The Teasing Effect: An Underappreciated Benefit of Creating and Resolving an Uncertainty

Abstract

Seven studies covering diverse contexts show an underappreciated benefit of teasing in information acquisition: first creating and then resolving an uncertainty can generate a net positive experience, yet laypeople do not seek out this process. For example, trivia readers report better hedonic experiences if they are first teased with some missing information and then given that information than if they receive all the information at the same time; however, when given a choice, readers prefer to receive all information at the same time. The authors further show that teasing is hedonically beneficial because uncertainty engenders curiosity and thereby builds a potential for a positive experience, whereas uncertainty resolution satisfies the curiosity and thereby realizes that potential. This research yields practical implications by demonstrating that imbuing an ad with an uncertainty creation-resolution process improves the viewer's attitude toward and increases the viewer's willingness to try the advertised product.

Full text

BOWEN RUAN, CHRISTOPHER K. HSEE, and ZOE Y. LU*
Seven studies covering diverse contexts show an underappreciated benefit
of teasing in information acquisition: first creating and then resolving an
uncertainty can generate a net positive experience, yet laypeople do not
seek out this process. For example, trivia readers report better hedonic
experiences if they are first teased with some missing information and then
given that information than if they receive all the information at the same time;
however, when given a choice, readers prefer to receive all information at the
same time. The authors further show that teasing is hedonically beneficial
because uncertainty engenders curiosity and thereby builds a potential for a
positive experience, whereas uncertainty resolution satisfies the curiosity and
thereby realizes that potential. This research yields practical implications by
demonstrating that imbuing an ad with an uncertainty creation–resolution
process improves the viewer’s attitude toward and increases the viewer’s
willingness to try the advertised product.
Keywords: curiosity, uncertainty, hedonic experience, affective forecasting,
teasing
Online Supplement: http://dx.doi.org/10.1509/jmr.15.0346
The Teasing Effect: An Underappreciated
Benefit of Creating and Resolving
an Uncertainty
Uncertainty is generally regarded as aversive (e.g., Buhr and
Dugas 2002; Castaño et al. 2008; Gneezy, List, and Wu 2006;
Gordon 2003; Hsee and Ruan 2016; Van den Bos 2001). When
delivering a message, marketers often avoid uncertainty and
tend to give all information directly. “Use the name within the
first ten seconds,”suggested David Ogilvy (1983), known as
the “father of advertising.”Indeed, most ads reveal the name of a
brand/product up front and involve no uncertainty (Armstrong
2010; Stewart and Furse 1986).
Is it hedonically beneficial to incorporate uncertainty in the
delivery of information? To illustrate, consider two versions
of a smartphone app that offers cool facts about animals: in
one version (the control version, which directly delivers all the
information), the user clicks a button and a piece of trivia
appears on the screen in its entirety (e.g., “Only one type of
mammal can fly. It is the bat.”); in the other version (the treatment
version, which incorporates uncertainty in the delivery of in-
formation), the user clicks a button and only the first part of the
trivia appears, along with a question (e.g., “Only one type of
mammal can fly. Do you know what it is?”), and then after a
short delay, the rest of the trivia appears (e.g., “It is the bat.”).
Does the treatment version create a better hedonic experience
than the control version? We believe so. In this research, we
show that first creating and then resolving an uncertainty can
generate a net positive hedonic experience. Furthermore, we
show that most laypeople underappreciate the hedonic benefit
of this process and do not seek it out.
In practical terms, our research highlights the importance
of using uncertainty in marketing, especially advertising. The
usage of uncertainty, both within and outside of marketing, is
still scant. Table 1 shows a few companies that do intentionally
incorporate uncertainty in information delivery. Apple, a
well-known example, often unveils no information about a
new product until its actual releasetocreatesuspenseand
*Bowen Ruan (corresponding author) is a doctoral candidate in mar-
keting, Wisconsin School of Business, University of Wisconsin–Madison
(email: bruan@wisc.edu). Christopher K. Hsee is the Theodore O. Yntema
Professor of Behavioral Science and Marketing, Booth School of Business,
University of Chicago (email: chris.hsee@chicagobooth.edu). Zoe Y. Lu
is a doctoral student in marketing, Wisconsin School of Business, Uni-
versity of Wisconsin–Madison (email: ylu248@wisc.edu). The first two
authors contributed equally. The authors thank George Loewenstein, Xilin
Li, Minzhe Xu, and Yanping Tu for their thoughtful comments on earlier
versions of this article, and the JMR review team for help and guidance
during the review process. Gerald H¨aubl served as associate editor for this
article.
© 2018, American Marketing Association Journal of Marketing Research
ISSN: 0022-2437 (print) Vol. LV (August 2018), 556–570
1547-7193 (electronic) DOI: 10.1509/jmr.15.0346556

fuel speculation. However, these companies primarily use
uncertaintytopiquecuriosityandcapitalizeonconsumers’
desire to resolve it. Prior research has shown that curiosity
can attract attention (Isikman et al. 2016), facilitate learning
(Kupor and Tormala 2015; Marvin and Shohamy 2016;
Menon and Soman 2002), and encourage purchases (Hill,
Fombelle, and Sirianni 2016; Laran and Tsiros 2013; Polman,
Ruttan, and Peck 2017). It is unclear, however, what happens
when curiosity is resolved. This research focuses on hedonic
experiences, studying how people feel both when they are
curious (i.e., they experience uncertainty) and when their
curiosity is resolved (i.e., they receive the information that
resolves the uncertainty). We show that when watching an ad,
consumers feel better and like the advertised brand more if they
first experience an uncertainty about the identity of the brand
and then resolve that uncertainty than if they experience no
uncertainty. Our research provides an additional reason for
using uncertainty in information delivery and highlights the
importance of uncertainty resolution.
This research also carries practical implications for every-
day life. Hedonic experiences are an important component of
subjective well-being (Kahneman, Wakker, and Sarin 1997),
and positive hedonic experiences often reside in seemingly
insignificant events such as receiving a greeting card from a
friend or seeing a rainbow in the sky (Kahneman et al. 2004).
The positive hedonic experience gained from uncertainty
creation–resolution, as studied in the present research, is one
example of an apparently insignificant event that can affect one’s
momentary hedonic experience and, potentially, one’s subjective
well-being. This research is an example of hedonomics, which
studies how we can increase happiness by improving the
presentation of and choices among external outcomes (such
as external information) without changing the objective levels
of the external outcomes (Hsee, Hastie, and Chen 2008; Tu
and Hsee 2018).
Theoretically, this research contributes to our understanding of
the effects of uncertainty and uncertainty resolution. Most extant
research on uncertainty studies how uncertainty affects judgment
and decision making; it views uncertainty as aversive. We
nevertheless study the hedonic consequence of uncertainty and
show that uncertainty can generate positive experiences, thus
joining a growing body of literature documenting these effects
(e.g., Lee and Qiu 2009; Shen, Fishbach, and Hsee 2015; Shen,
Hsee, and Talloen 2018; Whitchurch, Wilson, and Gilbert 2011).
Our research also complements the literature on curiosity. Prior
research has focused predominantly on the very state of curiosity,
examining its nature (e.g., Dashiell 1925; Fowler 1965; Litman
2005), determinants (e.g., Berlyne 1954a, b), and consequences
(e.g., Hsee and Ruan 2016; Isikman et al. 2016; Menon and
Soman 2002). In contrast, the current research studies the he-
donic experiences of both curiosity and curiosity resolution and
intends to show that, when we control for everything else, the
entire process is hedonically net positive.
Notably, we study only uncertainty about nonnegative
consequences. When uncertainty involves potentially negative
consequences (e.g., “Will my medical test turn out to be
positive?”), it likely causes negative feelings such as anxiety and
worry (Buhr and Dugas 2002; Chew and Ho 1994; Gordon
2003; Mandel and Nowlis 2008; Van den Bos 2001; Wu 1999).
Hence, it is unlikely that an uncertainty creation–resolution
process that contains potentially negative consequences would
engender a net positive hedonic experience.
THEORETICAL FRAMEWORK
Uncertainty Creation and Resolution
Recall the aforementioned animal cool facts app. Unlike
users of the control version, who experience no uncertainty,
users of the treatment version undergo two distinct phases: first,
uncertainty creation and, then, uncertainty resolution.
Will users of the two versions feel different? To answer this
question, we examine the uncertainty-creation phase and the
uncertainty-resolution phase in turn. In the uncertainty-creation
phase, users of the treatment version experience an uncertainty
that represents an information gap, and information gaps in-
duce curiosity (Loewenstein 1994). As summarized in Table 2,
Table 1
EXAMPLES OF COMPANIES USING UNCERTAINTY IN MARKETING
Tactic and Companies Example
Withholding Information About a New Product to Attract
Attention and Fuel Buzz
Sony Ericsson A countdown labeled “Smarter is coming”(referring to the Xperia X10) on its website
Apple Teaser trailer for Mac Pro 2013
Nike Teaser ads for Air Max 360, launched on January 21, 2006
Offering a Mystery Bonus or Discount to Encourage
Purchases or Other Behaviors
Alaska Airlines A mystery bonus of 35%, 40%, or even 50% more miles
American Airlines “Mystery Miles”: 100–100,000 miles earned for “liking”the AAdvantage program on Facebook
United Airlines “Mystery Sale”: Option to buy MileagePlus miles at up to an 80% bonus (account-specific)
InterContinental Hotels Group “Surprise—it’s a mystery bonus! Reveal up to 100% more on your points purchase now.”
Banana Republic “Mystery Money”: Play a game online and find out the prize in the store
Starwood Preferred Guest “Save up to 50% on starpoints. Buy now and see what you will save.”
Macy's “Receive a FREE Mystery Gift with any $50 online beauty purchase.”
Selling Mystery Products at a Relatively Low Price
Groupon “Goods Midsummer Mystery Deal: $10 for a Mystery Product Ranging from Watches to TVs”
Funjet Vacations “Deals so good we can't reveal the hotel name”
That Daily Deal “Mystery Box of Awesome”at $24.99 per month
The Teasing Effect 557

