The downside of surprise products
The effect of uncertainty labels on the subsequent
consumption experience
Lena Wegener
The downside of surprise products
The effect of uncertainty labels on the subsequent
consumption experience
Lena Wegener
S3812855
17
thJune 2019
Faculty of Economics and Business
Master of Science Marketing: Marketing Management Master Thesis
1st supervisor: Anika Schumacher, M.Sc
2nd supervisor: Dr. Jenny van Dorn
Lena Wegener, Höperhöfen 103, 27367 Bötersen, Germany Telephone: 0049 151 12096196
Abstract
In the field of research on perceived uncertainty this paper introduces a new antecedent of uncertainty: labels of uncertainty. We show that a label of uncertainty, like the word “surprise”, can trigger perceived uncertainty and that this has consequences on the labelled consumption experience with the product or service. We aim to investigate if a surprise label decreases consumption enjoyment and if this decrease in enjoyment results from a change in information processing due to the label. Using an online experiment, we assigned the participants either to the surprise label or control label condition and manipulated the processing type in order to evoke either narrative or analytical processing. The participants were shown a TV spot subsequent to the exposure to the label allowing us to measure their enjoyment of it afterwards. The analysis for the experiment revealed that the proposed model could not be proven. However, in the review of the experiment and data collection we identified several issues regarding the reliability of the experiments. Therefore, we suppose that the thoroughly developed theory could be proved under better circumstances.
Preface
Table of Contents
1 Introduction ... 6
2 Theoretical framework ... 8
2.1 Labels of uncertainty ... 8
2.2 Perceived uncertainty and information processing ... 8
2.2.1 Perceived uncertainty as undesirable state captures resources ... 8
2.2.2 Perceived uncertainty increases processing ... 10
2.2.3 Perceived uncertainty shifts attention to details ... 11
2.3 Uncertainty and enjoyment ... 12
2.4 Processing type ... 13
2.4.1 Narrative processing ... 13
2.4.2 Analytical processing ... 14
3 Research design ... 16
3.1 Participants and design ... 16
3.2 Stimuli development ... 16
3.2.1 Label (uncertainty label vs. control label) ... 16
3.2.2 Processing type ... 16
3.2.3 TV spot ... 18
3.2.4 Details in the TV spot ... 19
3.3 Procedure ... 21
4 Results ... 24
4.1 Scale reliability ... 24
4.1.1 Enjoyment scale ... 24
4.1.2 Processing type scales ... 24
4.1.3 Control variable scales ... 26
4.2 Manipulation checks ... 27
4.2.1 Narrative processing ... 27
4.2.2 Analytical processing ... 27
4.3 Testing our focal hypotheses: Uncertainty and enjoyment ... 28
4.3.2 Detail and peripheral cue ... 29
4.3.3 Moderated mediation of uncertainty label on enjoyment ... 31
4.4 Results for control variables ... 33
4.4.1 Involvement with the ad ... 33
4.4.2 Attitude toward the advertised brand ... 33
4.4.3 Gender effect ... 33
5 Conclusion and recommendations ... 35
6 References ... 37
1 Introduction
Subjective feelings of perceived uncertainty can have an impact on human behavior. In the marketing field concepts like attitude certainty as a form of perceived uncertainty have been studied in depth. According to a review on that issue by Tormala and Rucker (2018), there are several antecedents of perceived uncertainty like the assessment of accuracy, completeness, relevance, legitimacy, importance, and subjective experience (Tormala and Rucker 2018). We shed light on a new antecedent of perceived uncertainty: labels of uncertainty. In daily life consumers are faced with those labels in product names, brand slogans or in marketing promotions. To mention just a few examples, like the “Amazon Mystery Box” (Amazon EU S.à r.l. 2019; Amazon EU S.à r.l. 2017), “Lufthansa Surprise Booking” (Deutsche Lufthansa AG 2019) or the known “Sneak Preview” (CinemaxX Entertainment GmbH & Co. KG 2019) in the movie theatre. However, in consumer research the concept of labels of uncertainty has not received any attention yet to our best knowledge.
Consumer researchers usually study how product labels affect product choice (Hagtvedt 2011) or consumption experience (Dohle and Siegrist 2014). This study focuses on the latter aspect. Brand managers can use uncertainty appeals to create a brand name that is distinctive or emotionally interesting. Robertson (1989) described nine general guidelines considered to create a “good” brand name, from whom two are mentioned above. For example, the product label “Lufthansa surprise booking” is distinctive as surprise and booking a flight is usually not considered to be fulfilled by the same product and therefore, it can create feelings of interest. This may persuade the consumer to choose the product. However, the consumption experience might be perceived differently. As implicated in the examples, labels of uncertainty are not fixed in wording and can vary. In this study we focus on surprise as a representative label of uncertainty. It is used in branding in several product categories, like the “Kinder Surprise” (Ferrero 2019) for chocolate from Ferrero, the mentioned “Lufthansa Surprise Booking” (Deutsche Lufthansa AG 2019) from an airline or in the slogan of the car manufacturer “KIA – the power to surprise” (KIA MOTORS CORP. 2015). Because such brand names and labels of products can have a great impact on the brand image (Robertson 1989), it seems rather important to study this in more detail. The main question of this work is therefore to discover, which consequences a label of uncertainty, and the label surprise in particular, has on consumers.
that perceived uncertainty, evoked by the uncertainty label, will reduce enjoyment of the labelled consumption experience due to a psychological process in which the focus of cognitive resources is shifted towards details as consumers anticipate unexpected and thus, uncertain elements during the consumption experience. This prevents the mind from remaining involved in the consumption experience as such and a limited involvement in the consumption experience has been shown to decrease enjoyment (Higgins 2006; Killingsworth and Gilbert 2010; Barasch, Zauberman and Diehl 2018). In short, we expect that a surprise label decreases enjoyment of the subsequent consumption experience and that this decrease occurs due to a greater analytical information processing.
2 Theoretical framework
2.1 Labels of uncertainty
Labels and brand names are from particular interest in the marketing field. Managers often try to find the “correct” name for a product. Robertson (1989) early showed that managers are oversupplied by checklists for good brand names. He concluded that the perceived meaning of the brand name is an overlooked factor. Nowadays there is more research focusing on the perceived meaning of brand names, for example on the perceived difficulty due to a hard-to-read font type (Hagtvedt 2011) or the distinction between parent brand and brand extension (John and Monga 2010). In this paper we examine another type of labels: uncertainty labels.
As already mentioned, there are several antecedents of perceived uncertainty. In the review of Tormala and Rucker (2018) they have recently summarized the findings on these. Usually, the strength of uncertainty is determined in a meta-cognitive judgment process of a given piece of information. Several appraisal dimension are known to influence this process, such as accuracy or completeness (Tormala and Rucker 2018). It may be said here, that this study agrees on uncertainty labels to be an antecedent of perceived uncertainty because the perceived meaning of labels, including words like “surprise” or “mystery”, is obvious.
More interesting are the possible consequences of perceived uncertainty due to an uncertainty label. In the review of Tormala and Rucker (2018) the authors identify three main types of consequences. As one of them, the authors mention the effect of attitude certainty on information processing. Specifically, it can influence the extent to which consumers elaborate on a given piece information. In the following we will develop a conceptual model that shows which consequences perceived uncertainty has on a subsequent consumption experiences and which underlying psychological processes are causing this.
