Yu,Duan, and Zhou 2017).

Results

Behavioral results. We first tested for the previously docu-mented (negative) main effect of slow motion cinematography.Because each participant viewed two commercials (and thusevaluated two products) and there were two conditions foreach commercial (i.e., slow motion vs. natural speed), we con-ducted a hierarchical linear mixed-effects (HLM regression)model to control for the subject- and video-level randomeffects (Fisher, Newman, and Dhar 2018; Spiller andBelogolova 2016). Specifically, we ran the HLM regressionusing the “lme4” (Bates et al. 2014) and lmerTest(Kuznetsova, Brockhoff, and Christensen 2017) packages inR, with the slow motion condition (1= slow motion, 0=natural speed) as a fixed effect and random effect on interceptsfor subjects and commercials. Significant levels of the effectwere obtained using likelihood ratio tests of the full modelwith the fixed effect against the null model without the fixedeffect. As expected, the HLM regression replicated the effectof slow motion (likelihood ratio test, χ2(1)= 6.01, p= .014):Participants in the slow motion condition, compared withthose who viewed the same commercials in natural speed,reported worse attitudes toward the advertised product (b=−.56, SE= .23, t=−2.45, p= .015).

Eye tracking results. We next explored whether slowmotion shifted gaze patterns during commercial viewing asmeasured by the percentage of fixation duration on the areaof interest. As shown in Figure 2, the HLM analysis withvideo speed (natural speed=−1, slow motion= 1), area ofinterest (eye= 1, nose=−1), and their interaction as fixedeffects yielded a significant main effect in area of interest(χ2(1)= 88.21, p < .001), which was qualified by a significantarea of interest × video speed interaction (χ2(1)= 5.33,p= .021). In particular, participants in the slow motion condi-tion paid relatively more attention to the influence agents’ eyesthan did participants in the natural speed condition (b= .04,SE= .02, t= 2.39, p= .018; for raw values of fixations foreach condition and video, see Web Appendix F).

Because people may generally pay more attention to details inslowed videos (Hammerschmidt and Wöllner 2018), one maywonder if slowmotion video invites more visual attention to influ-ence agents and their facial reactions as a whole (as opposed toonly the eyes). To explore this possibility, we next testedwhether slow motion video affects gaze on the control AOI(i.e., nose area), which is still a prominent facial area. However,we observed no differences in fixation duration on the noseAOI between conditions (b= .01, SE= .01, t= .80, n.s.),

Yin et al. 7

suggesting that slow motion specifically increases consumers’visual attention to agents’ eye region rather than agents’ entireface. Finally, we tested for the possibility of gender or age differ-ences in gaze patterns (Mercer Moss, Baddeley, and Canagarajah2012; Vassallo, Cooper, and Douglas 2009) or visual interest andattention (for example, whether the stimuli we used was moreappealing to one demographic) by including demographic varia-bles (i.e., gender and age) as statistical controls to the HLMregression. We obtained consistent results and did not find evi-dence of differences across demographics. The HLM regressionincluding age and gender as covariates yielded a similar maineffect of slow motion, and these control variables did notpredict product rating or participants’ eye movement (ps > .1;Web Appendix G).

Mediation results. As a test for the proposed mechanism,we next tested whether visual attention to influence agents’eyes can explain and mediate the effect of slow motion videoon product evaluation. Because each participant viewed twocommercials and rated two different products (visual attention:r= .56, p < .001; product rating: r= .20, p= .005), we first aver-aged responses for each participant before running a mediationanalysis with 5,000 bootstrapping samples (PROCESS Model4; Hayes 2013). Results confirmed that the effect of slowmotion on product liking could be explained by greater atten-tion toward agents’ eyes as measured by fixation duration(95% CI: [−.281, −.022], excluding zero; Figure 3).

Discussion. Study 2a provides a preliminary theory test byshowing that consumers’ gaze patterns and visual attention toinfluence agents’ eyes mediate the effect. Besides offeringmechanism evidence, Study 2a also demonstrates that slowmotion ads may elicit a different cognitive and behavioral

response from the consumer compared to natural speed ads(i.e., by generating greater viewer cognition of action intentionsand subsequently greater visual attention on influence agents’eyes).

Nonetheless, the results of Study 2a also raise several alter-native possibilities. First, prior literature suggests that there arecultural differences in human eye movements. For example,people in cultures in which emotional subduction is the norm(e.g., the East Asian sample in this study) may focus relativelymore on the eyes when interpreting others’ emotion (Yuki,Maddux, and Masuda 2007). Thus, it might be possible thatthe mediating effect of visual focus on influence agents’ eyesis specific for East Asian cultures. Second, one may wonderwhether other factors such as the extent to which consumersperceive the video to be attention-grabbing, engaging, or evenboring may drive the effect of slow motion on ad persuasive-ness (and eye movements). We test these alternative explana-tions and provide more evidence of an intentionality-basedprocess in the following studies.

Study 2bThe goal of Study 2b is twofold. First, we aim to provide addi-tional process evidence by measuring and testing the mediatingrole of perception of intentionality. Second, we evaluatewhether slowing videos changes the visual aesthetic character-istics of the commercials or how participants perceive theviewing experience, which might be alternative explanationsfor the effect.

Method

Procedure. A total of 313 United States residents (165women, Mage= 37.38 years) recruited from AmazonMechanical Turk (MTurk) participated in the online study inexchange for a small amount of monetary compensation. Byrandom assignment, they viewed a commercial for Sanyoinstant noodles (from Study 2a) in either slow motion or

Figure 3. Gaze pattern mediates the effect of slow motion video on

product evaluation (Study 2a).

*p< .05.**p< .01.Notes: Value in parentheses indicates the effect of slow motion video on the

dependent variable after controlling for the mediator.

Figure 2. Slow motion video moderates consumers’ gaze pattern

(Study 2a).Notes: Error bar indicates ±1 SE; NS= not significant. *p< .05 based on HLM

regression model.

8 Journal of Marketing Research

natural speed. After viewing the video, participants were shownan image of the product and rated their willingness to pay (WTP)for the advertised product on an eight-point scale, with priceoptions increasing in one-dollar increments from one dollar toeight dollars, in response to the prompt “Please indicate themost amount of money you would be willing to pay for theproduct advertised in the video, if it is available locally.”

Next, participants saw a screenshot of the influence agent con-suming the noodles and responded to two questions measuringperceived intentionality: (1) “To what extent was this person’sconsumption behavior willful?” and (2) “To what extent do youthink this person’s consumption behavior was intentional” on anine-point scale (1= “not at all,” and 9= “to a great extent;”adapted from Caruso, Burns, and Converse 2016). We averagedthe two items to create a perceived intentionality index (r= .72,p< .001). To rule out other visual processes, we also measured(1) ease of visual processing (“How difficult was it to visuallyfollow what was happening in the video?”), (2) engagement(“How engaging was it to view this video?”), and (3) the extentto which participants perceived the video to be attention-grabbing(“Towhat extent was this commercial video attention-grabbing?”).We also considered the possibility that viewers felt human behav-ior in slow motion looked abnormal by asking, “to what extentwas this person’s consumption behavior weird” (all on nine-pointscales, 1= “not at all,” and 9= “extremely”). In addition, we alsoconducted a poststudy test to ensure that participants could under-stand the two-item intentionality measurement in the current study(Web Appendix H).