curiosity may cause both positive and negative feelings. On
one hand, curiosity is a state of information deprivation (Litman
and Jimerson 2004; Loewenstein 1994) and thus may cause
discomfort and displeasure (Berlyne 1954a, 1966b; Litman and
Jimerson 2004). On the other hand, curiosity is an arousal state
(Berlyne 1966a; Jepma et al. 2012; Litman 2005). Because we
study only uncertainty about nonnegative information
(e.g., trivia), the associated arousal is likely to be mild, and
mild arousal can be pleasurable (Berlyne 1960; Leuba
1955; Spielberger and Starr 1994). This analysis is con-
sistent with prior research that views curiosity as a feel-
ing of interest, which fits in the larger scheme of positive
emotions (Kashdan and Silvia 2009; Silvia 2008). Moreover,
because users of the treatment version know that the un-
certainty will resolve soon, curiosity may be associated with a
positive feeling of anticipation (Kang et al. 2009; Lee and Qiu
2009; Noordewier and Van Dijk 2015; Nowlis, Mandel, and
McCabe 2004).
Table 2
REVIEW OF RELEVANT RESEARCH
Category and Studies Relevant Conclusion(s)/Finding(s)
Curiosity as a Deprivation State
Berlyne 1950; Dashiell 1925; Dember 1956; Harlow,
Harlow, and Meyer 1950; Welker 1956
When presented with novel, complex, or ambiguous (curiosity-evoking) stimuli,
animals engaged in exploration without additional incentives or even at the cost of
enduring other deprivations.
Berlyne 1954a, 1966b; Litman and Jimerson 2004;
Loewenstein 1994
Curiosity could be viewed as a state of information/knowledge deprivation.
Negative-Valence Curiosity
Han et al. 2013 Participants who were made curious showed less positive brain activity than those
who were not made curious.
Jepma et al., 2012 The induction of curiosity activated brain regions sensitive to conflict and arousal.
Perry et al. 1990; Sieff, Dawes, and Loewenstein 1999 People felt worse when waiting for the result of an HIV test than after they learned
their test result was positive.
Shani and Zeelenberg 2007; Shani, Tykocinski, and
Zeelenberg 2008
People experienced emotional discomfort when they were uncertain whether an
unfavorable decision had been made.
Positive-Valence Curiosity
Hill, Fombelle, and Sirianni 2016 Curiosity is a state of high activation and positive valence.
Kang et al. 2009 The level of curiosity when reading trivia questions correlated with activity in brain
regions involved in anticipated reward.
Kashdan and Silvia 2009; Kashdan and Steger 2007;
Silvia 2008; Silvia and Kashdan 2009
Curiosity induction was rewarding and involved positive feelings of interest.
Knobloch-Westerwick and Keplinger 2006 Participants liked the first part of a mystery more if it involved greater uncertainty and
induced more curiosity.
Leavitt and Christenfeld 2011 Story spoilers might enhance reading enjoyment by increasing readers’curiosity
about a story.
Lee and Qiu 2009 Uncertainty associated with a positive event induced greater and longer-lasting
positive feelings.
Maner and Gerend 2007 Curiosity was more strongly associated with judgments of positive (vs. negative)
outcomes.
Wilson et al. 2005 Uncertainty about a positive experience prolonged the pleasure it caused.
Mixed-Valence Curiosity
Berlyne 1960, 1966a; Spielberger and Starr 1994 Curiosity/uncertainty induced positive (negative) feelings when the associated
arousal was low (high).
Berlyne 1966b Exploratory behavior could be associated with the condition of discomfort due to the
inadequacy of information and an optimal amount of stimulation (e.g., excitement).
Elsen, Pieters, and Wedel 2016 Viewers initially responded negatively toward mystery ads, but they responded more
positively over time.
Litman 2005; Litman and Jimerson 2004;
Litman and Silvia 2006
Curiosity represents both a negative feeling of deprivation and a positive feeling of
interest.
Loewenstein 1994 Curiosity could involve both the pleasant anticipation of acquiring information and
the unpleasant deprivation of not having access to information.
Noordewier and Van Dijk 2015 As people got closer to resolving their curiosity, they experienced more positive
affect, less discomfort, and less annoyance with the lack of information.
Valence of Curiosity Resolution
Jepma et al. 2012 The resolution of curiosity activated brain regions related to reward processing and
memory.
Knobloch-Westerwick and Keplinger 2006 Participants liked a mystery more if the first half of the mystery induced more
curiosity and this curiosity was resolved in the second half of the mystery.
Marvin and Shohamy 2016 The information that resolved curiosity was a reward. It had value in and of itself.
Yang, Gu, and Galak 2017 Favorable uncertainty resolution of the mere possibility of negative experiences
slowed hedonic adaptation of positive experiences.
558 JOURNAL OF MARKETING RESEARCH, AUGUST 2018

In the uncertainty-resolution phase, users of the treatment
condition receive the missing information that resolves the
uncertainty they experience. Because the users have been
curious, receiving the missing information satisfies their
curiosity. Although theorists disagree on whether curiosity
is an internally stimulated homeostatic state (Dashiell 1925;
Fowler 1965) or an externally stimulated deprivation state
(Berlyne 1960), it is generally acknowledged that curiosity
possesses many of the features of primary deprivation states
(Loewenstein 1994). Extensive research has found that people
experience a sense of joyful relief when their deprivation is re-
solved (Brown 1955; Cabanac 1971; Melis and Argiolas 1995).
Thus, people experiencing curiosity—a state of information
deprivation—also enjoy a positive hedonic experience when
receiving the missing information that resolves their curiosity
(Berlyne 1960; Golman and Loewenstein 2015; Litman 2005;
Loewenstein 1994; Marvin and Shohamy 2016). This prop-
osition is consistent with neurological evidence that curiosity
resolution activates brain regions responsible for reward pro-
cessing (Jepma et al. 2012).
The preceding analysis suggests that uncertainty creation is
unlikely to induce an overall negative experience, whereas
uncertainty resolution induces an overall positive experience.
Even if uncertainty creation induces a negative experience, it is
likely to be mild and outweighed by the positive experience of
uncertainty resolution. Therefore, the entire uncertainty creation-
resolution process will produce a net positive experience. This
constitutes our main hypothesis:
H
1
: When receiving information, people enjoy a better overall
experience if they first experience some uncertainty about the
information and then resolve that uncertainty than if they
receive the information directly with no uncertainty.
We refer to the proposed phenomenon as the “teasing
effect.”H
1
is reminiscent of the finding in Ge, H¨aubl, and
Elrod (2012) that although delaying the presentation of
some favorable information about an option reduces the
option’s chance of surviving the screening stage, the delay
can actually increase the likelihood of that option ultimately
being chosen.
We are not the first to study the effects of uncertainty on
feelings. However, the uncertainty studied in most prior re-
search has involved valenced outcomes such as winning a
gamble (positive valence) or losing a gamble (negative va-
lence). For this type of uncertainty, the hedonic experience of
uncertainty resolution depends primarily on the valence of the
outcome (e.g., Andrade and Iyer 2009; Chew and Ho 1994;
Yang, Gu, and Galak 2017). In contrast, we study the hedonic
experience generated by the uncertainty creation–resolution
process, over and above the hedonic experience generated by
the valence of the outcome. Therefore, in our studies, we try to
hold constant the valences of the alternative outcomes of an
uncertain prospect.
Some studies have also investigated the effects of un-
certainty without involving valenced outcomes, but as Table 2
shows, most of the existing studies on the topic have explored
only the effect of uncertainty creation (i.e., how people feel
when they experience uncertainty), not the effect of uncertainty
resolution (i.e., how people feel when the uncertainty is re-
solved). For example, Wilson et al. (2005) show that gift re-
cipients remain happy longer if they do not know the identity of
the gift giver than if they do know.
A few studies have examined the overall effect of uncertainty
and uncertainty resolution without distinguishing between the
two (e.g., Knobloch et al. 2004). One exception is Knobloch-
Westerwick and Keplinger’s (2006) study on mystery novels.
Those authors manipulate the degree of uncertainty early in a
story and the way the uncertainty is resolved later in the story to
examine readers’reactions during both the uncertainty-creation
and uncertainty-resolution phases. However, such manipulation
also changes the content of the story, making it difficult to
tell whether the difference between the uncertainty
creation–resolution condition and other conditions was due
to uncertainty creation-resolution or due to the difference in
the content of the information (the story) delivered.
The present research distinguishes between the uncertainty-
creation phase and the uncertainty-resolution phase. It
shows that when everything else is controlled, an un-
certainty creation–resolution manipulation can create a net
positive hedonic experience, compared with a baseline of
delivering information with no uncertainty.
Attitude and Behavior Toward Associated Stimuli
Prior research on advertising has demonstrated that ad-evoked
feelings influence brand attitudes (Batra and Ray 1986), with
a .35 average correlation between ad-evoked feelings and brand
evaluations (Brown, Homer, and Inman 1998; Pham, Geuens,
and De Pelsmacker 2013). Due to evaluative conditioning, the
mere paring of a brand with the feelings induced by an ad
may even lead to a change in brand evaluation (Gorn 1982). In
line with these findings, we propose that the positive experience
of uncertainty creation-resolution will spill over to the associated
stimuli:
H
2
: People who undergo an uncertainty creation-resolution process
about the information of a stimulus have a more positive attitude
and a more favorable behavioral tendency toward the stimulus
than people who do not undergo the process.
This hypothesized effect echoes existing research studying the
positive effect of uncertainty on marketing. Whereas the ma-
jority of existing research concerns people’s learning (Fazio,
Herr, and Powell 1992; Menon and Soman 2002; Swasy and
Rethans 1986) and behavior (Bernard and Schulze 2005;
Goldsmith and Amir 2010; Hill, Fombelle, and Sirianni 2016;
Laran and Tsiros 2013; Shen, Fishbach, and Hsee 2015) before
uncertainty is resolved and capitalizes on people’sdesireto
resolve uncertainty, the present research demonstrates that even
after uncertainty is resolved, people who have experienced
uncertainty treat the associated items more favorably than do
people who have not experienced uncertainty.
Will People Seek the Uncertainty Creation-
Resolution Process?
Given that an uncertainty creation-resolution process generates
a net positive experience, will laypeople appreciate this positive
teasing effect and voluntarily seek out this process? For example, if
consumers are given a choice between the previously mentioned
two versions of the animal cool facts app, will they choose the
treatment version, which first withholds some information and
then delivers the missing information, or the control version,
which directly delivers all the information? While this question
is largely exploratory, we offer our speculations next.
Notably, this question is about whether people actively seek
uncertainty in the first place, rather than whether they seek to
The Teasing Effect 559