2.2
Perceived uncertainty and information processing
2.2.1 Perceived uncertainty as undesirable state captures resources
Ruan, Hsee and Lu (2018) found in seven studies that a hedonic experience is more enjoyed when it was first teased with some uncertainty, like missing information. Compared to receiving all information immediately, they discovered a net positive experience from first creating and then resolving uncertainty (Ruan, Hsee and Lu 2018). To summarize uncertainty is not a desired state, whereas resolving it can provide benefits in different ways.
A paper by Pratto and John (1991) opens an interesting view on the latter findings. They argued and show that negative events directly attract attentional resources due to a function of automatic stimulus evaluation. As shown above uncertainty can represent an undesirable state and a possible negative outcome. We suggest that there is a link between both concepts and that perceived uncertainty will attract attentional resources due to its perceived negativity. As explained by Pratto and John a process called automatic stimulus evaluation directs attention to simply more urgent events, which are often the negative ones. They lead it back to an evolutionary need to detect danger, risks and threats and to deal with those immediately. Accordingly, this process was required to be quick, effortless and automatic (Pratto and John 1991). We can conclude here that uncertainty as possible negative outcome automatically and unconsciously attracts attentional and cognitive resources.
2.2.2 Perceived uncertainty increases processing
Following the same tone, we not only suggest that perceived uncertainty attracts more cognitive resources, moreover we propose that it also influences how we use these resources. It has been shown that perceived uncertainty increases processing of given information. There are several studies examining this effect under different conditions and with different research designs. They consistently show an increase in information processing due to a need to feel certain again (Tiedens and Linton 2001; Edwards 2003; Tormala, Rucker and Seger 2008; Karmarkar and Tormala 2010; Wan and Rucker 2013; Shen, Fishbach and Hsee 2015). For instance, Wan and Rucker (2013) showed in one of their experiments that a feeling of low confidence (a form of perceived uncertainty) is associated with greater information processing, as they observed an argument quality effect in a persuasive message evaluation.
In short, uncertainty labels evoking perceived uncertainty not only attract cognitive resources and change the allocation of attention to greater exogenous attention, but also increase the processing effort of any given information. This in turn has consequences on the subsequent consumption experience which is labelled. We will show in the following that not only more cognitive resources are tied but that that the focus is directed in a certain way.
2.2.3 Perceived uncertainty shifts attention to details
Building on the latest conclusion, we show in the following that uncertainty labels evoking perceived uncertainty shift the attention and cognitive processing towards more systematic information processing and thus a greater focus on details.
Most of the previous studies in marketing show that perceived uncertainty increases processing efforts, but they also imply what type of processing is operating. Tiedens and Linton (2001) showed that emotions representing certainty lead to a more heuristic processing whereas uncertainty evoked emotions foster a systematic processing. Edwards (2003), Tormala, Rucker and Seger (2008) and Karmarkar and Tormala (2010) all observed an increasing argument quality effect for conditions of perceived uncertainty (causal uncertainty, low confidence, source incongruity). In all of the studies an argument quality effect implied that consumers consider the objective quality of arguments and thus, evaluated the stimulus systematically.
Finally, in the studies of Wan and Rucker (2013) and Shen, Fishbach and Hsee (2015) they related perceived uncertainty to construals. In the former, it was shown that low confidence leads to a focus on information framed in a concrete way, while high confidence groups focused more on messages framed in an abstract manner (Wan and Rucker 2013). In the latter study, uncertainty increased motivation when the focus was on the process, while certainty influenced processing more when the participants focused on the outcome (Shen, Fishbach and Hsee 2015). It can be concluded that perceived uncertainty triggers a systematic way of processing and perceived certainty produces a processing more focused on abstract and heuristic information. As we suggest that a label of uncertainty, like surprise will evoke feelings of perceived uncertainty, we can hypothesise that a label of uncertainty will also increase information processing towards a focus on a more systematic way.
changes, risk or novel elements in the environment seems similar to the focus on details in information and represents a related cognitive processing.
Taking these findings together we can hypothesise that an uncertainty label, like “surprise”, will lead to an increase in processing and a shift in the attention towards the details of a given information or message.
We can find support for this thesis in a study of warning signs. Although it seems rather unrelated, a warning sign is nothing else than a label of uncertainty, as it indicates a possible risk with an unknown probability. In a series of five experiments it was found that warnings signs increase attentional vigilance, leading to a shift in attention to the surrounded environment and influence actual behavior (Cian, Krishna and Elder 2015).
2.3 Uncertainty and enjoyment
One might wonder, how the proposed effect of uncertainty labels can influence the product or service behind the label. Referring to research on the function of television advertisement some insights can be concluded. The excitation transfer theory examined for isntance how humor in commercials influence the entertainment with the following TV program, showing that humor in a TV advertisement increased enjoyment of the running sitcom (Perry 2001). Moreover, in studies on preference fluency, they have shown that subjective feelings during the consumption experience can eventually influence the judgment process and choice (Novemsky, et al. 2007). Based on that we can assume that a label, which induces uncertainty, will influence a subsequent consumption experience.
How will then the perceived uncertainty impact the consumption experience? As proposed above the label of uncertainty will attract cognitive resources and will shift the attention of the processing towards details in the surrounded environment. This implies tough that not all cognitive resources can be spent on the consumption experience and that there may be a stronger focus on details rather than on the gist of the experience.
engagement with the experience indirectly reduces the enjoyment of taking the photos (Barasch, Zauberman and Diehl 2018).
We propose that a similar effect can be observed when an uncertainty label is previously shown to the consumption experience. First, the perceived uncertainty due to the label will cause the mind to attract more resource on solving the perceived uncertainty and will in that way reduce cognitive resources on enjoying the consumption experience. Second, as the attention shifts towards details in a vigilant manner, the mind cannot focus on the experience during consumption rather than “looking out” for something possibly risky, dangerous or novel. Finally, this will lead to a lower engagement with the consumption experience and therefore to a reduce in an enjoyment of it. Based on this we can conclude that an uncertainty label will reduce enjoyment of a consumption experience due to a shift towards details compared to a consumption experience with a neutral label.
2.4 Processing type
In research about media enjoyment the term narrative processing is used as well as its counterpart analytical processing to describe a distinction between two types of processing media (Green, Brock and Kaufman 2004). As proposed above a label of uncertainty may reduce enjoyment of a subsequent consumption experience. Therefore, an important question is whether it will always reduce enjoyment.
2.4.1 Narrative processing
in study of Green, Brock and Kaufmann (2004) they found evidence that narrative processing can affect enjoyment.
The remaining question is still how narrative processing influences the effect of an uncertainty label on enjoyment. Nielsen and Escalas (2010) found in their study that a manipulated narrative processing of information diminished the effect of preference for fluency, meaning that a difficult-to-read font type increased product evaluation and not decreased it. Their explanation was that the difficult-to-read font added to the perceived complexity of the story and therefore, increased the value of it. Based on that it can be assumed that a possible negative effect of the uncertainty label on enjoyment can be reduced as well when a narrative processing is triggered.