Results. As hypothesized, we found that slow motion videoreduced self-reported WTP (Mnatural_speed= 3.58, SD= 1.96vs. Mslow_motion= 3.01, SD= 1.79; t(311)= 2.67, p= .008,d= .30). In addition, the slowed commercial (vs. commercialin natural speed) was associated with greater perceived inten-tionality behind the influence agents’ behavior (Mnatural_speed

= 7.06, SD= 1.71 vs. Mslow_motion= 7.49, SD= 1.38; t(311)=2.45, p= .015, d= .28). Furthermore, there was no correspond-ing effect on the visual and aesthetic properties of the video(i.e., visual difficulty, engagement, perceived attention grab-bing, and perceived weirdness; ps > .1).

Our conceptualization posits that slow motion video under-mines consumer evaluations of advertised products due toheightened consumer perceptions of intentionality (i.e., theovert intention and motivation to signal or persuade) inresponse to influence agents’ consumption behavior in suchcontexts. We conducted a bootstrapping mediation analysisusing 5,000 bootstrap samples (PROCESS Model 4; Hayes2013; Figure 4) to test this proposed mechanism and foundthat perceived intentionality significantly mediated the effectof slow motion on WTP for the advertised product (95% CI:[−.239, −.011], excluding zero).

Discussion. Using measures of perceived intentionality, we repli-cated the basic effect and provided support for the proposed mech-anism. We also did not find evidence that slow motion affects easeof visual processing, perceived engagement, and perceived

weirdness. This suggests that the effect is not driven merely bychanges in video aesthetic (and how consumers responded tothem) but rather by the way that consumers perceive and interpretthe slowed consumption behavior depicted by human influenceagents in the advertisements.

Study 3The purpose of Study 3 is twofold. First, we explore the moderat-ing role of individual differences in intentionality bias in theobserved effect, which serves as process evidence. As a corollaryof the proposed mechanism, we expect the effect to be moderatedby the viewer’s intentionality bias (Baldwin and Baird 2001;Rosset 2008; Slavny and Moore 2018), which indicates theextent to which individuals spontaneously tend to judge others’actions to be intentional. If the effect of slow motion on productevaluation is driven by viewers inferring influence agents’ con-sumption behavior and reactions as being motivated by extrinsicintentions, then the effect should be weaker among viewers withrelatively lower intentionality bias. Second, instead of using tradi-tional commercials as in previous studies, we played video clips ofprofessional live streamers. Live streaming, which occupies thespace between social and commercial media, is a major growingform of new media that is available globally on most majorsocial and video platforms including Facebook, TikTok,WeChat, and YouTube. For example, there were 425 millionusers of dedicated live streaming platforms in China alone inJuly 2018 (Zhou and Xiao 2019). Extending the effect into newsocial media domains, in which influence agents’ behaviors areless scripted, helps reduce concerns that the effect is limited to con-texts involving professional actors and enhances the ecologicalvalidity of our research.

MethodProcedure. A total of 348 United States residents (169 women,Mage= 40.30 years) recruited from MTurk participated in the

Figure 4. Intentionality mediates the effect of slow motion video on

product evaluation (Study 2b).

*p< .05.**p< .01.Notes: Value in parentheses indicates the effect of slow motion video on the

dependent variable after controlling for the mediator.

Yin et al. 9

experiments in exchange for monetary compensation.Participants were randomly assigned to one of two conditions(i.e., slow motion vs. natural speed) and viewed one realvideo steaming clip that showed streamers consuming andenjoying food (Web Appendix A). We edited the video clipin the same manner as in previous studies. After viewing theclip, participants were presented with an image of the productand reported their WTP for the advertised food on a drop-downlist with price options increasing in two-dollar increments from$2 to $16.

Next, participants completed an ostensibly unrelated taskthat measured their individual differences in intentionalitybias. The intentionality bias task was modified from Rosset(2008) and consisted of short sentences describing ambiguoushuman actions that could be deemed as either intentional orunintentional (Table 3).1 For each test sentence, participantsrated whether the action described in the sentence was done“on purpose” (i.e., intentionally) or “by accident” (i.e., uninten-tionally) and evaluated 22 sentences (in randomized order) intotal. Specifically, we instructed participants:

In this part, you will be asked to answer a few questions abouteveryday consumer behavior. For each question, you will read ashort sentence, which describes a simple action (for example, “hedeleted the email”), and you will need to decide whether youthink this action was generally done “on purpose” or ‘‘by accident.”For some actions, you may feel that both words are applicable butplease select just one option which you consider to be most suitable.You will need to evaluate 22 short sentences in total, and you willnot be timed and please take your time to read each short sentence.

To calculate the intentionality bias score, the total number ofintentional judgements was divided by the total number of sen-tences and multiplied by 100 to create a percentage score.

ResultsTo explore the impact of slow motion video on product evalu-ation, we ran a linear regression predicting participants’ WTPfor the advertised product, with video speed condition(natural speed=−1, slow motion= 1), intentionality biasscore (mean-centered for regression analysis), and the interac-tion between the two factors as independent variables. Ashypothesized, the regression analysis yielded a significantmain effect for video speed (b=−.41, SE= .17, t=−2.44,p= .015), which was qualified by a significant video speed ×intentionality bias interaction (b=−2.96, SE= 1.30, t=−2.27, p= .024; Figure 5). To decompose this interaction, weutilized the Johnson–Neyman technique (Johnson and Fay1950) to identify the range of intentionality bias for which thesimple effect of slow motion video was significant at a signifi-cance level of p < .05 (Spiller et al. 2013). This analysis

revealed a significant negative effect of slow motion video onWTP for any individual with an intentionality bias scorehigher than .25 (55.7%, or .19 SD below of the mean of .27)but not for those with an intentionality bias score less than .25.

DiscussionStudy 3 replicates the basic effect and provides further supportfor the intentionality account of the effect. In particular, themoderation effect suggests that perceived intentionality canexplain the impact of slow motion on ad efficacy, andwhether consumers perceive intentionality is moderated bythe degree of their own individual intentionality bias. Besidesshowing that the effect of slow motion can be explained bythe way consumers perceive and interpret the influenceagent’s intentionality, Study 3 shows that intentionality biasis an important individual-level trait that may limit the effec-tiveness of the advertising across mediums. Moreover, Study3 further suggests that individual differences in intentionalitymay also be a key factor for consumers to identify and infersocial influence attempts by monitoring influence agents invideo ads or social media. Finally, Study 3 extends the effectto the domain of live streaming, which suggests the effectmay extend to other video-based social media contexts, whichare now dominant contexts for marketing and advertising.