resolve an uncertainty when they are already experiencing
it and feeling curious. Ample extant literature indicates that
if people are already experiencing curiosity, they endeavor
to resolve it (Berlyne 1960; Litman and Jimerson 2004;
Loewenstein 1994; Van Dijk and Zeelenberg 2007). In fact,
the desire to resolve an already induced curiosity can be so
powerful that people seek the missing information even if
doing so may lead to predictably miserable consequences
(Hsee and Ruan 2016; Kruger and Evans 2009).
Returning to the question of whether trivia readers will
choose the control version or the treatment version before they
experience any uncertainty, we predict that people will choose
the control version, because we speculate that they heuristi-
cally dread uncertainty (“being in the dark”), even when the
uncertainty causes no harm. The uncertainty people encounter
in real life often involves potential negative consequences
(Kruger and Evans 2009; Shani, Tykocinski, and Zeelenberg
2008; Sieff etal. 1999; Perry et al. 1990),as with questions like
“Will the reviewers reject my paper?”or “Will my test result
turn out to be positive?”Such uncertainties likely evoke
worries and anxieties. In situations where uncertainty indeed
undermines well-being, people may overlearn the association
between uncertainty and negative feelings and carry it over to
situations in which that association does not apply (Amir and
Ariely 2007; Arkes and Ayton 1999; Hsee, Yang, and Ruan
2015). This analysis concurs with Wilson et al.’s(2005)
finding that people mistakenly predicted that gift recipients
would be happier if the recipients knew the identity of the gift
giver than if they did not. People may predict the happiness of
uncertainty resolution; however, because people are myopic
(e.g., Frederick, Loewenstein, and O’Donoghue 2002; Strotz
1955; Thaler et al. 1997), when deciding whether to undergo
an uncertainty creation-resolution process, they may focus more
on the impending uncertainty-creation phase, which, they be-
lieve, is aversive, than on the ensuing uncertainty-resolution
phase.
Some people do seek uncertainty, for example, reading
mysteries and working on puzzles. These people are likely to
choose the treatment version of the cool-fact app. However,
they may constitute only a minority of the general population,
just like risk-seeking individuals are only a minority of the
general population. Also note that in most cases where people
seek uncertainty, people have control over when they receive
the missing information (e.g., when to read the end of a
mystery and when to learn thesolutionof a puzzle).When such
control is absent (as is the case in our studies), we doubt that
those who seek uncertainty under other circumstances will still
do so here. Based on the speculations in the previous paragraphs,
we suspect that when given a choice between undergoing an
uncertainty creation-resolution process and directly receiving the
information, most people will choose the latter.
We next report seven studies, all conducted on Amazon
Mechanical Turk (MTurk). Unlike many MTurk studies that use
hypothetical scenarios, our studies involve real experiences.
STUDY 1 (TRIVIA)
Method
Study 1 mimicked the animal cool facts app example
mentioned earlier to test H
1
and examine people’schoices.The
stimuli were 11 interesting animal-related facts. For each fact,
we created a control version and a treatment version.
The control version consisted of two statements, one general
(e.g., “Three types of mammals undergo menopause”)and
one specific(e.g.,“They are killer whales, pilot whales, and
humans”). When participants clicked the button, both state-
ments appeared on the screen simultaneously and stayed there
for 20 sec. The treatment version was identical to the control
version except that when participants clicked the button, the
general statement appeared on the screen along with a question
(e.g., “Do you know what they are?”); after 12 sec, the specific
statement replaced the question and stayed for another 8 sec.
Thus, for both the control and the treatment versions, the total
display time for each trial was held constant at 20 sec. In none
of our studies did we ask participants inthe treatment condition
to answer the questions, because we did not want them to
suspect others might judge their performance.
Participants were 229 adults (116 women; M
age
= 39.60 years)
recruited on MTurk for a fixed nominal payment. They were
told they would first complete an animal-related survey, then
take a break, and then complete a second animal-related
survey. The animal-related surveys were a pretext; each
survey asked participants to rate a list of animals on traits such
as friendliness and aggressiveness. After participants com-
pleted the first survey, they were informed that there would
be a break before the second survey began, and they could
read some animal-related cool facts during the break to kill
time. Participants could view as few or as many of these
facts as they wanted. To view a fact, they clicked a button
on the computer screen. If participants no longer wanted to
view more cool facts, they could sit back and wait until the
break was over.
Unbeknownst to the participants, the actual experiment took
place during this break. Participants were assigned to one of
three conditions: control, treatment, and choice. Those in the
treatment condition viewed the treatment version of the trivia,
and those in the control condition viewed the control version.
Participants in the choice condition were given a choice be-
tween the two versions for every fact. Specifically, before a fact
appeared on the screen, they were told they had two viewing
options that would each take 20 sec. To choose, they clicked
one of two buttons on the screen. One said, “First read a
question about the cool fact and then read the cool fact that
answers the question”; the other said, “Directly read the cool
fact.”Once participants clicked a button, they viewed their
chosen version of the fact.
Four minutes into the break, participants were told that the
break was about to end and were asked, “Are you having a
good time during the break?”They rated their experience on a
five-point scale anchored by “not at all”(left) and “very much”
(right).
Results and Discussion
We first report the hedonic-experience result from the treat-
ment and the control conditions, then report the choice result
from the choice condition. Due to space constraints, we report the
main statistics of all studies in Table 3 rather than in the text.
Hedonic experience. To test our primary hypothesis about
the positive effect of uncertainty creation–resolution, we
compared the hedonic-experience ratings between the
treatment condition and the control condition. In support of
H
1
, participants in the treatment condition felt significantly
better. The two conditions did not differ in the number
of cool facts participants read (M = 10.15 and 10.29, and
560 JOURNAL OF MARKETING RESEARCH, AUGUST 2018

SD = 1.47 and 1.11, for treatment and control conditions,
respectively; t(153) = −.66, p= .512).
Choice. To see whether participants voluntarily chose to
undergo an uncertainty creation–resolution process, we ex-
amined the choice distribution of the two versions. In support
of our speculation, fewer than 50% of all facts were viewed in
the treatment version.
The results of Study 1 are susceptible to some alternative
explanations. Instead of our curiosity-based explanation, an
alternative explanation for the happiness result could be the joy
of guessing correctly: participants in the treatment condition
might have guessed most of the answers correctly and thus felt
happier. An alternative explanation for the choice result could
be misunderstanding. Some participants in the choice condi-
tion never tried the treatment version; they might have mis-
takenly believed they would have to answer the questions if
they chose the treatment version, and therefore avoided it.
Study 2 addresses these alternative explanations.
STUDY 2 (CITIES)
Study 2 replicates Study 1 in the context of viewing a slide
show, and it extends Study 1 in several ways. First, to test the
“joy from guessing right”alternative explanation, Study 2
asked participants in the treatment condition to report the
Table 3
RESULTS OF ALL THE STUDIES
Study and Condition N Dependent Variables Key Test Statistics
Study 1
Treatment 79 E = 3.61 (1.17) t(153) = 2.22, p= .028a
Control 76 E = 3.18 (1.21)
Choice 74 E = 3.22 (1.29)
Percentage who chose treatment = 37.0% c
2
(1) = 49.59, p<.001a
Study 2
Treatment 99 E = 3.65 (.99); percentage who continued = 69.7% E: F(1, 192) = 5.59, p= .019a
Control 95 E = 3.32 (.99); percentage who continued = 52.6% Percentage who continued: z = 2.43, p= .015*
Choice 99 Percentage who chose treatment = 37.4% c
2
(1) = 6.31, p= .012a
Chose treatment 37 E = 3.64 (1.01) F(1, 97) <.01, p= .968
Chose control 62 E = 3.63 (1.08)
Study 3
Treatment 98 E1 = 3.86 (.92); E2 = 4.33 (.81) Interaction: F(1, 202) = 11.33, p= .001a
E1: F(1, 202) = .29, p= .590
E2: F(1, 202) = 16.21, p<.001
Control 106 E1 = 3.78 (1.03); E2 = 3.76 (1.14)
Choice 101 Percentage who chose treatment = 19.8% c
2
(1) = 36.84, p<.001a
Percentage who predicted they would enjoy treatment
more = 21.8%
c
2
(1) = 32.17, p<.001a
Chose treatment 20 E1 = 3.95 (.95); E2 = 4.15 (1.23) Interaction: F(1, 99) = 5.08, p= .026
E1: F(1, 99) = .03, p= .869
E2: F(1, 99) = 1.00, p= .320
Chose control 81 E1 = 4.27 (.91); E2 = 3.75 (1.30)
Study 4
Treatment 66 E1 = 7.02 (1.98); E2 = 8.68 (2.10); A = 5.73 (1.21) Interaction: F(1, 126) = 48.54, p<.001a
E1: F(1, 126) = .39, p= .533
E2: F(1, 126) = 22.33, p<.001
A: F(1, 128) = 7.48, p= .007a
Control 64 E1 = 7.25 (1.97); E2 = 6.89 (2.09); A = 5.09 (1.42)
Choice 129 Percentage who chose treatment = 39.5% c
2
(1) = 5.65, p= .017*
Chose treatment 51 E1 = 7.55 (2.27); E2 = 9.02 (1.98) Interaction: F(1, 127) = 28.73, p<.001
E1: F(1, 127) = .36, p= .548
E2: F(1, 127) = 14.12, p<.001
Chose control 78 E1 = 7.79 (2.26); E2 = 7.50 (2.40)
Study 5
Treatment 104 E = 3.35 (.94); A = 3.98 (.61); T = 3.44 (.83) E: F(1, 186) = 70.08, p<.001a
Control 95 E = 2.17 (.88); A = 3.65 (.59); T = 3.01 (.83) A and T: F(2, 190) = 7.14, p= .001a
Study 6
Treatment 53 Percentage who chose the stimulus = 66.0% z = 2.15, p= .032a
Control 53 Percentage who chose the stimulus = 45.3%
Study 7
Treatment (M-F) 66 E1 = 2.62 (.63); E2 = 3.12 (.69) Treatment vs. controls interaction:
F(1, 218) = 66.62, p<.001a
E1: F(1, 218) = .54, p= .462
E2: F(1, 218) = 40.66, p<.001
Control (M-M) 53 E1 = 2.75 (.65); E2 = 2.26 (.81)
Control (F-F) 49 E1 = 2.63 (.60); E2 = 2.39 (.79)
Control (F-M) 54 E1 = 2.69 (.70); E2 = 2.59 (.74)
aTest statistics included in the p-curve analysis.
Notes: E = overall experience; E1 = experience in Phase 1; E2 = experience in Phase 2; A = attitude; T = willingness to try. In the “Dependent Variables”column,
values in parentheses are standard deviations. Because some participants skipped some questions, the degrees of freedom of some tests are smaller than they would
be otherwise.
The Teasing Effect 561