A supporting argument can be found in the work by Green and Brock (2000). They examined the role of transportation in more depth finding that narrative processing can lead by transportation to immersion into the story, in which the consumer gets less aware of its environment. This is the opposite effect of what we expect an uncertainty label will evoke. Therefore, we can assume that a narrative processing after the exposure to an uncertainty label will contribute to a less negative or even to a net positive experience during consumption. 2.4.2 Analytical processing
In the research on enjoyment a paper by Raney and Byrant (2002) studied an integrated model of enjoyment for crime drama showing that as well affective but also cognitive factors can affect enjoyment of media. This connects the counterpart of narrative processing, analytical processing, to the subject of enjoyment.
Analytical processing, classified under the rational system by Epstein (1996), represents the opposite of immersion into a story. Escalas (2007) has shown that analytical processing influences judgment through the dual response system. The literature on that states that attitude change is achieved on the peripheral or central route depending on processing resources, meaning that under limited cognitive resources attitude change is based on more affective reasoning and can be persuaded by weak arguments while under cognitive elaboration only strong arguments can persuade (Petty, Cacioppo and Schumann 1983). This for example leads to the known results on preference fluency studies, that under limited cognitive resources easy-to-read fonts are more liked that difficult-easy-to-read because they are easier to process (Reber, Winkielman and Schwartz 1998; Novemsky, et al. 2007; Nielsen and Escalas 2010).
uncertainty attracts additional cognitive resources as a result of the uncertainty label and amount of resources is a critical factor influencing the dual processing. For example, Shiv and Fedorikhin (1999) have shown that under limited resources affective reactions influence choices. This in turn means that under limited cognitive resources the mind is not able to differentiate the cause of the meta-cognitive feeling of uncertainty evoked by the label from meta-cognitive feelings evoked by the labelled consumption experience.
Transferring these insights to the issue of an uncertainty label effect, we can assume that the evoked feeling of uncertainty through the label will negatively affect the enjoyment, as meta cognitive thoughts are crucial under analytical processing. Further, we can suspect that the automatic vigilant behavior of searching for details will be supported by the analytical processing resulting in even less cognitive resource for the involvement in the story and therefore, less enjoyment. Lastly, as a side effect it can be assumed that the supporting “detail focus” can cause the consumer of the story to notice details that would be undetected otherwise – e.g. details that are not central to the ad’s narrative.
In short, we have presented the role of processing type as a moderator of the effect on uncertainty labels on enjoyment resulting in the hypothesis of this paper. In figure 1 and in the hypotheses we display these explained relationships in a conceptual model.
H1a. An uncertainty label reduces, through a process in which the focus shifts towards details, enjoyment of a consumption experience, compared to a neutral label when a manipulated analytical processing enhances the focus on details.
H1b. An uncertainty label will not differ compared to a neutral label in its effect on enjoyment of a consumption experience when a manipulated narrative processing prevents the focus towards details.
3 Research design
3.1 Participants and design
We tested our hypotheses in an online experiment with 185 participants. In total 262 participants started the survey, from whom 70 did not finished and 7 participants reported substantial technical issues, resulting in 185 participants. The participants were recruited in Germany and in the Netherlands, 36% of them were male and 64% female (see Appendix A) for more details on the data cleaning).
Specifically, we assigned the participants either to an uncertainty label or control label condition and then manipulated the processing type. This resulted in a 2 (label: surprise vs. control) x 2 (processing type: analytical vs. narrative) design. We used a video with an edited TV commercial as a subsequent consumption experience to the label and measured the enjoyment of it on a 2-item scale adapted from Redden and Haws (2013). Specifically, we expect to find an interaction between label condition and processing type when analysing the enjoyment of watching the TV spot.
3.2 Stimuli development
3.2.1 Label (uncertainty label vs. control label)
As proposed in chapter 2.1 the wording of uncertainty labels can differ. A variety of words and phrases can evoke feelings of uncertainty. As a representative example the word “surprise” was used in this study. The study started with exposure to either the surprise label or control label (Appendix B) announcing a new (Surprise) TV commercial to be coming out soon. To ensure enough time to process the picture, the respondents could not proceed until 5 seconds have passed. Further, in the following instructions the word “surprise” was used as a salient cue to evoke the hypothesized reactions.
3.2.2 Processing type
Nielsen and Escalas (2010) have shown that analytical and narrative processing can be manipulated using instructions on how to consume the stimuli. In the narrative processing condition, we used their approach on instructing to review the TV commercial as if they should give feedback on the ad concept, while in the analytical processing condition, they were asked to review the concept critically and carefully evaluate the details of the whole movie. In addition, inspired by experiments of Green and Brock (2000) we used their approach of telling the respondents that the shown video is an assignment of students from a movie academy, as it will ensure that the viewer will keep their attention more likely. We combined this with a fictional expert explaining the core elements to consider. In the narrative processing condition, they were told by the fictional expert that an ad concept is the overarching “big” idea of the commercial and that the decision basis should be one’s own feelings. Together the instruction in that condition are designed to enhance transportation into the story and in detail the instruction was phrased like the following:
The commercial was an assignment of students from a movie academy to create a commercial with a famous brand. Your instructions are to review the commercial as if you have been asked by an advertising agency for feedback on an ad concept. According to John Gilovich (a professor at Stanford University), an ad concept is the overarching “big” idea of the commercial. Gilovich states: “It’s best to decide based on one’s general feeling (“Do I like the ad?”) when giving feedback on an ad concept." Please watch the following (Surprise) TV commercial as if you watched something at home on TV.
On the other hand, in the analytical processing condition, they were told from the same expert that there are three core elements (story line, color combination, flow) to consider. Together with the information that the video is an assignment of students, the task was designed to reduce transportation into the story and evoke a systematic evaluation of the video. We phrased it as followed:
John Gilovich (professor at Stanford University) three core elements are always considered when judging commercials:
1. Story line: Is the story easily understandable?
2. Color combination: Is the color combination pleasant and appealing? 3. Movement or flow: Is there a sense of flow throughout the scenes?
Please consider these three criteria when evaluating the (Surprise) TV commercial and watch it carefully.
To ensure that all participants have read the instruction carefully they were asked on the next page to answer a question about the previous read task showing that they have read the task (adapted from Green’s and Brock’s (2000) experiment 2). If they were not able to answer the question correctly, they were proceeded to a reminder page repeating the instructions. 3.2.3 TV spot
We used a video showing an actual TV commercial for several reasons, although in previous experiments (Green and Brock 2000; Escalas 2007; Nielsen and Escalas 2010) textual or static visual material was used. First, according to Green and Brock (2000) a vivid story is necessary in order to ensure that consumers of the story can actually engage with it. They referred to texts mainly, but made it clear that narrative worlds can be consumed by listening, viewing or any other form of consuming. Anyway, they defined some indicators of a good way to design a narrative: “[…] a story that raise unanswered questions […]. A story line, with a beginning, middle, and end” (Green and Brock 2000). The used video meets all these conditions. Second, a video compared to a single print ad (as used in Escalas 2007) or storyboards (as used in Nielsen and Escalas 2010) provides an easier processing of the story.