Study 4Studies 4a and 4b explore an important boundary condition ofthe effect. Extant literature suggests that inferring others’ inten-tionality and motives requires sufficient mental resources(Campbell and Kirmani 2000; Waytz et al. 2010; Williams,Fitzsimons, and Block 2004). As such, without adequate cogni-tive resources, consumers may be less able to infer agents’underlying motives in the first place. We therefore expect thathampering consumers’ overall cognitive capacity will impedetheir ability to perceive influence agents’ intentionality, thusmitigating the negative effect of slow motion on consumer eval-uations of video ads.

Table 3. Test Sentences in Individual Intentionality Bias Task (Study 3).

He hit the man with his car. He gave her the wrong change.

She burnt the meal. She broke the vase.

He tracked mud inside. He forgot his homework.

He arrived 5 min late

for class.

He bumped into a classmate

in the hall.

He broke the window. The painter inhaled the fumes.

He drank the spoiled milk. She woke the baby up.

He stepped in the puddle. He set off the alarm.

He jumped when the bell rang. He dripped paint on the canvas.

She kicked her dog. She left the water running.

He set the house on fire. He ate the bruised part of the apple.

She told the same joke twice. The girl popped the balloon.

1 For simplicity, we did not include reading comprehension filler questions fromthe original task (Rosset 2008) because they primarily serve as attention checks.

10 Journal of Marketing Research

Study 4aMethod

Procedure. A total of 501 United States residents (253women, Mage= 39.52 years) recruited from MTurk were ran-domly assigned to watch a commercial for Sony Bluetoothearbuds (Web Appendix A) in a 2 (video condition: slowmotion vs. natural speed) × 2 (cognitive load: low-load sound-track vs. high-load soundtrack) between-subjects design. Tomimic real-world marketing practices, we manipulated cogni-tive load by embedding auditory information into the ad sound-track, which we expected viewers to find distracting (Horvathand Burgyan 2011; Rees, Frith, and Lavie 2001). Thelow-load soundtrack condition included a soft music clipplayed in the background, while in the high-load soundtrackcondition, a voice-over narration of product features (overlaidon rhythmic music) was added into the background soundtrack.The voice-over narration was played only once in the middle ofthe videos in the two high-load soundtrack conditions. Afterviewing the video, participants rated their WTP for theearbuds using a drop-down list with price options increasingin $25 increments from $50 to $255.

Pretest. A total of 271 participants recruited from MTurk(135 women, Mage= 41.55 years) participated in the pretestwhich featured a 2 (slow motion vs. natural speed) × 2(low-load soundtrack vs. high-load soundtrack) between-subjects design. After viewing the video, participants ratedthe extent to which they found the background sounds to be dis-tracting on a nine-point scale (e.g., “To what extent do youthink the background sound used in the ads is distracting?” 1= “not at all,” and 9= “very”). An ANOVA test confirmed

that the soundtrack in the high-load condition (vs. low-loadcondition) was indeed perceived as more distracting (Mlow-load

= 2.39, SD= 1.88 vs. Mhigh-load= 3.08, SD= 1.98; F(1, 267)= 8.42, p= .004, η2p= .03). Perceived distraction did not differbetween the two video speed conditions (F(1, 267)= .12, NS;interaction effect: F(1, 267)= .10, NS).

Results. A 2 (video condition: slow motion vs. natural speed) ×2 (cognitive load: low-load soundtrack vs. high-load sound-track) ANOVA with WTP as the dependent variable revealeda significant main effect of cognitive load (F(1, 497)= 12.64,p < .001, η2p= .02), which was qualified by a significant interac-tion effect (F(1, 497)= 4.54, p= .034, η2p= .01; Figure 6).When the ad had a low cognitive load soundtrack, we observedthe same effect as previous studies: The slowed commercial (vs.natural speed commercial) yielded lower product WTP ratings(Mnatural_speed= 87.50, SD= 41.87 vs. Mslow_motion= 75.20,SD= 35.28; t(497)= 2.38, p= .018, d= .32). When the adshad a cognitive load soundtrack, however, there was no signifi-cant difference in WTP ratings between the two conditions(Mnatural_speed= 92.72, SD= 41.56 vs. Mslow_motion= 95.97,SD= 44.22; t(497)=−.63, NS).

Discussion. Study 4a provides an additional theory test andshows that the effect weakens when consumers lack the cogni-tive resources to think carefully about influence agents’ inten-tionality. This study also has direct marketing implications:Under conditions of cognitive busyness, which is commonplacein real-world physical environments and online social environ-ments such as live streaming (where viewers are constantlytyping comments that other viewers can see), the negativeeffect of slow motion video can be weakened by the distractor.Indeed, taken one step further, the moderation by cognitive

Figure 5. WTP as a function of video speed and intentionality bias

(Study 3).Note: Participants reported significantly lower WTP toward the advertised

product in the slow motion condition when their intentionality bias scores

were higher than .25.

Figure 6. WTP as a function of video speed and background

soundtrack (Study 4a).

*p< .05.Notes: Error bar indicates ±1 SE.

Yin et al. 11

load suggests that marketers can actively mitigate the negativeimpact of slow motion video by inducing cognitive load.Examples of unobtrusive forms of cognitive load in video adsinclude playing background music, additional audio information(from a narrator or commentator), or even displaying bulletscreen comments (a video–comment integration techniquepopular on live streaming platforms, especially in East Asia).

Study 4b: Manipulating Cognitive Resources Usingan Alternative ApproachOne may wonder whether it was induced emotion or other idio-syncratic psychological or physiological factors triggered by theaudio content, rather than cognitive load per se, that mitigated theeffect (Allan 2006; Gorn 1982; Kellaris, Cox, and Cox1993). Toaddress this concern, we conducted an additional study in whichwe manipulated cognitive load by asking participants to memo-rize an alphanumeric string. Although lower in ecological valid-ity, this manipulation does not create the aforementionedspillover effects and is well established in marketing research(Cian, Longoni, and Krishna 2020; Kwan, Dai, and Wyer2017; Trendel, Mazodier, and Vohs 2018). We obtained analo-gous results using this alternative manipulation of cognitiveload. In the low-load condition, the slow motion condition (vs.natural speed) yielded lower consumer interest (t(503)= 2.77,p= .006, d= .35), while the effect was weakened in the high-loadcondition (t(503)=−.26, NS). Overall, this supports our hypoth-esis that constraining consumers’ cognitive resources can preventthem from perceiving intentionality and subsequently mitigatethe effect of slow motion. The details of Study 4b (method,results, and discussions) are presented in Web Appendix I.

Study 5Study 5 serves two purposes. First, we provide additional processinsight by exploring whether the effect diminishes when the aduses nonhuman agents (e.g., an animated object, animal, brandmascot, which are commonplace in such videos). We postulatethat the perceived intentionality mechanism does not apply fornonhuman agents because social heuristics relating totheory-of-mind are no longer relevant, particularly when nonhu-man agents have no anthropomorphized features, such as a faceor limbs (Epley,Waytz, and Cacioppo 2007), as is the case in thisstudy. Second, in previous experiments, the commercials in theslow motion condition had a longer temporal duration (andthus potentially more visual information), which also gave con-sumers more time to analyze and react to the advertising informa-tion in the advertisement. The current study rules out thispossibility by holding viewing time and amount of informationconstant across conditions.