number of correct guesses they had made. Second, to minimize
misunderstanding, Study 2 required participants in the choice
condition to sample both the treatment and control versions of
the stimuli before they made their choice. Third, unlike Study
1, which measured hedonic experience only at the end, Study
2 measured hedonic experience multiple times throughout the
process. Finally, to see whether our manipulation had any
behavioral consequence, Study 2 gave participants the op-
portunity to view more slides after they had viewed the main
slide show, examining whether those in the treatment con-
dition were more willing to view the additional slides.
Method
The stimulus for the study was a slide show of 24 color
pictures of city scenes. The cities were all well-known places,
such as Philadelphia, Tokyo, and Paris, but it was not obvious
from the pictures which cities were depicted. We compiled a
treatment version and a control version for the slide show. In
both versions, each picture stayed on the screen for 12 sec
before switching to the next picture. In the control version, the
name of the city stayed above the picture for the entire display
duration. In the treatment version, a question mark stayed above
the picture for the first 6 sec of the display period, after which the
name of the city replaced the question mark and stayed for the
remaining 6 sec. In short, the two display versions were identical
except that in the treatment version, the name of the city was
concealed during the firsthalfofeachpicture’s display period.
Participants were 293 adults (151 women; M
age
= 36.84 years)
recruitedonMTurkforafixed nominal payment. They were told
they would enjoy a slide show of city scenes and report on their
experience. Study 2 consisted of three conditions: treatment,
control, and choice. Participants in the treatment and the control
conditions viewed the treatment and the control versions of the
slide show, respectively. Participants in the choice condition were
told that the slide show had two versions. They viewed four
sample slides, two from the treatment version and two from the
control version, in a counterbalanced order. The participants
were then asked to choose which version they preferred, and they
viewed the remaining slides in their chosen version.
We measured hedonic experiences six times during the
slide show, once after every four pictures (i.e., after the fourth,
eighth, etc.). On each occasion, participants saw the question,
“Are you having a good time?”and gave their answer on a
five-point scale anchored by “not at all”(left) and “very much”
(right). After viewing all 24 pictures in the slide show, par-
ticipants were asked whether they were willing to view six more
similar pictures without extra compensation. Those who agreed
viewed the additional pictures. At the end, we asked participants
in the treatment condition how many cities they had figured out
correctly before seeing the city names.
Results and Discussion
Hedonic experience. A 2 (condition) × 6 (time) mixed analysis
of variance (ANOVA) found a main effect of condition. Repli-
cating the teasing effect, participants in the treatment condition
felt significantly better than participants in the control condition
(F(1, 192) = 5.59, p= .019). The ANOVA also showed a main
effect of time (F(5, 960) = 23.56, p<.001) and a condition ×
time interaction effect (F(5, 960) = 3.72, p=.002).
An additional analysis using a linear mixed-effects model
(which treated time as a continuous variable and included by-
participant random intercept and random slope) compared the
Figure 1
HEDONIC EXPERIENCE RESULTS OF STUDIES 2–4
A: Study 2
B: Study 3
3.75 3.72 3.77 3.72
3.52 3.44
3.62 3.51 3.48
3.21 3.11 2.97
1
2
3
4
5
Time 1 Time 2 Time 3 Time 4 Time 5 Time 6
Treatment Control
3.86
4.33
3.78 3.76
1
2
3
4
5
Phase 2
Phase 1
Treatment Control
C: Study 4
7.02
8.68
7.25 6.89
1
2
3
4
5
6
7
8
9
10
11
Treatment Control
Phase 1 Phase 2
562 JOURNAL OF MARKETING RESEARCH, AUGUST 2018

temporal slopes of the two conditions and revealed a significant
effect of time (F(1, 192) = 44.28, p<.001) and a significant
condition × time interaction effect (F(1, 192) = 6.10, p= .014). As
Figure 1 shows, although the average hedonic experience declined
over time, the difference between the treatment and the control
conditions increased over time, suggesting that the teasing effect
was not transient and slowed hedonic adaptation to the slide show.
Notably, our results cannot be attributed to boredom be-
cause even toward the end of the slide show, participants in the
control condition did not feel negative (t(94) = −.23, p=.816,
relative to the midpoint). Nor can our results be attributed to the
joy of guessing correctly, because we found no correlation
between self-reported correct guesses and hedonic experience
(r=.08;t(97)=.78,p=.437).
Willingness to continue. Compared with the control con-
dition, more participants in the treatment condition volunteered
to view the additional slides at the end of the main study. This
result suggests that our manipulation influenced not only
experience but also behavior.
Choice. Replicating the choice result of Study 1, more than
half of the participants in the choice condition chose to view
the control version of the slide show, in which they would
know the identity of the city from the beginning. Notably, the
choice result could not be attributed to misunderstanding or
fear of answering questions because before choosing, the
participants had sampled both the treatment and control ver-
sions of the slide show, so they knew they would not need to
answer the questions in the treatment version.
At first glance, the finding that most participants in the
choice condition preferred the control version ofthe slide show
seems at odds with the finding that more participants in the
treatment condition than in the control condition volunteered
to continue. However, this ostensible inconsistency is con-
sistent withour theory: participants in the choice condition saw
both the treatment and the control versions of the stimulus,
could easily tell the difference between the two versions, and,
perhaps because of myopia and aversion to uncertainty, chose
the control version. In contrast, participants in the treatment
condition and in the control condition encountered only their
own version, and because those in the treatment condition had a
better experience, they were more willing to continue.
In all the studies that had a choice condition, we also mea-
sured the hedonic experiences of the participants in the choice
condition while they went through the study in their chosen
version. These results are not highly informative because (1) the
participants were aware of both versions and self-selected the
treatment or the control version, and (2) the numbers of par-
ticipants choosing the two versions were unbalanced (fewer
chose the treatment version). We report these results in Table 3
and Web Appendix A.
STUDY 3 (BIOGRAPHY)
To test our theory further, Study 3 separates the uncertainty-
creation phase from the uncertainty-resolution phase and
measures participants’hedonic experiences in each of the two
phases separately. This deviates from the first two studies, in
which participants in the treatment condition repeatedly
underwent many brief uncertainty creation–resolution cycles;
for example, in Study 2, they underwent an uncertainty
creation–resolution cycle first on the first slide, then on the
second slide, and so on. Because of the brevity of each cycle,
we were unable to assess participants’hedonic experiences in
the uncertainty-creation phase and in the uncertainty-resolution
phase separately. Unlike the first two studies, Study 3 involved
only one uncertainty creation–resolution cycle. The cycle com-
prised two distinct phases: uncertainty creation (Phase 1) and
uncertainty resolution (Phase 2); each phase was relatively
long. Specifically, participants in the treatment condition first
read a series of questions about Einstein (Phase 1) and then
read a biography of Einstein (Phase 2), whereas participants
in the control condition first viewed a series of pictures of Einstein
(Phase 1) and then read the same biography of Einstein (Phase 2).
At the end of each phase, we measured participants’experiences.
According to our theory, the positive effect of an uncertainty
creation–resolution process is not realized until uncertainty
resolution. Thus, in Study 3, we expected to observe not only a
main effect of condition but also an interaction effect between
condition and phase, such that the difference in hedonic ex-
perience between the treatment and the control conditions would
be greater in Phase 2 than in Phase 1.
Another noteworthy feature of Study 3 is that in the choice
condition, we asked the participants not only to choose their
condition (control or treatment) but also to predict which con-
dition they would feel better in. Because we attribute the un-
willingness to choose the treatment condition to participants’
underappreciation of the hedonic benefitofthatcondition,we
expected the prediction results to match the choice results.
Method
The Einstein biography used in the study was adapted from
Biography.com and was divided into ten paragraphs of similar
length, eachabout 200 words. The ten questions were all about
Einstein, for example, “According to Einstein, which two events
had a marked effect on his childhood?”and “What insight led
Einstein to develop his theory of relativity?”The ten pictures all
depicted Einstein in black and white and were available in the
public domain. For examples, see Web Appendix B.
Participants in the study were 305 adults (165 women; M
age
=
36.94 years) recruited on MTurk for a fixed nominal payment.
They were assigned to one of three conditions: treatment,
control, and choice. Participants in the treatment condition went
through two phases. In Phase 1, they viewed the ten Einstein
questions one by one, with each question staying on the screen
for 10 sec before participants could move to the next question;
participants were told they needed only to read the questions, not
to answer them. In Phase 2, participants viewed the ten para-
graphs of the Einstein biography one by one, with each para-
graph staying on the screen for 10 sec before participants could
move to the next paragraph. Participants in the control condition
followed the same procedure, except that in Phase 1, they
viewed the Einstein pictures instead of the Einstein questions.
Participants in the choice condition were first told about the
procedures of the treatment and the control conditions, including
the facts that the two conditions took the same amount of time;
that in the treatment condition they needed only to read the
questions, not to answer them; and that they could find the correct
answers to those questions in the biography. The participants then
viewed two sample Einstein questions and two sample Einstein
pictures. After that, the participants were asked, in a counter-
balanced order, (1) to choose their preferred condition, and, on a
separate screen, (2) to predict the condition in which they would
have a better time. Finally, the participants went through their
chosen condition as described previously.
The Teasing Effect 563