We considered the limitations of the latter study into account. Nielsen and Escalas (2010) noted that a storyboard was probably too difficult to process, which would prevent immersion into the story. In a storyboard the consumers have to imagine the transition between the different pictures and the action taking place in the pictures. The used video instead provided a complete story to process. Finally, a TV spot provides greater insights into the marketing effectiveness of it. Respectively, print ads (as used in Escalas 2007) are a common practice in marketing and consumers are faced with print ads in their environment on a daily basis. However, videos provide a more suitable basis to allow for transportation into the story.
out in the middle of the night, taking a long trip to a big city. Obviously, he sneaks out on purpose, packing a lunch bag carefully, taking his teddy bear with him and planning the route on a map. He finally arrives at a police station, where he records that he is lost. The mystery is then resolved as the police men indicates that he knows the boy and that the boy has been “lost” before. The goal of the boy was to be transported home in the Mercedes car. Due to a piano music playing in the background and the cute young boy with his teddy bear and the scenes on his way (walking alone over a lonely bridge), the video can evoke strong emotions and allows the viewer to be transported. Further, the video raises questions which are unanswered until the solution, like “Why is the boy sneaking out?” “Where does he go?” or “Will something bad happen to him?”. In short, the video is well suited to allow for transportation and narrative processing.
3.2.4 Details in the TV spot
In addition, to test our hypothesis that uncertainty labels will increase the sensitivity for details, we edited the video slightly by adding small symbols on the borders of the video (not in the video). As shown in the chapter 2.2. the perceived uncertainty through the label transforms the consumer into a vigilant manner, more likely to detect changes in its environment. After the video was shown, the participants were asked to answer five questions regarding details on the border of the video (Appendix C). These questions served as indicators for a detail score.
Besides, we also asked five questions about details from the story itself, which served as indictor for a peripheral score. We want to stress that the detail cues were designed at the border (black bar) of the video as we expect participants in the uncertainty label condition to be more alert what happens around the video. While the peripheral questions were specifically designed for details from the story. Meaning that respondents who are immersed in the story and take great attention towards the journey of the boy may answer these questions more accurately. An overview of the questions and the answers is given in table 1.
TABLE 1 SURVEY QUESTIONS PERIPHERAL AND DETAIL CUES Peripheral Cues
What colour was the boy's hat?
(1) With coloured stripes.
What was the boy eating?
(1) A sandwich.
(2) An apple. (3) A chocolate bar. (4) I don't know.
What was the man in the car doing, who stopped for the boy at the zebra crossing?
(1) He was at his phone.
(2) He was drinking water. (3) He honked the horn. (4) I don't know.
What kind of transportation was the boy using?
(1) Bike.
(2) Public transportation (Bus or Train).
(3) A taxi. (4) I don't know.
What was the boy carrying with him all the time?
(1) A teddy bear.
(2) A car toy. (3) A book. (4) I don't know.
Detail cues
In part of the video there was a symbol in the upper
left corner. Which was it? (1) A star.
(2) A pencil.
(3) A light bulb.
(4) I don't know. Some scenes had a symbol in the upper right corner.
Which was it? (1) #
(2) > (3) +
(4) I don't know. How often have you seen this symbol?
(1) Once. (2) Five Times.
(3) Three Times.
(4) I don't know. In some scenes a shape occurred at the bottom. Which
was it? (1) a)
(2) b) (3) c)
3.3 Procedure
As we used an online survey to perform the experiment, the participants were first welcomed in written form and informed about their rights and incentivized to finish the study with the chance to win one of two 20€ Amazon vouchers in a lottery (Appendix D for the whole survey). With the beginning of the study, the participants were asked to look on the figure on the next page carefully and informed that they can proceed at earliest after 5 seconds on the next page. Followed by either the uncertainty label or the control label (see 3.2.1.) they were confronted with one of the manipulations of the processing type explained in chapter 3.2.2. Then the video was shown to them, where they were automatically proceeded to the next page after the video finished to ensure no repetitions of the video. Directly after the video we included two items to measure the overall enjoyment. As also used by Redden and Haws (2013) we contained one item “How much did you enjoy watching the commercial?” from 0 = “not at all” to 100 = “very much” to measure enjoyment and one item “How much would you like to continue watching the commercial?” from 0 = “not at all” to 100 = “very much”, which indicates future behavioral intentions.
After the enjoyment measures the participants had to answer the questions explained in chapter 3.2.2 from table 1. From each set of questions a scale could be determined, indicating how many correct answers were given. Further, by computing a difference measure between the peripheral cues and the detail cues we were able to interpret the scores. High scores indicated that more details at the border were detected compared to details in the video, whereas low scores or even negative ones imply that the participant was more focused on the story.
Respondents were then asked to complete a set of questions in a cross table on a 7-point Likert scale from “1” = not at all to “7” = very much as displayed in table 2. These questions functioned as a manipulation check for narrative processing. We used ten well-tested narrative-transportation items adapted by Nielsen and Escalas (2010). The original scale was designed by Green and Brock (2000) and was meant to measure transportation in a reading context. However, Nielsen and Escalas changed the items to adapt it to an advertisement context. In this study, we only deleted one item (“After I finished looking at the advertisement, I found it easy to put it out of my mind”), because this specific item was not useable in our procedure, as the
Which movie academy is responsible for the ad?
“Filmakademie ...” (1) Baden-Württemberg.
(2) Berlin
questions about the TV commercial are already preventing the consumer to put it out of his mind. Furthermore, in this section we also included the reading check item, asking the respondent to choose the number “2” to show that he or she is focused.
Further, we included a scale with four items, from 0 = “not at all” to 100 = “very much, adapted from the same study (Nielsen and Escalas 2010) measuring analytical processing. This set of questions will be used as a manipulation check and are displayed in table 2 as well.
Finally, the participants were asked to answer several questions on control variables. First, they indicated on two questions their involvement with the advertisement: "How involved did you
TABLE 2 NARRATIVE AND ANALYTICAL SCALE ITEMS
Narrative Scale
1 While thinking about the advertisement, I could easily picture the events in it taking place.
2 While watching the advertisement, activity going on in the room around me was on my mind (R).
3 I could picture myself in the scene of events shown in the advertisement.
4 I was mentally involved in the advertisement while looking at it.
After I finished looking at the advertisement, I found it easy to put it out of my mind. (R)
5 The advertisement affected me emotionally.
6 I found myself thinking of ways the advertisement could have turned out differently.
7 I found my mind wandering while I looked at the advertisement (R).
8 The advertisement is relevant to my everyday life.
9 My thoughts about the advertisement changed how I think about the product being advertised.
10 While looking at the advertisement, I had a vivid image of myself using the product being
advertised.