MethodProcedure. A total of 506 United States residents (274 women,Mage= 41.09 years) recruited from MTurk participated in the

experiment in exchange for monetary compensation. Thestudy adopted a 2 (video condition: slow motion vs. naturalspeed) × 2 (influence agent: animated agent vs. human agent)between-subjects design. Participants were randomly assignedto one of four conditions and viewed one food commercial inwhich the product qualities were demonstrated by either an ani-mated character or a human influence agent (Web Appendix A).The videos were both food commercials, either for Oreo choc-olate cookies or Sanyo instant noodles (the same ad fromStudies 2a–b). The cookie commercial was an ad for Cadburychocolate-flavored Oreos, in which an animated Oreo cookieand Cadbury chocolate hop around and travel together. Bothentities resembled the physical product and were not anthropo-morphized with a face or limbs, etc. To equalize deliberationtime and amount of information across video conditions, partic-ipants in the natural speed condition were asked to watch the advideo twice (as the slow motion ad played at half the speed ofthe ad in the natural speed condition). After watching the com-mercial, participants rated their interest in trying the advertisedproduct on a nine-point scale (“How interested are you in tryingthe product you just saw in the video, if it is available locally?”1= “not at all,” and 9= “very”).

ResultsIn the human influence agent conditions, we observed the sameeffect as in previous studies: Slow motion (vs. natural speed)decreased consumers’ interest in trying the advertised product(Mnatural_speed= 5.78, SD= 2.41 vs. Mslow_motion= 5.07, SD=2.58; t(502)=−2.32, p= .021, d= .28). However, slowmotion actually yielded a positive effect on product evaluation(vs. natural speed) when the ad used nonhuman animatedfigures as influence agents (Mnatural_speed= 6.33, SD= 2.57 vs.Mslow_motion= 6.97, SD= 2.10; t(502)= 2.12, p= .035,d= .27). Although we caution that an interaction analysismay not be a fair comparison (since the subject of the adswere different), a 2 (video condition: slow motion vs. naturalspeed) × 2 (influence agent: human vs. animated figure)ANOVA with consumer interest toward the advertisedproduct as the dependent variable nonetheless yielded a signifi-cant main effect of influence agent (F(1, 502)= 31.50, p < .001,η2p= .06), which was qualified by a significant interaction effect(F(1, 502)= 9.83, p= .002, η2p= .02).

DiscussionThe replication of the effect in the human influence agent con-ditions helps rule out longer viewing time, which was equalacross conditions, as an alternative explanation of the effect.Study 5 also provides a theory test for the proposed processand shows that the effect reverses for an ad with no humanagents—a context in which perceiving an influence agent’sintentionality is not relevant. Of note, slow motion cinematog-raphy actually leads to greater interest in the product for thenonhuman agent ad. In this context, the positive impact ofslow motion might be due to increased attention elicited by

12 Journal of Marketing Research

the slow motion cinematographic effect or aesthetic apprecia-tion for the advertised product portrayed in slow motion. Ofmanagerial relevance, this suggests that the effect can bereversed and that slow motion may abet the persuasivenessof ads when advertisers avoid using human influence agents.

This study is a first step in showing how the humanness ofthe influence agent may moderate the detrimental effect ofslow motion. However, more research is needed to explorethe precise boundaries and moderating role of anthropomor-phism, as these questions are beyond the scope of this research.Although both ads in Study 5 had the same product category(food commercials), the ads differed in many aspects (e.g.,precise consumption behavior, video length), which makes itunfair to interpret the interaction effect. Thus, to address thisissue and rigorously test how the humanness of the influenceagent may moderate the slow motion effect, future studiescould use computer generated imagery to manipulate theextent to which a nonhuman agent is humanized whileholding other factors (e.g., video property, agent actions)constant.

In addition, the nonhuman characters in Study 5 had minimalanthropomorphic features and were simply inanimate objects(i.e., a cookie and a piece of chocolate) engaging in humanlikebehaviors (e.g., going to the beach, taking selfies). It remains anopen question whether slow motion cinematography wouldyield a similar effect if a more humanized character wereused: Do people make social inferences and gauge intentional-ity for lifelike animations of humans, highly anthropomor-phized nonhuman characters (e.g., Mickey Mouse), orquasi-anthropomorphized characters (e.g., a cookie with asmiley face)? These boundary conditions remain questions forfuture research but are of managerial importance given thewidespread use of nonhuman or animated brand mascots, par-ticularly for grocery products, children’s products, and digitalproducts such as video games (Kim, Chen, and Zhang 2016).

General DiscussionThe common use of slow motion effects in video advertisingsuggests that marketers generally assume this tactic nets posi-tive benefits such as greater visual attention or stronger aes-thetic appreciation for the advertised product. However, wefind that using slow motion to depict consumption can backfireand make video marketing less persuasive and advertisedproducts less attractive. The effect is robust across a varietyof real commercials (including live streaming videos), for awide range of products of different countries of origin, andamong participants from student to general population poolsin both East Asia and North America. Study 1 shows thatthe effect persists for online consumers’ responses to realadvertising. Studies 2a–b demonstrate that the negativeimpact of slow motion video is driven by consumers’ attemptsto infer influence agents’ underlying intentionality, which isalso reflected by greater visual attention to influence agents’eyes (Study 2a). In further support, the pattern of results inthe moderation conditions is consistent with previous research

on how people infer intentionality, namely that the backlashcaused by slow motion attenuates for individuals with lowerintentionality bias (Study 3), is mitigated under cognitiveload (Study 4), and reverses when ads use nonhuman influenceagents (Study 5). We also show that the effect cannot beexplained by changes in ease of visual processing, perceivedengagement, and deliberation time (Studies 2b and 5).Finally, we conducted a within-paper meta-analysis on theeffects of experiments reported in current research (Grewal,Puccinelli, and Monroe 2018), which shows that our effectsizes across studies are both homogeneous (Q(6)= .66,p= .995) and significant (η= .32, p< .001), demonstratingthe robustness of our findings (Web Appendix J).

Theoretical ContributionAlthough the slow motion effect is visual in nature, our concep-tual contribution is primarily to the literature on visual-basedsocial inference. We reveal a novel antecedent (i.e., slowmotion) that may trigger the perception of intentionality in con-sumer inferences, and we identify a process by which slowmotion affects consumer preference. Previous literature hasshown that slow motion may highlight intentionality underlyinghuman actions in very specific decision-making domains, par-ticularly in legal adjudication contexts that involve legal ormoral infractions (e.g., referee’s judgment of violent fouls,jury’s perceptions of criminal behavior [Caruso, Burns, andConverse 2016; Spitz et al. 2018]). Our research makes a firststep in showing that a similar slow motion intentionality biascan extend to more general social contexts and visual marketinginvolving influence agents in particular. In these contexts, con-sumers may use their internal speed comparison mechanisms tocompare “ideal” action against observed action to conduct anintentionality test. Contrary to the intuition that slowed advideo should present more visual details and thus lead togreater trustworthiness (DePaulo et al. 2003; Hartwig andBond 2011), our research shows that slow motion effects canbackfire. We document the specific conditions under whichintentionality inferences can override any aesthetic and visualbenefits to the detriment of the advertisement’s appeal and per-suasiveness. In doing so, we also explore how (dynamic) visualmarketing design aesthetics can have a significant and unin-tended impact on consumers’ social cognition, which directlyaffects marketing efficacy.