We measured participants’hedonic experiences twice, once
at the end of Phase 1 and once at the end of Phase 2. At the end
of Phase 1, we asked, “Did you have a good time reading the
questions (viewing the pictures)?”At the end of Phase 2, we
asked, “Did you have a good time reading the biography?”
Participants answered each question on a five-point scale an-
chored by “notatall”(left) and “very much”(right). We also
asked participants about their curiosity after surveying their he-
donic experiences. At the end of Phase 1, we asked, “Do you feel
curious about Einstein now?”At the end of Phase 2, we asked,
“Did the biography satisfy your curiosity about Einstein?”Par-
ticipants answered each question on the same five-point scale.
Results and Discussion
Curiosity. As expected, during Phase 1, participants in the
treatment condition experienced more curiosity than did those
in the control condition (M = 4.56 and 4.00, and SD = .69
and 1.02, for treatment and control conditions, respectively;
t(202) = 4.56, p<.001), and during Phase 2, those in the
treatment condition also experienced greater satisfaction of
curiosity than did those in the control condition (M = 4.56 and
4.07, and SD = .72 and 1.12, for treatment and control con-
ditions, respectively; t(202) = 3.74, p<.001).
Hedonic experience. To analyze hedonic experience, we
performed a 2 (condition) × 2 (phase) mixed ANOVA. Rep-
licating the previous findings, the analysis found a main effect of
treatment (F(1, 202) = 7.27, p= .008), a main effect of phase
(F(1, 202) = 9.65, p= .002), and a condition × phase interaction
effect (F(1, 202) = 11.33, p= .001). As Figure 1 shows, the
Phase 1 hedonic experiences were similar between participants
in the treatment condition and in the control condition, but the
Phase 2 hedonic experiences were more positive in the treatment
conditionthaninthecontrolcondition. These results reinforced
our proposition that the effect of uncertainty creation (Phase 1)
manifests itself in uncertainty resolution (Phase 2).
Choice and prediction. Of the participants in the choice
condition, fewer than 50% opted into the treatment condition, and
fewer than 50% predicted a better experience in the treatment
condition. The two results were highly consistent (exact
McNemar's test: p= .625). Even when we analyzed choices and
predictions between participants by examining only the choices
of those who answered the choice question first and the pre-
dictions of those who answered the prediction question first, the
choice and the prediction results were still consistent (18.0% vs.
21.6%; z = .45, p= .653). Thus, the consistency between choice
and prediction was not a within-participants carryover effect.
This suggests that the participants based their decisions on
predicted experience and that their unwillingness to seek the
uncertainty creation–resolution process reflected their under-
appreciation of the hedonic advantage of that process.
Study 3 replicates the positive effect of uncertainty creation
and resolution and shows that the effect arises in uncertainty
resolution rather than in creation. In other words, uncertainty
creation induces a potential for happiness, and uncertainty res-
olution realizes that potential. Furthermore, the choice and
prediction results suggest that laypeople’s failure to seek
the uncertainty creation–resolution process stems from their
failure to predict its hedonic advantage.
STUDY 4 (TARGET CARD)
Study 4 seeks to replicate the findings of the previous studies
using a marketing-relevant stimulus (a gift card). It also tests
H
2
: that an uncertainty creation–resolution process will make
consumers like the associated stimulus (product) more.
Method
We conducted the study on MTurk as an add-on to an
unrelated experiment. After participants completed the un-
related experiment, they were asked whether they were willing
to participate in an extra study about sound evaluation, for
which they would not get paid but instead would get a small
chance to win a $50 gift card. Participants then learned that to
proceed, they had to have speakers or earphones attached to
their computer. We administered an audio test and dismissed
those who chose to do the extra study but failed the test. Out of
632 participants who completed the unrelated study, 259 (151
women; M
age
= 35.85 years) agreed to do the extra study and
passed the audio test.
Upon entering the study, participants first read that their task
was to listen to a sound for about two minutes and report their
feelings. While listening to the sound (repeated piano tones),
participants saw, at the top of the computer screen, “Below is
the gift card you may actually receive. It can be used at...”and,
below the words, a 100-second animation of the $50 gift card.
During the animation, the gift card gradually moved from the
left side to the right side of the screen.
The study consisted of three between-participants condi-
tions: control, treatment, and choice. In the control condition,
participants immediately saw the entire gift card, on which the
store name “Target”was printed; thus, they immediately knew
the card could be used at Target. In the treatment condition, the
gift card was initially inside an envelope. During the anima-
tion, the gift card gradually emerged from the envelope.
Because participants could not tell where the gift card could
be used until the second half of the animation, the first half of
the animation represented the uncertainty-creation phase,
and the second half represented the uncertainty-resolution
phase. In the choice condition, participants were asked to
choose, before the animation began, whether they wanted to
find out where the gift card could be used right away or 60 sec
later. After making their choice, the participants then watched
the animation and went through the study in their chosen
condition.
We measured hedonic experience twice, once at the end of
the first half of the animation and once at the end ofthe second
half, using the question “Overall, how good do you feel now?”
Participants answered the question on an 11-point scale
anchored by “very bad”(left), “neutral”(middle), and “very
good”(right). At the end of the entire study, participants
answered two questions about their attitudes toward Target:
“How do you feel about Target (the store)?”and “How much
do you like Target (the store)?”The first question used a
seven-point scale anchored by “very bad”(left), “neutral”
(middle), and “very good”(right), and the second question
used a seven-point scale anchored by “not at all”(left) and
“very much”(right).
Unlike the previous studies, in which we assigned an equal
number of participants to the choice condition as to each of the
other conditions, in this study we assigned twice as many
participants to the choice condition as to each of the other
conditions. We did so because we were curious whether those
in the choice condition who opted into the treatment condition
were happier than those in the choice condition who opted into
the control condition. The previous studies were unable to
564 JOURNAL OF MARKETING RESEARCH, AUGUST 2018