Analytical Scale
1 To what extent did you systematically evaluate the commercial?
2 Did you think about the product's attributes or characteristics?
3 Were you skeptical about the product in the ad?
4 Did you think of reasons why you don't like the ad?
feel with the commercial video?” on a 7-point Likert scale from “1” = not involved at all to “7” = very involved and “How interested were you in the commercial video?” on a 7-point Likert scale from “1” = not interested at all to “7” = very interested. These questions were adapted from Karmarkar and Tormala (2010): Petty and Cacioppo 1979 and designed to rule out a possible alternative explanation. As shown in chapter 2.3 involvement and engagement are critical elements in influencing enjoyment. Therefore, by measuring involvement with the advertisement itself, we can conclude if an effect in enjoyment is due to involvement or due to the manipulations. Second, the participants answered the question whether they have seen the TV commercial before. As we used an actual TV spot, there is a chance that a participant has seen the ad before and therefore, may show different results.
In examining the effects of advertisement attitude toward the advertised brand or product is an important variable (Tormala 2016). As attitudes can influence the evaluation processes, it seems important to explore whether any observed effects in this study are influenced by attitude toward the brand Mercedes. Therefore, we included four items adapted from Tormala, Rucker and Seger (2008) to control for the attitude towards the advertised brand. The respondents were asked to “Please indicate your general attitude towards Mercedes” on semantic differential scales ranging from 1 to 7 with the following anchors: positive – negative; good – bad; favourable – unfavourable; in favour – against (Tormala, Rucker and Seger 2008). Moreover, we tested for a possible gender effect. According to Green and Brock (2000) in narrative processing a gender effect can be observed among narratives in particular. They found ambiguous results testing this possibility. In some narratives a gender effect was observed, whereas in other narrative the gender seems to had no impact on transportation. However, when a gender effect occurred, women were more transported than men (Green and Brock 2000).
4 Results
4.1 Scale reliability
4.1.1 Enjoyment scale
First, starting with the dependent variable, enjoyment was measured using two items: “How much did you enjoy watching the commercial?” from 0 = “not at all” to 100 = “very much” and “How much would you like to continue watching the commercial?” from 0 = “not at all” to 100 = “very much”. The two items were highly correlated with each other, r =.63, p < .01, and therefore, averaged in one reliable scale called enjoyment, α = .73 (Appendix E1). With respect to the data, it has to be stated that the distribution of the enjoyment scale is not normal. As shown in figure 2 extreme values created the distribution of the histogram. Especially the relative high number of respondents reporting values “0” (2.2%) and “100” (4.5%) created a non-normal distribution. As this represent a unique characteristic of the data, they were not excluded from the analysis. Further, the enjoyment scale functions as the main dependent variable and should not be manipulated in order to gain the most truthful results. Anyway, the distribution affects the statistical power of the data set.
FIGURE 2: HISTOGRAM ENJOYMENT
4.1.2 Processing type scales
Two other scales were adapted from the study by Nielsen and Escalas (2010) measuring narrative and analytical processing of the respondents. The narrative scale consisted of ten items as already shown in table 2. The factor analysis suggested either a 4-factor solution or a 1-factor
solution (Appendix E2). On the one hand, a 4-factor solution would be based on an eigenvalue of 1.01 and explained cumulative variance of 69.68%, but only explained 10.10% variance per factor. On the other hand, a 1-factor solution is based on a higher eigenvalue of 3.26 and higher explained variance of 32.55%, but a lower cumulative variance of 32.55%.
In this study and in the following analysis we used one factor to proceed for several reasons. First, the scree plot shows an edge for the 1-factor solution as seen in figure 3. Second, the variance explained in the 4-factor solution of 10.10% is not worth the trade-off for the cumulative variance explained. Lastly, as seen in table 3 the distribution of number of items in the extracted components shows that in a 4-factor solution the item “I found myself thinking of ways the advertisement could have turned out differently” was assigned to a single component. Respectfully, the item could have been deleted from the scale to get better results, but this would manipulate the results. Therefore, we proceeded with one scale called narrative processing in which all ten items were averaged.
FIGURE 3: SCREE PLOT FACTOR ANALYSIS NARRATIVE PROCESSING
The analytical scale did not show the expected reliability (α = .41) for the total scale. However, a performed factor analysis revealed, that a 2-factor solution will provide most solid results with KMO = .51 and a rejected Bartlett's Test of Sphericity. The 2-factor solution does not only show statistically reliability (eigenvalue = 1.36, 69.64% cumulative variance explained, all rotated factor loading > .80) but also can be argued on the content. On the one hand, the first factor contains the items “To what extent did you systematically evaluate the commercial?” and “Did you think about the product attributes or characteristics?” from 0 = “not at all” to 100 = “very much” and obviously implies information about active systematic evaluation. On the other hand, the second factor contains the items “Were you sceptical about the product in the ad?” and “Did you think of reasons not to buy the product?” from 0 = “not at all” to 100 = “very much”, which are more intending for a negative and critical perspective.
Therefore, the analytical processing was measured on two scales with two items each. The first scale is called systematic analytical processing, α = .54, and includes the two items “To what extent did you systematically evaluate the commercial?” and “Did you think about the product attributes or characteristics?” from 0 = “not at all” to 100 = “very much” which are correlated significantly, r = .37, p < .01, while the second scale is called critical analytical processing, α = .55, and includes the two items “Were you sceptical about the product in the ad?” and “Did you think of reasons not to buy the product?” from 0 = “not at all” to 100 = “very much” which are correlated significantly, r = .38, p < .01 (Appendix E3).
4.1.3 Control variable scales
Further, we also included four items adapted from Tormala, Rucker and Seger (2008) to control for the attitude towards the advertised brand. The four items were highly correlated and a factor solution supported the scale’s reliability with eigenvalue = 3.45, cumulative variance explained with 86.22% and all factor loadings > .90. Therefore, they could be averaged together in one scale called attitude with a high reliability, α = .95 (Appendix E4). Lastly, involvement was measured with two items adapted from Karmarkar and Tormala (2010): Petty
TABLE 3 FACTOR ANALYSIS SOLUTIONS FOR NARRATIVE SCALE Factor
Solution
N items in extracted components
I II III IV
I 10
II 7 3
III 5 3 2
and Cacioppo (1979). The two items were correlated significantly, r = .66 and could also be averaged together into one scale called involvement with a high reliability, α = .8.1 (Appendix E5).
4.2 Manipulation checks
4.2.1 Narrative processing
The manipulation check for narrative processing could not be confirmed. A two-way ANOVA (Appendix F) showed no significant effect for the processing type on narrative processing, F (1,181) = .02, p = .88. Figure 4 supports these findings, showing a mean of 3.9 with really low standard deviations for all condition, SD < 1. Rather, it seems that all participants were all immersed into the story.
FIGURE 4: RESULTS FOR NARRATIVE PROCESSING MANIPULATION CHECK
Note: Narrative processing was measured on a 7-point Likert Scale; Error bars denote standard errors
4.2.2 Analytical processing
As figure 5 indicates, the analytical processing conditions scored slightly higher on both analytical scales, compared to the narrative processing condition. A two-way ANOVA (Appendix G) revealed that the processing type condition had a significant effect on the systematic analytical processing scale, F (1,181) = 4.23, p = .04. However, the significant effect of processing type on the second analytical scale, F (1,181) = 1.43, p =.23, was very weak. Therefore, the manipulation of an analytical processing type could be confirmed.