Our work also sheds light on visual attentional processes(and eye movements) that occur when consumers make social-cognitive inferences about influence agents (and brand repre-sentatives). Indeed, our eye tracking study showed that con-sumers’ gaze patterns and visual attentional strategieschanged in response to the relatively subtle cue of slowmotion behavior. We make a methodological contribution bydemonstrating that visual attention, eye movement, andactive tracking of influence agents’ eyes can serve as proxymeasures of consumer attributional inference and suspicions(of intentionality) in marketing contexts. We directly linksuch visual behavioral patterns to perceptions of agents’

Yin et al. 13

intentionality and changes in consumers’ attitude toward theadvertised product. As marketing becomes increasingly “indi-rect” and subtle, particularly in new media contexts, measure-ments of when consumers “actively monitor” for marketingand persuasion attempts are important marketing efficacy mea-sures in their own right.

This work also adds to the growing research on how timeperception affects consumer judgment (Rudd, Catapano, andAaker 2018; Tonietto, Malkoc, and Nowlis 2019). We studyhow a common manipulation of time perception in mediaand advertising, namely slow motion, affects social inferencesand ultimately changes consumer attitudes and preferences.We concurrently contribute to an emerging stream of visualmarketing research that examines how aesthetic strategiesinfluence the evaluation and perception of products (Argoand Dahl 2017; Buechel and Townsend 2018; Cian, Krishna,and Elder 2014; Hagtvedt and Patrick 2008). Whereas priorresearch has focused on static visual cues, we investigatehow a unique, dynamic visual cue affects how consumers con-strue influence agents in marketing. Finally, we propose onespecific trait—intentionality bias—as a critical social cognitionqualification that modulates the efficacy of how visual-basedpersuasive messages are delivered. This trait can be readilyextended to the many marketing contexts in which social cog-nition is required to process persuasive information comingfrom a marketing or influence agent (e.g., face-to-face sales,online social media platforms, live streaming, as shown inStudy 3).

We also contribute to literature on consumer response to per-suasion and advertising by considering how visual effects andpersuasion intersect in new media contexts (e.g., live streaming,reality TV shows, social influencers) in which advertising is more“hidden” and less explicit than in legacy marketing mediums.Such contexts do not fit precisely into the traditionalPersuasion Knowledge Model, which was developed for mid-to late-twentieth-century media and communication contextsand posits that consumers need recognition and awareness ofadvertising to cope and defend against persuasive messages(Evans and Park 2015; Friestad and Wright 1994; Obermillerand Spangenberg 1998). Our findings suggest that the extent towhich consumers are able to resist persuasion in new media con-texts, in which there is less scripted behavior by social agents andno blatant signs of advertising (e.g., the live streaming context inStudy 3), may depend on their ability to muster social cognitionheuristics. Once active, these heuristics can devalue the credibil-ity of influence agents’ behaviors. Of course, these findings arestill applicable in traditional advertising contexts, which stillinclude overt advertising signals (Friestad and Wright 1994;Obermiller and Spangenberg 1998). In these contexts, heightenedperceived intentionality may directly trigger persuasion knowl-edge and skepticism toward advertisers.

Managerial ImplicationsOur research has numerous applications for visual marketing,particularly for video, TV, mobile advertising, and emerging

social media contexts, in which behavioral social proof frominfluence agents (e.g., testing or reacting to a product) is acommon persuasion strategy. Our findings highlight the needfor more programmatic studies of how visual marketing strat-egies and social cognition interact. This question is particularlyrelevant for visual portrayals of active consumption, in whichvisual and cinematographic effects can potentially trigger aminefield of visual and cognitive heuristics. These questionshave received little attention in marketing research despitetheir wide-ranging implications for both theory and practice.However, it is equally likely that such heuristics can be easilyde-biased. For example, our findings suggest that the slowmotion effect can be de-biased by cognitive load (e.g., back-ground sound) or if the consumption decision seems more inci-dental rather than intentional. Moreover, results of our eyetracking study also imply that marketers can potentiallydodge the slow motion inference bias by using tools to nudgeconsumers’ visual attention away from the influence agent(e.g., captions, narrative preceding the ads, visual markers) toprevent the establishment of action intentionality.

In addition, our research provides a starting point for under-standing how visual marketing affects consumer trust. A trust-worthiness alarm has been raised in advertising recently,especially for digital advertising: According to Nielsen’sreport on global trust in advertising (Nielsen 2015), less thanhalf (48%) of respondents reported they entirely or partiallytrust online video ads, and mobile ads had an even lower trust-worthiness rating (43%). Our studies suggest that marketersshould be even more mindful of how the delivery and executionof advertisements (including the interaction effect betweenvideo/visual aesthetics and choice of influence agent) affect per-ceived trustworthiness, particularly as more and more advertis-ing spending migrates toward emerging new media platformssuch as (mobile) short-form videos, live streaming, andvirtual reality. Our conceptualization suggests that resistanceto overt intentionality can be lowered when ads (1) are endorsedby real consumers rather than professional actors (or celebri-ties), (2) look more like real life (e.g., consumption in naturalspeed), and (3) present natural contexts rather than exaggeratedcaricatures (e.g., prolonged savoring in slow motion).

Limitations and Directions for Future ResearchThis research is intended to be a first step in exploring thegeneral question of how social cognition effects mediate theefficacy of visual marketing strategies. In investigating howone particular (but common) visual marketing technique gener-ates unintended social cognition effects, we have by necessityleft many questions unanswered and have spurred even morenew questions. We next discuss questions that are beyond thescope of this research but may nonetheless be important limita-tions and potential moderators for our findings.

While our research focuses on a particular cinematographiceffect (i.e., slow motion) and its undesirable marketing conse-quences, one may wonder whether analogous effects occurfor other cinematographic effects (e.g., speeding up, color

14 Journal of Marketing Research

filtering) or whether consumers have a negative reaction tovideo editing in general. Initial evidence reported in WebAppendix K suggests this is not the case, and it shows thatthe negative main effect and perceived intentionality-drivenmechanism does not occur for two other major cinematographictechniques (e.g., fast speed, color filtering in black-and-white).Furthermore, we find no evidence that the process underlyingthe slow motion effect is due to some other visual aestheticcharacteristics (e.g., perceived information quantity andboredom). This suggests that the negative impact of slowmotion cinematography is not merely due to a general prefer-ence for “natural” speed. Of course, our tests were not exhaus-tive, and we hope that future research can systematically test themarketing efficacy of all known cinematographic techniques,which are usually evaluated from an aesthetic or psychologicalrather than marketing perspective (Argo and Dahl 2017;Buechel and Townsend 2018; Cian, Krishna, and Elder 2014;Hagtvedt and Patrick 2008). Doing so will deepen our under-standing of the mechanism behind consumer visual processingof dynamic visual stimuli and also have direct managerialimplications for how commercials are designed, calibrated,edited, and played.