address this question because the sample size in the choice
condition was small and thus rendered insufficient power for
the comparison.
Results and Discussion
Hedonic experiences. The results, presented in Figure 1,
replicated the previous findings: a 2 (condition) × 2 (phase)
mixed ANOVA found a main effect of uncertainty (F(1, 126) =
5.34, p= .023), a main effect of phase (F(1, 126) = 19.57, p<
.001), and a condition × phase interaction effect (F(1, 126) =
48.54, p<.001). Relative to participants in the control con-
dition, participants in the treatment condition felt similar
during the uncertainty-creation phase, and felt significantly
better during the uncertainty-resolution phase. The results
suggest that uncertainty creation induces a potential for a
positive hedonic experience, which is not realized until the
uncertainty is resolved.
Choice. More than 50% of participants in the choice
condition chose to find out where the gift card could be used
right away rather than to wait 60 sec before finding out. This
result replicates the findings of our previous studies and
parallels Galak, Kruger, and Loewenstein’s(2013)finding
about eating: people tend to eat quickly rather than slowly,
even though eating slowly is more pleasurable.
Attitude. As noted earlier, we asked two questions about
attitude toward Target. Because responses to the two questions
were highly correlated (r = .91; t(128) = 24.57, p<.001), we
averaged them to create a composite attitude score; we found
that participants in the treatment condition had a more positive
average attitude toward Target than did participants in the
control condition. In support of H
2
, the result suggests that an
uncertainty creation-resolution process not only makes people
happier but also makes them like the associated brand more.
To show that uncertainty creation–resolution is vital to
producing the teasing effect, and to rule out a mere-stimulus-
change account and a mere-uncertainty account, we conducted
Study 7 and manipulated uncertainty in both Phase 1 and Phase
2. The results were consistent with our theory and ruled out the
two alternative accounts (for details see Web Appendix C).
STUDY 5 (ADS)
We designed Studies 5 and 6 primarily to test H
2
and show
the implication of our research for ads. Study 5 comprised
a treatment condition and a control condition. Participants
viewed four ads and reported their overall hedonic expe-
rience, their attitudes toward the advertised brands/products,
and their willingness to try the brands/products. Each ad had
two versions, which were identical except that one revealed
the identity of the brand/product at the beginning (the control
version), whereas the other revealed the identity at the end
(the treatment version). We predicted that viewers of the
treatment version would feel better, would like the brand/
product more, and would be more willing to try it.
Method
This study showed ads for four rather diverse brands/products:
Burberry, Toyota Corolla, Windex, and IBM. Each ad con-
sisted of an animated PowerPoint slide show. Each slide show
had two versions: treatment and control. The treatment version
first prompted viewers to guess what the brand/product was,
then provided descriptions of the brand/product, and finally
displayed the image of the brand/product along with its name.
The control version was identical except that the identity of the
brand/product was revealed at the beginning. For example, in
the Burberry slide show, the treatment and the control versions
differed only in the first slide. In the treatment version, the first
slide read, “Guess what the brand to be described is.”In the
control version, the first slide read, “Introducing Burberry.”
Then, both versions provided several lines of descriptions
about Burberry (see Web Appendix B), with each line grad-
ually appearing on the screen and then gradually giving way to
the next line. On the final slide, both versions presented a
picture of Burberry’s brand logo along with the brand name,
“Burberry.”The logo and the brand name stayed on the screen
for 5 sec before they disappeared, marking the end of the
slideshow. The ads for the other three brands/products fol-
lowed the same structure.
Participants were 199 adults (100 women; M
age
=37.33
years) recruited from MTurk for a nominal payment. They
were assigned randomly to one of two conditions: treatment and
control. In each condition, participants viewed, in a random
order, the four ads in their respective versions. At the end of each
slide show, they were first asked, “Did you feel curious when
viewing the ad?”After that, they were asked (1) “Overall, did
you have a good time when viewing the ad?”(2) “How do you
feel about the brand/product in the ad?”and (3) “Would you like
to try the brand/product if you have never tried it?”The par-
ticipants answered each question on a five-point scale an-
chored by “notatall”(left) and “very much”(right). In
addition, participants in the treatment condition were asked if
they had guessed the identity of the brand/product correctly
before it was revealed in each ad.
Results and Discussion
Curiosity. As expected, participants in the treatment con-
dition reported greater curiosity for each ad (M = 3.68, 3.92,
3.60, and 3.66, and SD = 1.07, .91, 1.00, and 1.00, for the four
treatment ads, respectively) than participants in the control
condition (M= 2.42, 2.15, 2.25, and 2.17, and SD = 1.08, 1.11,
1.07, and .97, for the four control ads, respectively; all t >8.27,
all p<.001).
Hedonic experience. A 2 (condition) × 4 (brand/product)
mixed ANOVA on participants’overall hedonic-experience
ratings generated a significant effect of condition (F(1, 186) =
70.08, p<.001), a main effect of brand/product (F(3, 558) =
3.07, p= .027), and no interaction effect (F(3, 558) = 1.44,
p= .232). Participants in the treatment condition were sig-
nificantly happier than participants in the control condition,
replicating the teasing effect. As in Study 2, we found no
correlation in the treatment condition between self-reported
correct guesses and hedonic experience (r = −.05; t(102) = −.52,
p= .603), further ruling out “joy from guessing right”as an
alternative explanation for the teasing effect.
Attitude and willingness to try. Because attitude and will-
ingness to try were highly correlated (all r >.71, all p<.001,
across different brands/products), we conducted a 2 (condition) ×
4 (brand/product) mixed multivariate analysis of variance
(MANOVA) on these two measures. The results, based
on Wilks’s lambda, revealed a significant effect of con-
dition (F(2, 190) = 7.14, p= .001), a significant effect of brand/
product (F(6, 186) = 11.90, p<.001),andnosignificant
interaction effect (F(6, 186) = 1.43, p= .204), suggesting
that our manipulation yields positive effects even on a
brand/product.
The Teasing Effect 565

Our results may appear contrary to Baker, Honea, and
Russell’s(2004)finding that participants had better associative
learning of and thus more favorable attitudes toward a brand
in an ad if the identity of that brand was revealed at the be-
ginning than at the end of the ad. However, a closer look
suggests a major difference between the two streams of re-
search: whereas we study well-known brands, Baker, Honea,
and Russell (2004) investigated unknown brands. In line with
prior research showing that curiosity is positively related to
one’s prior knowledge in a domain (Berlyne 1954b; Jones
1979; Loewenstein 1994), we suspect that imbuing the ad of an
unknown brand with an uncertainty creation–resolution pro-
cess did not induce much curiosity from the viewer and
therefore did not generate a net positive experience.
STUDY 6 (STARBUCKS CARD)
Study 6 replicates Study 5 with real consequences to par-
ticipants. In Study 5, we asked participants about their curiosity
and attitudes before measuring their behavioral tendency
(willingness to try), possibly biasing the behavioral-tendency
response. InStudy 6, we directly asked participants about their
choice without first asking them about psychological variables.
Method
Participants in the study were 106 adults (54 women; M
age
=
35.46 years) recruited from MTurk for a nominal payment.
We embedded this study in an unrelated survey. At the end of
the unrelated survey, participants were told that beyond the
promised payment for completing the survey, they had a
chance to win a gift.
Participants were assigned to either a treatment or a control
condition. Participants in the treatment condition were first
asked, “Wanna know what the gift is?”They were then
presented, line by line, with some descriptions of the gift
(see Web Appendix B). At the end, participants were informed,
“It is a $15 Starbucks gift card”and shown the image of the
card. Participants in the control condition were first informed,
“It is a $15 Starbucks gift card”and then saw the same de-
scriptions and image as those in the treatment condition.
After these steps, participants in both conditions were asked
whether they would prefer the $15 Starbucks gift card or a $5
Amazon gift card; they were told they would actually receive
the card of their choice if they won. The Amazon gift card
served as a common alternative option and enabled us to assess
participants’relative preferences for the Starbucks gift card
(the focal stimulus) between the treatment and the control
conditions. Upon completion of the study, we sent the winners
the gift cards of their choice.
Results and Discussion
Because we wanted to keep this study as natural as possible
and minimize the influence of asking questions per se, we
asked participants only to indicate their preferences for the
Starbucks card relative to the Amazon card and did not ask
them to report their feelings. In support of H
2
and replicating
the behavioral results of Study 5, more participants in the
treatment condition than in the control condition chose the
Starbucks card.
A subtle uncertainty manipulation can influence even real
choices. Ceteris paribus, people who first experience un-
certainty about the identity of a product and then resolve that
uncertainty are more likely to choose the product than are
people who learn the product’s identity directly.
GENERAL DISCUSSION
We are living in an epoch of abundant information, yet we
have limited knowledge of how we can deliver and receive
information better. The present research enriches our un-
derstanding of the topic and makes multiple contributions. We
propose and identify a positive hedonic effect of teasing: that
first creatingand then resolving an uncertainty engenders a net
gain in happiness. This effect exemplifies the hedonomics
approach to happiness, increasing happiness by improving the
presentation of and choices among external stimuli without
changing their objective levels (Hsee, Hastie, and Chen 2008;
Tu and Hsee 2018). Furthermore, we find that laypeople
underappreciate the hedonic benefit of such a process and do
not seek it out. Finally, we show thatimbuing an ad withsuch a
process enhances consumers’attitudes toward, willingness to
try, and choice of the advertised product.
To test the reliability of the teasing effect, we conducted a
single-paper meta-analysis (SPM; McShane and B¨ockenholt
2017) including all the studies except Study 6, which did not
measure hedonic experience. The SPM estimates the teasing
effect at .55 (95% confidence interval = [.26, .85]), demon-
strating the robustness of the effect. We also submitted key
statistical results into a p-curve analysis (Simonsohn, Nelson,
and Simmons 2014). Both the half p-curve test (z = −10.80,
p<.001) and the full p-curve test (z = −11.27, p<.001) are
right-skewed, providing strong support for the presence of
evidential value.
Applied Implications
This research carries applied implications for marketing and
beyond. First, as demonstrated in Studies 5 and 6, it has
implications for designing ads that are both more pleasurable
to watch and more effective. In real life, many ads are dis-
played in situations that involve waiting; for example, waiting
for a bus, riding an elevator or subway, waiting for a website or
video to load, and so on. In these situations, our research is
helpful from the perspectives of both consumer well-being and
advertising effectiveness. Another implication regards the
design of software and smartphone apps, as we alluded to in
our opening example. Consider two versions of a “word-a-
day”app, an example similar to the animal cool facts app
example introduced earlier. Every morning, one version of the
app delivers an esoteric word along with its definition (e.g.,
“‘Brogue’means an Irish accent.”). The other version first
teases the reader with a question (e.g., “Does the word ‘brogue’
mean (a) an American accent, (b) a British accent, or (c) an
Irish accent?”) and, after a delay, delivers the answer. Most
existing apps resemble the first version. But we predict that the
second version would be better at brightening one’sday.
A third implication concerns the design of information
guides (booklets and websites). Consider two versions of an art
museum pamphlet that museum visitors receive when pur-
chasing their tickets. One version directly relays facts about the
artwork in the museum, whereas the other version simply
poses questions about the artwork and lets the visitors find out
the answers themselves while visiting the museum. We predict
that visitors receiving the second version would have a better
time in the museum.
566 JOURNAL OF MARKETING RESEARCH, AUGUST 2018