0 1 2 3 4 5 6
Surprise Label Control Label
N ar rat iv e P ro ces si n g
FIGURE 5: RESULTS FOR ANALYTICAL PROCESSING MANIPULATION CHECK
A: Results for systematic analytical processing
B: Results for critical analytical processing
Note: Analytical processing was measured on a scale from “0” = not at all to “100” = very much; Error bars denote standard errors
4.3 Testing our focal hypotheses: Uncertainty and enjoyment
As proposed in our hypothesis, we assumed that the effect of uncertainty labels on enjoyment of a consumption experience would be mediated by an attention shift towards details and that this will be moderated by the processing type. Accordingly, we test if uncertainty labels will reduce enjoyment of a consumption experience compared to the control label when an analytical processing is manipulated not when a narrative processing is evoked.
4.3.1 Enjoyment
A two-way ANOVA revealed a not significant interaction effect of processing type and label condition, F (1,181) = .43, p = .51. Despite the statistical significance, figure 6 shows that there is a slight increase in the enjoyment of the narrative processing type compared to the analytical processing type in the uncertainty label condition. Though, all other three conditions show equal scores on the enjoyment variable (Appendix H).
0 10 20 30 40 50 60 70 80
Surprise Label Control Label
Sco re An al yt ic al Scal e 1
Analytical Processing Narrative Processing
0 10 20 30 40 50 60 70 80 90
Surprise Label Control Label
An al yt ic al Scal e 2
FIGURE 6: RESULTS FOR ENJOYMENT
Note: Enjoyment was measured on a scale from “0” = not at all to “100” = very much; Error bars denote standard errors
4.3.2 Detail and peripheral cue
The aim of this measure was to infer if the focus during consuming the video differed between the conditions and in that way indicate the effect of the possible mediator. As we subtracted the number of correct peripheral cues from the number of correct detail cues, a high score indicated a shift of focus away from the actual video and more on the environment (the border of the video) of it.
As shown in figure 7 there was a slightly lower score for the narrative processing and analytical processing in both label groups. However, we performed a two-way ANOVA to compare the difference measure of vigilance and peripheral cue in the conditions. There was no significant interaction effect, F (1,181) = .04, p = .84 (Appendix I).
0 10 20 30 40 50 60 70 80 90 100
Surprise Label Control Label
En
jo
yment
FIGURE 7: RESULTS ON DIFFERENCE MEASURE DETAIL CUES MINUS PERIPHERAL
CUES
Note: difference measure ranges from max 5 (five correct detail cues minus zero correct peripheral cues) to a minimum of -5 (zero correct detail cues minus five correct peripheral cues); Error bars denote standard errors
Unfortunately, the vigilance cues had some weakness in the design. As it turned out some participants were asked four and other participants five questions. In the set of questions for the vigilance cues, there was one question which was a subquestion to the previous question. Specifically, the respondents were first asked if they have seen a symbol in the upper left corner (# as correct answer) and if they answered correctly, they were asked how often they have seen that symbol. This is a weakness of the study design and probably affected the results.
Accordingly, we tried to control for that effect and created a reduced vigilance score and peripheral score. We deleted the subquestion “How often have you seen this symbol” from the vigilance score and reduced the peripheral score deleting the item with the smallest amount of false answers “What was the boy carry with him all the time” (97.3% correct). We plotted the frequencies of each of the scores in order to compare the effect. As seen in figure 8 both scores follow now a more synchronic path. However, it had no effect on the results of the two-way ANOVA. The interaction effect of processing type and label condition remained still not significant, F (1,180) = .002, p = .97. -5 -4 -3 -2 -1 0
Surprise Label Control Label
D If fer en ce M ea su re
FIGURE 8: FREQUENCY PERIPHERAL AND DETAIL CUES A: Frequency of correct cues with five items
per scale
B: Frequency of correct cues with four items per scale
4.3.3 Moderated mediation of uncertainty label on enjoyment
Due to figure 6 we performed a bootstrap analysis for estimating moderated mediation with 20,000 samples (SPSS Macro PROCESS, Model 7) based on Hayes (2018) (Appendix J). He has noted in his chapter “Miscellaneous Topics and Some Frequently Asked Questions” that there are possibilities to explore and find relationships between variables even if the main interaction effect is not significant (Hayes 2018). Table 4 and figure 9 show the results of the mediation analysis using Process model 7. Supporting the results from the 2-way ANOVA the focus on details (difference measure vigilance cue – peripheral cue) does not mediate the relationship between uncertainty label and enjoyment, c’1 = 2.51, SE = 3.70; 95% CI = [-4.79,
9.8]. However interestingly, there is one path significant, b1 = -3.40, SE = 1.40; 95% CI =
[-6.17, -.63], indicating a negative relationship between the difference measure of vigilance and peripheral cues with enjoyment of the consumption experience. According to the model, a higher difference measure score would decrease enjoyment of the consumption experience. As a higher score indicates a focus shift towards details on the border, this model may provide a first evidence for the overall hypothesis. Though, as seen in table 4, the uncertainty label does not seem to cause this as neither the direct effect on enjoyment, c1 = 2.51, SE = 3.70; 95% CI
= [-4.79, 9.8], neither the direct path to the difference measure, a1 = -.13, SE = .28; 95% CI =
[-.68, 4.3], or the interaction effect of uncertainty label and processing type, a3 = .08, SE = .39;
95% CI = [-.69 , .85], show significant paths. Therefore, we cannot confirm the mediation
0 10 20 30 40 50 60 70 80 90 100 4 3 2 1 0 Fr eq u en cy correc cues
Peripheral Cues Vigilance Cues
0 10 20 30 40 50 60 70 80 90 5 4 3 2 1 0 Fr eq u en cy correct cues
hypothesis of the effect of uncertainty labels. Further, the model showed that the manipulation of the processing type had no moderated mediation effect, a2 = -.13, SE = .28; 95% CI = [-.68,
4.2]. The analysis allowed to interpret the indirect effect of uncertainty label on enjoyment through the mediator on the two levels of the moderator. In the narrative processing condition, 95% CI = [-1.00, 3.04], as well as in the analytical processing condition, 95% CI = [-1.94, 2.67], the 95% confidence interval included zero. Overall, we cannot find evidence for the proposed model in the pure analysis. However, as indicated in the method section, there are several possible variables that could have influences the results. We will test in the following if we still cannot confirm the hypothesis when accounting for possible covariates.
FIGURE 9: MODERATED MEDIATION MODEL
4.4 Results for control variables
4.4.1 Involvement with the ad
The involvement with the TV advertisement did not differ significantly among the conditions, F (1,181) = .97, p = .33. Participants were equally involved with the ad whether they were exposed to an uncertainty or control label or manipulated in their processing type (Appendix K).
4.4.2 Attitude toward the advertised brand
We measured the attitude towards the brand Mercedes as attitudes can have a sufficient impact on one’s evaluation process. Therefore, we used a 4-item measured adapted from Tormala, Rucker and Seger (2008). The data revealed that the attitude did not differ among the four conditions, F (1,181) = .22, p = .64 (Appendix L).