Although we tested the effect using numerous real ads thatvaried in product category and cinematographic features(length, sound, and original speed), and we ruled out severalobvious alternative explanations (e.g., ease of visual process-ing, boredom, engagement), the slow motion effect, as withmost marketing phenomenon, is likely to be multiply deter-mined. Consequently, these and other factors may co-occurand help drive the effect in the real world. For example, affec-tive factors may override (or at least interplay with) theintentionality-driven mechanism when videos are immersed instrong emotions (e.g., charity ads using slow motion toamplify a touching story). Indeed, because we found thatinformation-intensive background sounds can mitigate thebacklash caused by slow motion (Study 4), one may wonderwhether sound effects can interact with slow motion effects inother ways. For example, prior studies find that backgroundmusic can shape ad processing through attention gain andmusic–message congruency (Allan 2006; Gorn 1982;Kellaris, Cox, and Cox 1993). Thus, future research can manip-ulate nuances of background sound, which may interact withslow motion video and product category to determine an ad’sefficacy. Such research can help draw a more complete andmultimodal understanding of how consumers respond tovisual cues in the real (multimodal) world.

For consistency, we used real commercials that showedinfluence agents reacting positively to consumption (e.g.,smiling while chewing on food) to serve as social proof forthe advertised product. Although most real-world slowmotion commercials fall within this category, there are alsoproduct categories in which social proof plays a smaller role.For instance, advertisements for products that require closescrutiny (e.g., medications, tools) typically emphasize productperformance qualities rather than qualities that rely on socialproof. Future research may investigate the impact of slow

motion effects in contexts in which social proof is less relevantor informative. Related to this question, one might also wonderif our effect extends to contexts in which there are no influenceagents at all, for example, a car commercial depicting a car trav-elling in slow motion. Indeed, we found that slow motion can bebeneficial to ad persuasiveness when influence agents are ani-mated, nonhuman objects (Study 5). When no influenceagents are present, we predict that slow motion should onlyhave visual aesthetic effects and no social cognition effects.However, the characteristics of such effects are beyond thescope of our research and conceptual framework, which focuson contexts in which influence agents are present and socialcognition effects are possible.

Finally, future investigations may consider the impact ofslow motion on downstream effects such as memory. In thisresearch, we studied how slow motion video affects persuasionby measuring consumer evaluations immediately after commer-cial viewing. However, could slow motion videos affect, andpossibly improve, how well consumers recall the informationpresented in the commercial? These are but a few examplesof the multitudinous directions that future research could taketo build on our opening endeavors. These questions willbecome increasingly important as marketing and daily con-sumer behavior migrate into new media domains, particularlylive video–enabled social platforms and AI-assisted visualsearch platforms such as TikTok. More generally, we hopethat a stream of future research can help create a comprehensiveunderstanding of how visual presentation and social cognitioneffects interact to influence perceptions, persuasion, and trustacross a variety of marketing contexts.

AcknowledgmentsThe authors would like to thank the JMR editorial team, Echo WenWan, and Sara Kim for their feedback at various stages of this research.This research is part of the first author’s doctoral dissertation. Pleaseaddress all correspondence to Yunlu Yin (yinyunlu@fudan.edu.cn).

Associate EditorDhruv Grewal

Declaration of Conflicting InterestsThe author(s) declared no potential conflicts of interest with respect tothe research, authorship, and/or publication of this article.

FundingThe author(s) disclosed receipt of the following financial support forthe research, authorship, and/or publication of this article: Thisresearch was supported by research grants from the Hong KongResearch Grants Council (14505217 and 17506316) awarded to thesecond author.

ORCID iDYunlu Yin https://orcid.org/0000-0002-6433-3870

Yin et al. 15

References

Allan, David (2006), “Effects of Popular Music in Advertising onAttention and Memory,” Journal of Advertising Research, 46 (4),434–44.

Areni, Charles S. and Iain Black (2015), “Consumers’ Responses toSmall Portions: Signaling Increases Savoring and Satiation,”Psychology & Marketing, 32 (5), 532–43.

Argo, Jennifer J. and Darren W. Dahl (2017), “Standards of Beauty:The Impact of Mannequins in the Retail Context,” Journal ofConsumer Research, 44 (5), 974–90.

Arstila, Valtteri (2012), “Time Slows Down During Accidents,”Frontiers in Psychology, 3, 196.

Auyeung, Bonnie, Michael V. Lombardo, Markus Heinrichs,Bhismadev Chakrabarti, Akeem Sule, Julia B. Deakin, et al.(2015), “Oxytocin Increases Eye Contact During a Real-Time,Naturalistic Social Interaction in Males with and WithoutAutism,” Translational Psychiatry, 5 (2), e507.

Baldwin, Dare A. and Jodie A. Baird (2001), “Discerning Intentions inDynamic Human Action,” Trends in Cognitive Sciences, 5 (4),171–78.

Barnett, Brooke and Maria Elizabeth Grabe (2000), “The Impact ofSlow Motion Video on Viewer Evaluations of Television NewsStories,” Visual Communication Quarterly, 7 (3), 4–7.

Baron-Cohen, Simon, Ruth Campbell, Annette Karmiloff-Smith,Julia Grant, and Jane Walker (1995), “Are Children with AutismBlind to the Mentalistic Significance of the Eyes?” BritishJournal of Developmental Psychology, 13 (4), 379–98.

Bates, Douglas, Martin Maechler, Ben Bolker, and Steven Walker(2014), “lme4: Linear Mixed-Effects Models Using Eigen andS4,” R Package Version, 1 (7), 1–23.

Birmingham, Elina, Walter F. Bischof, and Alan Kingstone (2008),“Social Attention and Real-World Scenes: The Roles of Action,Competition, and Social Content,” Quarterly Journal ofExperimental Psychology, 61 (7), 986–98.

Biteable (2021), “55 Video Marketing Statistics for 2021,” (accessedJuly 15, 2021), https://biteable.com/blog/video-marketing-statistics/.

Blakemore, Sarah-Jayne and Jean Decety (2001), “From thePerception of Action to the Understanding of Intention,” NatureReviews Neuroscience, 2 (8), 561–67.

Bowman, Matt (2017), “Video Marketing: The Future of ContentMarketing,” Forbes (February 3), https://www.forbes.com/sites/forbesagencycouncil/2017/02/03/video-marketing-the-future-of-content-marketing/.

Bryant, Fred B. and Joseph Veroff (2007), Savoring: A New Model ofPositive Experience. Mahwah, NJ: Lawrence Erlbaum AssociatesPublishers.

Buechel, Eva C. and Claudia Townsend (2018), “Buying Beauty forthe Long Run: (Mis)predicting Liking of Product Aesthetics,”Journal of Consumer Research, 45 (2), 275–97.

Campbell, Margaret C. and Amna Kirmani (2000), “Consumers’ Useof Persuasion Knowledge: The Effects of Accessibility andCognitive Capacity on Perceptions of an Influence Agent,”Journal of Consumer Research, 27 (1), 69–83.