Fourth, this research also suggests that ads that interrupt a
television program can enhance viewers’enjoyment. In fact,
Nelson, Meyvis, and Galak (2009) have already documented
this effect and explain it interms of adaptation, positing thatthe
ads disrupt the viewers’adaption of the program. Our research
offers an additional (though not necessarily alternative) ex-
planation: the presence of an ad creates uncertainty about the
upcoming content of the program, which is later resolved when
the program resumes after the ad. The viewers thus experience
an uncertainty creation–resolution process and, therefore, a net
gain in happiness.
Fifth, our findings carry implications for teaching and in-
terpersonal communications. Some teachers and speakers al-
ready take advantage of uncertainty by first asking questions
and then giving answers (Gaylen 1998), but many still do not.
Even those who do primarily use uncertainty to help the au-
dience learn (Hawes 2004). Our research suggests that doing
so has the additional benefit of making the audience feel
happier. We further speculate that if teachers routinely
create uncertainty about some knowledge and later resolve
the uncertainty, students will learn to enjoy the process
(Csikszentmihalyi 1990; Kapp 2012), develop long-lasting
habits of knowledge seeking (Kashdan and Silvia 2009;
Silvia 2008; Silvia and Kashdan 2009), and live a happier
life for years to come (Kashdan and Steger 2007).
Finally, our research provides insights about gamification.
Uncertainty and curiosity are important components of a game
(Seaborn and Fels 2015). Our research suggests that the un-
certainty creation–resolution manipulation can make a non-
game activity, such as reading a biography or watching an ad,
enjoyable and gamelike. Shen, Hsee, and Talloen (2018) have
recently documented a robust reinforcing uncertainty effect:
people repeated a boring task more when the incentive for the
activity was uncertain a priori (e.g., either $1 or $2) than when
it was certain and larger a priori (e.g., $2 for sure). Although
this finding is counternormative, it agrees with our current
theory. Every time a person in the uncertain-incentive con-
dition repeated the activity, he or she underwent an uncertainty
creation–resolution cycle. This process made the otherwise
boring task enjoyable and gamelike. Also consistent with our
theory, this counternormative reinforcing uncertainty effect
occurred only when people in the uncertain-incentive condi-
tion could find out the incentive immediately and thus resolve
the uncertainty every time they repeated the activity; it dis-
appeared when the uncertainty was not resolved immediately
(Shen, Hsee, and Talloen 2018).
Remaining Questions
What role does curiosity play? According to our theory,
curiosity is vital in producing the teasing effect. Indeed, in all
studies in which we asked participants to rate their curiosity, the
ratings not only differed between the treatment and the control
conditions but also fully mediated the effect of our uncertainty
creation–resolution manipulation on hedonic experience (see
Web Appendix A), which attests to the importance of curiosity
in the process.
Is curiosity good or bad? Curiosity carries both risk and
potential. Hsee and Ruan (2016) find that people experiencing
curiosity acted to resolve it even if the outcome was expectedly
aversive, such as receiving electric shocks. Dubbed the
“Pandora effect,”this result exemplifies the risk of curiosity.
On the other hand, the present research reveals the potential of
curiosity. Notably, the research on the Pandora effect focuses
on uncertainty about potentially negative outcomes (e.g.,
electric shocks), and shows that after curiosity was induced,
people could not resist the temptation to resolve their curiosity
despite the negative consequences. In contrast, the present
research concerns uncertainty about nonnegative outcomes
and demonstrates that before curiosity is induced, people do
not seek out curiosity despite its positive potential.
What roles do specific feelings (such as engagement and
relief) play? Like much of the existing research on hedonic
experience (e.g., Galak, Kruger, and Loewenstein 2013;
Mandel and Nowlis 2008; Wilson et al. 2005), the present
research focuses on overall experience, as measured by questions
such as “Are you having a good time now?”rather than on the
specific feelings (besides curiosity) involved in the overall ex-
perience. However, we speculate that participants in the treat-
ment conditions of our studies may have experienced different
specific feelings during uncertainty creation and uncertainty
resolution: tension and engagement during uncertainty creation,
and relief and satisfaction during uncertainty resolution. Al-
though they are beyond the scope of the present research, such
specific feelings can shed light on what constitutes overall he-
donic experience, and they deserve future research.
Does the length of the uncertainty-creation phase matter?
We suspect that for uncertainty creation–resolution to yield a
positive effect, the length of the uncertainty-creation phase
should be neither too brief nor too long. Too brief a phase is
insufficient to evoke curiosity, and too long a phase may cause
anxiety or lose the interest of the person (Noordewier and Van
Dijk 2015).
Can evaluation mode explain the choice/experience
discrepancy? In studies that included a choice condition,
the experiencers (i.e., participants in the treatment and the
control conditions) favored the treatment condition, while the
choosers (i.e., participants in the choice condition) favored
the control condition. A potential explanation for the in-
consistency is that the choosers were in the joint-evaluation
mode—they knew both the treatment and the control condi-
tions and could compare the two—whereas the experiencers
were in the single-evaluation mode, exposed to one condition
only (for further discussion of joint vs. single evaluations, see
Hsee 1996). Given the studies in this research, we cannot tell
whether this is an alternative explanation or part of our ex-
planation (e.g., that joint-evaluation mode highlights the initial
information deficiency in the treatment condition), and we
look forward to further research on this issue.
Would choosers learn? We speculate that people have
learned from their lifelong experience that uncertainty is
generally aversive and thus do not actively seek uncertainty.
However, if people experience a benign learning environment
(e.g., one in which uncertainty is always about innocuous
outcomes and is always resolved) with repeated trials, they will
learn to appreciate the benefit of temporary uncertainty and
will seek it out.
Broader Implications
In addition to the aforementioned specific implications, this
research also holds broader implications about desire in
general. The overarching theme of this research is to use a
process of desire creation and resolution to distill happiness.
This theme is more important today than ever before. For most
of human history, our ancestors lacked the resources to satisfy
The Teasing Effect 567

even their most basic desires (Newman 1990). In those times,
one of the most effective ways to gain pleasure was to satisfy a
basic desire. Historically, this is probably why it is popular in
many cultures to treat friends by cooking them a meal. Relative
to our ancestors, we are living in a world of abundance. For
many people, there is too much to eat, not too little. With the
advances of technology and the accumulation of wealth, more
and more of us face a “ceiling effect.”To generate additional
happiness, we need additional desires. Thus, the key to hap-
piness is not to satisfy existing desires but to induce new
desires. But most basic desires cannot easily be induced, and
even if they could, inducing such desires would entail costs and
side effects. Curiosity—at least the type of curiosity studied in
this research—is a desire that is readily induced and can be
satisfied with little to no cost. This research shows the potential
of using curiosity induction and resolution to boost happiness.
REFERENCES
Amir, On, and Dan Ariely (2007), “Decisions by Rules: The Case of
Unwillingness to Pay for Beneficial Delays,”Journal of Marketing
Research,44(1),142–52.
Andrade, Eduardo B., and Ganesh Iyer (2009), “Planned Versus
Actual Betting in Sequential Gambles,”Journal of Marketing
Research, 46 (3), 372–83.
Arkes, Hal R., and Peter Ayton (1999), “The Sunk Cost and Con-
corde Effects: Are Humans Less Rational Than Lower Animals?”
Psychological Bulletin, 125 (5), 591–600.
Armstrong, J.S. (2010), Persuasive Advertising: Evidence-Based
Principles. Hampshire, UK: Palgrave MacMillan.
Baker, William E., Heather Honea, and Cristel A. Russell (2004),
“Do Not Wait to Reveal the Brand Name: The Effect of Brand-
Name Placement on Television Advertising Effectiveness,”
Journal of Advertising, 33 (3), 77–85.
Batra, Rajeev, and Michael L. Ray (1986), “Affective Responses
Mediating Acceptance of Advertising,”Journal of Consumer
Research, 13 (2), 234–49.
Berlyne, Daniel E. (1950), “Novelty and Curiosity as Determinants
of Exploratory Behavior,”British Journal of Psychology,41(1/2),
68–80.
Berlyne, Daniel E. (1954a), “A Theory of Human Curiosity,”British
Journal of Psychology, 45 (3), 180–91.
Berlyne, Daniel E. (1954b), “An Experimental Study of Human
Curiosity,”British Journal of Psychology, 45 (4), 256–65.
Berlyne, Daniel E. (1960), Conflict, Arousal, and Curiosity. New
York: McGraw-Hill.
Berlyne, Daniel E. (1966a), “Conflict and Arousal,”Scientific
American, 215 (2), 82–87.
Berlyne, Daniel E. (1966b), “Curiosity and Exploration,”Science,
153 (3731), 25–33.
Bernard, John C., and William Schulze (2005), “The Next New
Thing: Curiosity and the Motivation to Purchase Novel Products,”
Economic Bulletin,3(32),1–8.
Brown, Judson S. (1955), “Pleasure-Seeking Behavior and the
Drive-Reduction Hypothesis,”Psychological Review, 62 (3),
169–79.
Brown, Steven P., Pamela M. Homer, and J. Jeffrey Inman (1998),
“A Meta-Analysis of Relationships Between Ad-Evoked Feelings
and Advertising Responses,”Journal of Marketing Research,
35 (1), 114–26.
Buhr, Kristine, and Michael J. Dugas (2002), “The Intolerance of
Uncertainty Scale: Psychometric Properties of the English Version,”
Behaviour Research and Therapy,40(8),931–45.
Cabanac, Michel (1971), “Physiological Role of Pleasure,”Science,
173 (4002), 1103–07.
Castaño, Raquel, Mita Sujan, Manish Kacker, and Harish Sujan
(2008), “Managing Consumer Uncertainty in the Adoption of
New Products: Temporal Distance and Mental Simulation,”
Journal of Marketing Research, 45 (3), 320–36.
Chew, Soo Hong, and Joanna L. Ho (1994), “Hope: an Empirical
Study of Attitude Toward the Timing of Uncertainty Resolution,”
Journal of Risk and Uncertainty, 8 (3), 267–88.
Csikszentmihalyi, Mihaly (1990), Flow. New York: Harper & Row.
Dashiell, John F. (1925), “A Quantitative Demonstration of Animal
Drive,”Journal of Comparative Psychology, 5 (3), 205–08.
Dember, W.N. (1956), “Response by the Rat to Environmental Change,”
Journal of Comparative and Physiological Psychology,49(1),93–95.
Elsen, Millie, Rik Pieters, and Michel Wedel (2016), “Thin Slice
Impressions: How Advertising Evaluation Depends on Exposure
Duration,”Journal of Marketing Research, 53 (4), 563–79.
Fazio, Russell H., Paul M. Herr, and Martha C. Powell (1992), “On
the Development and Strength of Category-Brand Associations in
Memory: The Case of Mystery Ads,”Journal of Consumer
Psychology, 1 (1), 1–13.
Frederick, Shane, George Loewenstein, and Ted O’Donoghue
(2002), “Time Discounting and Time Preference: A Critical
Review,”Journal of Economic Literature, 40 (2), 351–401.
Fowler, Harry (1965), Curiosity and Exploratory Behavior. New
York: Macmillan.
Galak, Jeff, Justin Kruger, and George Loewenstein (2013), “Slow
Down! Insensitivity to Rate of Consumption Leads to Avoidable
Satiation,”Journal of Consumer Research, 39 (5), 993–1009.
Gaylen, Nancy (1998), “Encouraging Curiosity at Home,”Science
and Children, 35 (4), 24–25.
Ge, Xin, Gerald H¨aubl, and Terry Elrod (2012), “What to Say When:
Consumer Choice by Delaying the Presentation of Favorable
Information,”Journal of Consumer Research, 38 (6), 1004–21.
Gneezy, Uri, John A. List, and George Wu (2006), “The Uncertainty
Effect: When a Risky Prospect Is Valued Less Than Its Worst
Possible Outcome,”Quarterly Journal of Economics, 121 (4),
1283–309.
Goldsmith, Kelly, and On Amir (2010), “Can Uncertainty Improve
Promotions?”Journal of Marketing Research, 47 (6), 1070–77.
Golman, Russell, and George Loewenstein (2015), “Curiosity,
Information Gaps, and the Utility of Knowledge,”working paper,
Carnegie Mellon University.
Gordon, Kerry (2003), “The Impermanence of Being: Toward a
Psychology of Uncertainty,”Journal of Humanistic Psychology,
43 (2), 96–117.
Gorn, Gerald J. (1982), “The Effects of Music in Advertising on
Choice Behavior: A Classical Conditioning Approach,”Journal
of Marketing, 46 (1), 94–101.
Han, Chunhui, Peng Li, Christopher Warren, Tingyong Feng,
Jordan Litman, and Hong Li (2013), “Electrophysiological Ev-
idence for the Importance of Interpersonal Curiosity,”Brain
Research, 1500, 45–54.
Harlow, Harry F., Margaret K. Harlow, and Donald R. Meyer
(1950), “Learning Motivated by a Manipulation Drive,”Journal
of Experimental Psychology, 40 (2), 228–34.
Hawes, Jon M. (2004), “Teaching Is Not Telling: The Case Method
as a Form of Interactive Learning,”Journal for Advancement of
Marketing Education, 5 (4), 47–54.
Hill, Krista M., Paul W. Fombelle, and Nancy J. Sirianni (2016),
“Shopping Under the Influence of Curiosity: How Retailers Use
Mystery to Drive Purchase Motivation,”Journal of Business
Research, 69 (3), 1028–34.
Hsee, Christopher K. (1996), “The Evaluability Hypothesis: An
Explanation for Preference Reversals Between Joint and Separate
Evaluations of Alternatives,”Organizational Behavior and
Human Decision Processes, 67 (3), 247–57.
568 JOURNAL OF MARKETING RESEARCH, AUGUST 2018