4.4.3 Gender effect
Lastly, an independent-samples t-test was performed to compare narrative processing and analytical processing in female and male groups of participants. The test revealed that the analytical processing was not affected by gender. The systematic analytical processing scale, t (183) = 1.13, p = .26, and the critical analytical processing scale, t(183) = .209, p = .835, did not show significant differences. However, there was a significant difference in the scores on narrative processing for female, M = 4,06, SD = .80, and male participants, M = 3,64, SD = .89); t (183) = -3.32, p = .001 (see Appendix L). Due to this we checked the manipulation check of narrative processing as well as the main hypothesis including gender as a covariate.
TABLE 4 COEFFICIENTS FOR THE CONDITION PROCESS MODEL
Consequent
M(Diff) Y (Enjoy)
Antecedent Coeff SE P Coeff SE p
First, we repeated the two-way ANOVA with the narrative processing scale as dependent variable. Including gender as a covariate in the model, we conducted the ANCOVA accordingly. The results (Appendix M) indicated that there is a significant gender effect as expected, F (1,180) = 11.90, p < .001, but still no significant effect of processing type on the narrative scale, F (1,180) = .97, p = .33. Therefore, we can still not conclude that the manipulation of the narrative processing was successful.
We included gender as a covariate in the two-way ANOVA testing the effect of processing type and label condition on enjoyment. The results of the ANCOVA indicated that there is, as assumed, a significant gender effect, F (1,180) = 12.20, p < .001. However, the interaction effect remains not significant despite including the covariate, F (1,180) = .33, p = .57. In figure 10 we displayed the reduce in enjoyment among male participants compared to female participants. Therefore, we cannot confirm the proposed model.
FIGURE 10: RESULTS FOR ENJOYMENT
Note: Enjoyment was measured on a scale from “0” = not at all to “100” = very much
5 Conclusion and recommendations
We developed our conceptual model showing that a moderated mediation effect will determine the effect of uncertainty labels on enjoyment of a consumption experience. We hypothesized that an uncertainty label reduces, through a process in which the focus shifts towards details, enjoyment of a consumption experience, compared to a neutral label. This focus shift is moderated by the processing type. Accordingly, we expected that uncertainty labels will reduce enjoyment of a consumption experience compared to the control label when an analytical processing is manipulated not when a narrative processing is evoked.
Therefore, we conducted an online experiment exposing the participants to an uncertainty or control label and manipulated the processing type. The participants watched a TV spot as a subsequent consumption experience to the label and reported their enjoyment.
The main analysis revealed no significant interaction effect between the label condition and processing type for the enjoyment of the TV spot. Though we could not accept the hypotheses, we gained some insights in the research. First, we could not replicate the reliability of the scales for the processing type developed by Green and Brock (2000) and adapted by Nielsen and Escalas (2010). We extended the scale from the context of textual narratives to audio-visual narratives and could not agree with Green and Brock (2000) that the narrative processing scale can be transferred from texts to audio-visual material. Second, we provided a new approach on how to measure sensitivity towards details on peripheral or detail cues in the environment. Although we could not report significant results with this measure either, we could present at least a possible measurement. Lastly, we found a significant gender effect on the enjoyment of the video and the narrative processing. This connects to the first argument, that there is a difference between textual or audio-visual material. It may be stressed out that the TV spot is very emotional due to the audio background and the story which may supported the gender effect. Anyway, although we could not contribute to the uncertainty research, this paper may contribute to the research on the use of emotions in ad (e.g. Berger and Milkman 2012).
As already implied in the discussion of the results, there are several limitations of the study and their design. First, the data itself showed some weaknesses. The dataset with only 185 participants did not have high statistical power, although all found significant results can be seen as somewhat valid. Next, there were some issues with some variables. The processing type scales did not show the expected reliability and the main dependent variable enjoyment was not normal distributed. Besides, the design of the study has to be reviewed critically. Especially, the peripheral and detail cues were designed without any tested template. In particular a subquestion turned out to affect the overall scale and the difficulty of the questions was not equal in peripheral and detail cues. Finally, the form of an online experiment may not be the ideal setting to test effects of labels of uncertainty. Delays from exposure to consuming the TV spot may have influences the results and there may be to many distractions under non laboratory conditions.
This paper introduced an experimental, simple and convenient way to follow the focus of consumers of video material without using eye-tracking. This should be improved and tested in future experiments. Moreover, future work should try to extend the findings on narrative and analytical processing with different narratives next to text. This will contribute to the management issue of creating appealing and emotional advertisement for digital use or TV. Finally, the research on perceived uncertainty should elaborate stronger on the psychological process on how it effects subsequent consumption experience. Though our experiment could not prove the theory that perceived uncertainty will decrease enjoyment through a psychological process of a shifting focus, that does not mean that this phenomenon cannot be proved under better experimental circumstances.
6 References
Amazon EU S.à r.l. (2017), "Mit der Amazon Surprise Box innovative Food-Trends entdecken,” (accessed June 06, 2019), [available at https://amazon-presse.de/Top-Navi/Pressetexte/Pressedetail/amazon/de/Produkte/170601_Amazon_Surprise-Box/].
- - - (2019), "Amazon Surprise,” (accessed June 06, 2019), [available at https://www.amazon.de/b?node=13513182031].
Barasch, Alixandra, Zauberman, Gal and Diehl, Kristin (2018), "How the Intention to Share Can Undermine Enjoyment: Photo-Taking Goals and Evaluation of Experiences," Journal
of Consumer Research, 44(6), 1220–1237.
Berger, Jonah and Milkman, Katherine L. (2012), "What Makes Online Content Viral?,"
Journal of Marketing Research, 49(2), 192–205.
Boyd, Brian K. and Fulk, Janet (1996), "Executive Scanning and Perceived Uncertainty: A Multidimensional Model," Journal of Management, 22(1), 1–21.
Bruns, Andreas and Heuwold, Gunnar (2012), "Mercedes-Benz --The Journey (90"),” (accessed June 16, 2019), [available at https://www.youtube.com/watch?v=-V2vHSuUJog].
Chun, HaeEun H., Diehl, Kristin and MacInnis, Deborah J. (2017), "Savoring an Upcoming Experience Affects Ongoing and Remembered Consumption Enjoyment," Journal of
Marketing, 81(3), 96–110.
Cian, Luca, Krishna, Aradhna and Elder, Ryan S. (2015), "A Sign of Things to Come: Behavioral Change through Dynamic Iconography," Journal of Consumer Research, 41(6), 1426–1446.
Deutsche Lufthansa AG (2019), "Lufthansa Surprise,” (accessed June 06, 2019), [available at https://www.lufthansa-surprise.com/#opq_retrieve].
Dohle, Simone and Siegrist, Michael (2014), "Fluency of pharmaceutical drug names predicts perceived hazardousness, assumed side effects and willingness to buy," Journal of Health
Psychology, 19(10), 1241–1249.
Duke, Kristen E., Goldsmith, Kelly and Amir, On (2018), "Is the Preference for Certainty Always So Certain?," Journal of the Association for Consumer Research, 3(1), 0.
Edwards, John A. (2003), "The interactive effects of processing preference and motivation on information processing: Causal uncertainty and the MBTI in a persuasion context," Journal
of Research in Personality, 37(2), 89–99.