Caruso, Eugene M., Zachary C. Burns, and Benjamin A. Converse(2016), “Slow Motion Increases Perceived Intent,” Proceedings

of the National Academy of Sciences of the United States ofAmerica, 113 (33), 9250–55.

Cialdini, Robert B. (2007), Influence: The Psychology of Persuasion.New York: Collins New York.

Cian, Luca, Aradhna Krishna, and Ryan S. Elder (2014), “This LogoMoves Me: Dynamic Imagery from Static Images,” Journal ofMarketing Research, 51 (2), 184–97.

Cian, Luca, Chiara Longoni, and Aradhna Krishna (2020),“Advertising a Desired Change: When Process SimulationFosters (vs. Hinders) Credibility and Persuasion,” Journal ofMarketing Research, 57 (3), 489–508.

Clement (2019), “U.S. User Reach of Leading Video Platforms 2018,”https://www.statista.com/statistics/266201/us-market-share-of-leading-internet-video-portals/.

Colombatto, Clara, Yi-Chia Chen, and Brian J. Scholl (2020), “GazeDeflection Reveals How Gaze Cueing is Tuned to Extract theMind Behind the Eyes,” Proceedings of the National Academy ofSciences, 117 (33), 19825–29.

DePaulo, Bella, James J. Lindsay, Brian Malone, Laura Muhlenbruck,Kelly Charlton, and Harris Cooper (2003), “Cues to Deception,”Psychological Bulletin, 129 (1), 74–118.

Emery, Nathan J. (2000), “The Eyes Have It: The Neuroethology,Function, and Evolution Of Social Gaze,” Neuroscience &Biobehavioral Reviews, 24 (6), 581–604.

Epley, Nicholas, AdamWaytz, and John Cacioppo (2007), “On SeeingHuman: A Three-Factor Theory of Anthropomorphism,”Psychological Review, 114 (4), 864–86.

Evans, Nathaniel and Dooyeon Park (2015), “Rethinking thePersuasion Knowledge Model: Schematic Antecedents andAssociative Outcomes of Persuasion Knowledge Activation forCovert Advertising,” Journal of Current Issues & Research inAdvertising, 36 (2), 157–76.

Fisher, Matthew, George E. Newman, and Ravi Dhar (2018),“Seeing Stars: How the Binary Bias Distorts the Interpretationof Customer Ratings,” Journal of Consumer Research, 45 (3),471–89.

Friestad, Marian and Peter Wright (1994), “The Persuasion KnowledgeModel: How People Cope with Persuasion Attempts,” Journal ofConsumer Research, 21 (1), 1–31.

Gilbert, Daniel T. and Patrick S. Malone (1995), “The CorrespondenceBias,” Psychological Bulletin, 117 (1), 21–38.

Gorn, Gerald J. (1982), “The Effects of Music in Advertising onChoice Behavior: A Classical Conditioning Approach,” Journalof Marketing, 46 (1), 94–101.

Grewal, Dhruv, Nancy Puccinelli, and Kent B. Monroe (2018),“Meta-Analysis: Integrating Accumulated Knowledge,” Journalof the Academy of Marketing Science, 46 (1), 9–30.

Grossmann, Tobias (2017), “The Eyes as Windows into Other Minds:An Integrative Perspective,” Perspectives on PsychologicalScience, 12 (1), 107–21.

Guillon, Quentin, Nouchine Hadjikhani, Sophie Baduel, andBernadette Rogé (2014), “Visual Social Attention in AutismSpectrum Disorder: Insights from Eye Tracking Studies,”Neuroscience & Biobehavioral Reviews, 42, 279–97.

Hagtvedt, Henrik and Vanessa M. Patrick (2008), “Art Infusion: TheInfluence of Visual Art on the Perception and Evaluation of

16 Journal of Marketing Research

Consumer Products,” Journal of Marketing Research, 45 (3),379–89.

Hamilton, Antonia F. de C. (2016), “Gazing at Me: The Importance ofSocial Meaning in Understanding Direct-Gaze Cues,”Philosophical Transactions of the Royal Society of London.Series B, Biological Sciences, 371 (1686), 20150080.

Hammerschmidt, David and Clemens Wöllner (2018), “The Impact ofMusic and Stretched Time on Pupillary Responses and EyeMovements in Slow-Motion Film Scenes,” Journal of EyeMovement Research, 11 (2).

Hartwig, Maria and Charles Bond (2011), “Why Do Lie-Catchers Fail?A Lens Model Meta-Analysis of Human Lie Judgments,”Psychological Bulletin, 137 (4), 643–59.

Harvey, John H. and Gifford Weary (1984), “Current Issues inAttribution Theory and Research,” Annual Review of Psychology,35 (1), 427–59.

Hayes, Andrew F. (2013), Introduction to Mediation, Moderation, andConditional Process Analysis: A Regression-Based Approach.New York: The Guilford Press.

Heider, Fritz (1958), The Psychology of Interpersonal Relations.Hoboken, NJ: Wiley.

Horvath, Janos and Annamaria Burgyan (2011), “Distraction and theAuditory Attentional Blink,” Attention Perception & Psychophysics,73 (3), 695–701.

Irvine, Mark (2019), “Facebook Ad Benchmarks for Your Industry,”Wordstream (accessed July 19, 2021), https://www.wordstream.com/blog/ws/2017/02/28/facebook-advertising-benchmarks.

Johnson, Palmer O. and Leo C. Fay (1950), “The Johnson–NeymanTechnique, its Theory and Application,” Psychometrika, 15 (4),349–67.

Kellaris, James J., Anthony D. Cox, and Dena Cox (1993), “The Effectof Background Music on Ad Processing: A ContingencyExplanation,” Journal of Marketing, 57 (4), 114–25.

Kelley, Harold H. and John L. Michela (1980), “Attribution Theoryand Research,” Annual Review of Psychology, 31 (1), 457–501.

Khalid, Saara, Jason C. Deska, and Kurt Hugenberg (2016), “The EyesAre the Windows to the Mind: Direct Eye Gaze Triggers theAscription of Others’ Minds,” Personality and Social PsychologyBulletin, 42 (12), 1666–77.

Kim, Sara, Rocky Peng Chen, and Ke Zhang (2016),“Anthropomorphized Helpers Undermine Autonomy and Enjoymentin Computer Games,” Journal of Consumer Research, 43 (2),282–302.

Kuhn, Gustav, Benjamin W. Tatler, and Geoff G. Cole (2009), “YouLook Where I Look! Effect of Gaze Cues on Overt and CovertAttention in Misdirection,” Visual Cognition, 17 (6/7), 925–44.

Kuznetsova, Alexandra, Per B. Brockhoff, and Rune H. B. Christensen(2017), “lmerTest Package: Tests in Linear Mixed Effects Models,”Journal of Statistical Software, 82 (13), 1–26.

Kwan, Canice M. C., Xianchi Dai, and Robert S. Wyer Jr. (2017),“Contextual Influences on Message Persuasion: The Effect ofEmpty Space,” Journal of Consumer Research, 44 (2), 448–64.