Epstein, Seymour, Pacini, Rosemary, Denes-Raj, Veronika and Heier, Harriet (1996), "Individual Differences in Intuitive-Experiential and Analytical-Rational Thinking Styles,"
Journal of Personality and Social Psychology, 71(2), 390–405.
Escalas, Jennifer E. (2007), "Self‐Referencing and Persuasion: Narrative Transportation versus Analytical Elaboration," Journal of Consumer Research, 33(4), 421–429.
Ferrero (2019), "Kinder Surprise Webpage,” (accessed June 06, 2019), [available at https://www.kinder.com/uk/en/kinder-surprise].
Green, Melanie C. and Brock, Timothy C. (2000), "The Role of Transportation in the Persuasiveness of Public Narratives," Journal of Personality and Social Psychology, 79(5), 701–721.
Green, Melanie C., Brock, Timothy C. and Kaufman, Geoff F. (2004), "Understanding Media Enjoyment: The Role of Transportation Into Narrative Worlds," Communication Theory, 14(4), 311–327.
Hayes, Andrew F. (2018), "Introduction to mediation, moderation, and conditional process analysis. A regression-based approach. Second edition. New York, London: The Guilford Press.
Higgins, E. Tory (2006), "Value From Hedonic Experience and Engagement," Psychological
Review, 113(3), 439–460.
Isikman, Elif, MacInnis, Deborah J., Ülkümen, Gülden and Cavanaugh, Lisa A. (2016), "The Effects of Curiosity-Evoking Events on Activity Enjoyment," Journal of Experimental
Psychology: Applied, 22(3), 319–330.
Jauch, Lawrence R. and Kraft, Kenneth L. (1986), "Strategic Management of Uncertainty,"
Academy of Management Review, 11(4), 777–790.
John, Deborah R. and Monga, Alokparna B. (2010), "What Makes Brands Elastic? The Influence of Brand Concept and Styles of Thinking on Brand Extension Evaluation," Journal
of Marketing, 74(3), 80–92.
Karmarkar, Uma R. and Tormala, Zakary L. (2010), "Believe Me, I Have No Idea What I’m Talking About: The Effects of Source Certainty on Consumer Involvement and Persuasion,"
Journal of Consumer Research, 36(6), 1033–1049.
KIA MOTORS CORP. (2015), "KIA Brand Identity,” (accessed June 06, 2019), [available at http://m.kia.com/worldwide/about_kia/brand/brand_identity.do].
Killingsworth, Matthew A. and Gilbert, Daniel T. (2010), "A Wandering Mind Is an Unhappy Mind," Science (New Series), 330(6006), 932.
"Intensive Meditation Training Improves Perceptual Discrimination and Sustained Attention," Psychological Science, 21(6), 829–839.
Meissner, Philip and Wulf, Torsten (2014), "Antecendents and effects of decision comprehensiveness: The role of decision quality and perceived uncertainty," European
Management Journal, 32(4), 625–635.
Nielsen, Jesper H. and Escalas, Jennifer E. (2010), "Easier is not always better: The moderating role of processing type on preference fluency," Journal of Consumer Psychology, 20(3), 295–305.
Novemsky, Nathan, Dhar, Ravi, Schwarz, Norbert and Simonson, Itamar (2007), "Preference Fluency in Choice," Journal of Marketing Research, 44(3), 347–356.
Perry, Stephen D. (2001), "Commercial Humor Enhancement of Program Enjoyment: Gender and Program Appeal As Mitigating Factors," Mass Communication and Society, 4(1), 103– 116.
Petty, Richard E., Brinol, Pablo, Tormala, Zakary L. and Wegener, Duane T. (2007), "The Role of Metacognition in Social Judgment," in Social psychology: Handbook of basic principles, Higgins, Edward T. and Kruglanski, Arie W., New York, NY, Guilford Press, 254–284.
- - - and Cacioppo, John T. (1979), "Issue Involvement Can Increase or Decrease Persuasion by Enhancing Message-Relevant Cognitive Responses," Journal of Personality and Social
Psychology, 37(10), 1915–1926.
Petty, Richard E., Cacioppo, John T. and Schumann, David (1983), "Central and Peripheral Routes to Advertising Effectiveness: The Moderating Role of Involvement," Journal of
Consumer Research, 10(2), 135.
Raney, Arthur A. and Bryant, Jennings (2002), "Moral Judgment and Crime Drama: An Integrated Theory of Enjoyment," Journal of Communication, 52(2), 402–415.
Reber, Rolf, Winkielman, Piotr and Schwartz, Norbert (1998), "Effects of Perceptual Fluency on Affective Judgments," Psychological Science, 9(1), 45–48.
Redden, Joseph P. and Haws, Kelly L. (2013), "Healthy Satiation: The Role of Decreasing Desire in Effective Self-Control," Journal of Consumer Research, 39(5), 1100–1114.
Robertson, Kim (1989), "Strategically desirable brand name characteristics," The Journal of
Consumer Marketing, 6(4), 61–71.
Ruan, Bowen, Hsee, Christopher K. and Lu, Zoe Y. (2018), "The Teasing Effect: An Underappreciated Benefit of Creating and Resolving an Uncertainty," Journal of Marketing
Research, 55(4), 556–570.
Shen, Luxi, Fishbach, Ayelet and Hsee, Christopher K. (2015), "The Motivating-Uncertainty Effect: Uncertainty Increases Resource Investment in the Process of Reward Pursuit,"
Journal of Consumer Research, 41(5), 1301–1315.
Shiv, Baba and Fedorikhin, Alexander (1999), "Heart and Mind in Conflict: The Interplay of Affect and Cognition in Consumer Decision Making," Journal of Consumer Research, 26(3), 278–292.
Tiedens, Larissa Z. and Linton, Susan (2001), "Judgment Under Emotional Certainty and Uncertainty: The Effects of Specific Emotions on Information Processing," Journal of
Personality and Social Psychology, 81(6), 973–988.
Tormala, Zakary L. (2016), "The role of certainty (and uncertainty) in attitudes and persuasion,"
Current Opinion in Psychology, 10, 6–11.
- - -., Rucker, Derek D. and Seger, Charles R. (2008), "When increased confidence yields increased thought: A confidence-matching hypothesis," Journal of Experimental Social
Psychology, 44, 141–147.
Wan, Echo W. and Rucker, Derek D. (2013), "Confidence and Construal Framing: When Confidence Increases versus Decreases Information Processing," Journal of Consumer
Research, 39(5), 977–992.
Warm, Joel S., Parasuraman, Raja and Matthews, Gerald (2008), "Vigilance Requires Hard Mental Work and Is Stressful," Human Factors, 50(3), 433–441.
Whalen, Paul J. (1998), "Fear, Vigilance, and Ambiguity: Initial Neuroimaging Studies of the Human Amygdala," Current Directions in Psychological Science, 7(6), 177–188.
Wilson, Timothy D., Gilbert, Daniel T. and Centerbar, David D. (2002), "Making Sense: The Cause of Emotional Evanescence," Economics and Psychology, 209–233.
Yang, Yang, Gu, Yangjie and Galak, Jeff (2016), "When It Could Have Been Worse, It Gets Better: How Favorable Uncertainty Resolution Slows Hedonic Adaptation," Journal of