MacInnis, Deborah J., Ambar G. Rao, and Allen M. Weiss (2002),“Assessing When Increased Media Weight of Real-WorldAdvertisements Helps Sales,” Journal of Marketing Research, 39(4), 391–407.

Mercer Moss, Felix, Roland Baddeley, and Nishan Canagarajah(2012), “Eye Movements to Natural Images as a Function of Sexand Personality,” PloS One, 7 (11), e47870.

Newman, George E., Margarita Gorlin, and Ravi Dhar (2014), “WhenGoing Green Backfires: How Firm Intentions Shape the Evaluationof Socially Beneficial Product Enhancements,” Journal ofConsumer Research, 41 (3), 823–39.

Nielsen (2015), “Global Trust in Advertising,” (September 28), https://www.nielsen.com/wp-content/uploads/sites/3/2019/04/global-trust-in-advertising-report-sept-2015-1.pdf.

Obermiller, Carl and Eric R. Spangenberg (1998), “Development of aScale to Measure Consumer Skepticism Toward Advertising,”Journal of Consumer Psychology, 7 (2), 159–86.

Petty, Richard E. and John T. Cacioppo (1986), “The ElaborationLikelihood Model of Persuasion,” in Communication andPersuasion: Central and Peripheral Routes to Attitude Change,Richard E. Petty and John T. Cacioppo, eds. New York: SpringerNew York.

Rees, Geraint, Chris Frith, and Nilli Lavie (2001), “Processing ofIrrelevant Visual Motion During Performance of an AuditoryAttention Task,” Neuropsychologia, 39 (9), 937–49.

Reich, Taly, Daniella M. Kupor, and Rosanna K. Smith (2017), “Madeby Mistake: When Mistakes Increase Product Preference,” Journalof Consumer Research, 44 (5), 1085–103.

Rosset, Evelyn (2008), “It’s No Accident: Our Bias for IntentionalExplanations,” Cognition, 108 (3), 771–80.

Rudd, Melanie, Rhia Catapano, and Jennifer Aaker (2018), “MakingTime Matter: A Review of Research on Time and Meaning,”Journal of Consumer Psychology, 29 (4), 680–702.

Saxe, Rebecca and Nancy Kanwisher (2003). “People Thinking AboutThinking People: The Role of the Temporo-Parietal Junction in‘Theory of Mind,’” Neuroimage, 19 (4), 1835–42.

Senju, Atsushi and Mark H. Johnson (2009), “The Eye Contact Effect:Mechanisms and Development,” Trends in Cognitive Sciences, 13(3), 127–34.

Settle, Robert B. and Linda L. Golden (1974), “Attribution Theory andAdvertiser Credibility,” Journal of Marketing Research, 11 (2),181–85.

Singh, Deepak (2019), “10 Best Phones to Capture Slow MotionVideos in 2019,” Smartprix (accessed July 19, 2021), https://www.smartprix.com/bytes/best-phones-to-capture-slow-motion-videos/.

Slavny, Rachel J. M. and James W. Moore (2018), “IndividualDifferences in the Intentionality Bias and its Association withCognitive Empathy,” Personality and Individual Differences,122, 104–08.

Sparkman, Richard M. Jr. and William B. Locander (1980),“Attribution Theory and Advertising Effectiveness,” Journal ofConsumer Research, 7 (3), 219–24.

Spiller, Stephen A. and Lena Belogolova (2016), “On ConsumerBeliefs About Quality and Taste,” Journal of ConsumerResearch, 43 (6), 970–91.

Spiller, Stephen A., Gavan J. Fitzsimons, John G. Lynch, and GaryH. McClelland (2013), “Spotlights, Floodlights, and the MagicNumber Zero: Simple Effects Tests in Moderated Regression,”Journal of Marketing Research, 50 (2), 277–88.

Yin et al. 17

Spitz, Jochim, Put Koen, Johan Wagemans, Mark A. Williams, andWerner F. Nelsen (2017), “Does Slow Motion Impact on thePerception of Foul Play in Football?” European Journal of SportScience, 17 (6), 748–56.

Spitz, Jochim, Pieter Moors, Johan Wagemans, and Werner F. Nelsen(2018), “The Impact of Video Speed on the Decision-MakingProcess of Sports Officials,” Cognitive Research: Principles andImplications, 3, 1–10.

Tonietto, Gabriela N., Selin A. Malkoc, and Stephen M. Nowlis(2019), “When an Hour Feels Shorter: Future Boundary TasksAlter Consumption by Contracting Time,” Journal of ConsumerResearch, 45 (5), 1085–102.

Trendel, Olivier, Marc Mazodier, and Kathleen D. Vohs (2018),“Making Warnings About Misleading Advertising and ProductRecalls More Effective: An Implicit Attitude Perspective,”Journal of Marketing Research, 55 (2), 265–76.

Vassallo, Suzane, Sian L. Cooper, and Jacinta M. Douglas (2009),“Visual Scanning in the Recognition of Facial Affect: Is There anObserver Sex Difference?” Journal of Vision, 9 (3), 11.

Waytz, Adam, Kurt Gray, Nicholas Epley, and Daniel M. Wegner(2010), “Causes and Consequences of Mind Perception,” Trendsin Cognitive Sciences, 14 (8), 383–88.

Williams, Patti, Gavan J. Fitzsimons, and Lauren G. Block(2004), “When Consumers Do Not Recognize ‘Benign’ Intention

Questions as Persuasion Attempts,” Journal of ConsumerResearch, 31 (3), 540–50.

Wöllner, Clemens, David Hammerschmidt, and Henning Albrecht(2018), “Slow Motion in Films and Video Clips: MusicInfluences Perceived Duration and Emotion, AutonomicPhysiological Activation, and Pupillary Responses,” PLOS ONE,13 (6), e0199161.

Yu, Hongbo, Yunyan Duan, and Xiaolin Zhou (2017), “Guilt in theEyes: Eye Movement and Physiological Evidence forGuilt-Induced Social Avoidance,” Journal of Experimental SocialPsychology, 71, 128–37.

Yuki, Masaki, William W. Maddux, and Takahiko Masuda(2007), “Are the Windows to the Soul the Same in the Eastand West? Cultural Differences in Using the Eyes and Mouthas Cues to Recognize Emotions in Japan and the UnitedStates,” Journal of Experimental Social Psychology, 43 (2),303–11.

Zettl, Herbert (2011), Television Production Handbook. Boston:Cengage Learning.

Zhou, Christina and Bang Xiao (2019), “China’s Digital Gold Rush:Meet the New Breed of Entertainers who Found a Career in LiveStreaming,” ABC News (February 16), https://www.abc.net.au/news/2019-02-17/inside-chinas-lucrative-livestreaming-industry/10810788.

18 Journal of Marketing Research

Copyright of Journal of Marketing Research (JMR) is the property of American MarketingAssociation and its content may not be copied or emailed to multiple sites or posted to alistserv without the copyright holder's express written permission. However, users may print,download, or email articles for individual use.