Hsee, Christopher K., Reid Hastie, and Jingqiu Chen (2008),
“Hedonomics: Bridging Decision Research with Happiness Re-
search,”Perspectives on Psychological Science, 3 (3), 224–43.
Hsee, Christopher K., and Bowen Ruan (2016), “The Pandora
Effect: The Power and Peril of Curiosity,”Psychological Science,
27 (5), 659–66.
Hsee, Christopher K., Yang Yang, and Bowen Ruan (2015), “The
Mere Reaction Effect: Even Non-Positive and Non-Informative
Reactions Can Reinforce Actions,”Journal of Consumer Re-
search, 42 (3), 420–34.
Isikman, Elif, Deborah J. MacInnis, G¨ulden ¨
Ulk¨umen, and Lisa A.
Cavanaugh (2016), “The Effects of Curiosity-Evoking Events on
Activity Enjoyment,”Journal of Experimental Psychology. Ap-
plied, 22 (3), 319–30.
Jepma, Marieke, Rinus G. Verdonschot, Henk van Steenbergen,
Serge A.R.B. Rombouts, and Sander Nieuwenhuis (2012),
“Neural Mechanisms Underlying the Induction and Relief of Per-
ceptual Curiosity,”Frontiers in Behavioral Neuroscience, 6 (5), 1–9.
Jones, R. Stewart (1979), “Curiosity and Knowledge,”Psycho-
logical Reports, 45 (2), 639–42.
Kahneman, Daniel, Alan B. Krueger, David A. Schkade, Norbert
Schwarz, and Arthur A. Stone (2004), “A Survey Method for
Characterizing Daily Life Experience: The Day Reconstruction
Method,”Science, 306 (5702), 1776–80.
Kahneman, Daniel, Peter P. Wakker, and Rakesh Sarin (1997),
“Back to Bentham? Explorations of Experienced Utility,”
Quarterly Journal of Economics, 112 (2), 375–405.
Kang, Min Jeong, Ming Hsu, Ian M. Krajbich, George Loewenstein,
Samuel M. McClure, Joseph Tao-yi Wang, and Colin F. Camerer
(2009), “The Wick in the Candle of Learning: Epistemic Curiosity
Activates Reward Circuitry and Enhances Memory,”Psycho-
logical Science, 20 (8), 963–73.
Kapp, Karl M. (2012), The Gamification of Learning and In-
struction: Game-Based Methods and Strategies for Training and
Education. San Francisco: John Wiley & Sons.
Kashdan, Todd B., and Paul J. Silvia (2009), “Curiosity and Interest:
The Benefits of Thriving on Novelty and Challenge,”in Oxford
Handbook of Positive Psychology, S.J. Lopez, ed. Oxford, UK:
Oxford University Press, 367–75.
Kashdan, Todd B., and Michael F. Steger (2007), “Curiosity and
Pathways to Well-Being and Meaning in Life: Traits, States, and
Everyday Behaviors,”Motivation and Emotion, 31 (3), 159–73.
Knobloch, Silvia, Grit Patzig, Anna-Maria Mende, and Matthias
Hastall (2004), “Affective News Effects of Discourse Structure in
Narratives on Suspense, Curiosity, and Enjoyment While Reading
News and Novels,”Communication Research,31(3),259–87.
Knobloch-Westerwick, Silvia, and Caterina Keplinger (2006),
“Mystery Appeal: Effects of Uncertainty and Resolution on the
Enjoyment of Mystery,”Media Psychology, 8 (3), 193–212.
Kruger, Justin, and Matt Evans (2009), “The Paradox of Alypius and
the Pursuit of Unwanted Information,”Journal of Experimental
Social Psychology, 45 (6), 1173–79.
Kupor, Daniella, and Zakary L. Tormala (2015), “Persuasion,
Interrupted: The Effect of Momentary Interruptions on Message
Processing and Persuasion,”Journal of Consumer Research,
42 (2), 300–15.
Laran, Juliano, and Michael Tsiros (2013), “An Investigation of the
Effectiveness of Uncertainty in Marketing Promotions Involving
Free Gifts,”Journal of Marketing, 77 (2), 112–23.
Leavitt, Jonathan D., and Nicholas J.S. Christenfeld (2011), “Story
Spoilers Don’t Spoil Stories,”Psychological Science, 22 (9),
1152–54.
Lee, Yih Hwai, and Cheng Qiu (2009), “When Uncertainty Brings
Pleasure: The Role of Prospect Imageability and Mental Imag-
ery,”Journal of Consumer Research, 36 (4), 624–33.
Leuba, Clarence (1955), “Toward Some Integration of Learning
Theories: The Concept of Optimal Stimulation,”Psychological
Reports,1,27–33.
Litman, Jordan (2005), “Curiosity and the Pleasures of Learning:
Wanting and Liking New Information,”Cognition and Emotion,
19 (6), 793–814.
Litman, Jordan, and Tiffany L. Jimerson (2004), “The Measurement
of Curiosity as a Feeling of Deprivation,”Journal of Personality
Assessment, 82 (2), 147–57.
Litman, Jordan, and Paul J. Silvia (2006), “The Latent Structure of
Trait Curiosity: Evidence for Interest and Deprivation Curiosity
Dimensions,”Journal of Personality Assessment, 86 (3), 318–28.
Loewenstein, George (1994), “The Psychology of Curiosity: A
Review and Reinterpretation,”Psychological Bulletin, 116 (1),
75–98.
Mandel, Naomi, and Stephen M. Nowlis (2008), “The Effect of
Making a Prediction About the Outcome of a Consumption
Experience on the Enjoyment of That Experience,”Journal of
Consumer Research, 35 (1), 9–20.
Maner, Jon K., and Mary A. Gerend (2007), “Motivationally Se-
lective Risk Judgments: Do Fear and Curiosity Boost the Boons or
the Banes?”Organizational Behavior and Human Decision
Processes, 103 (2), 256–67.
Marvin, Caroline B., and Daphna Shohamy (2016), “Curiosity and
Reward: Valence Predicts Choice and Information Prediction
Errors Enhance Learning,”Journal of Experimental Psychology.
General, 145 (3), 266–72.
McShane, Blakeley B., and Ulf B¨ockenholt (2017), “Single Paper
Meta-Analysis: Benefits for Study Summary, Theory-Testing,
and Replicability,”Journal of Consumer Research, 43 (6),
1048–63.
Melis, Maria Rosaria, and Antonio Argiolas (1995), “Dopamine and
Sexual Behavior,”Neuroscience and Biobehavioral Reviews,
19 (1), 19–38.
Menon, Satya, and Dilip Soman (2002), “Managing the Power of
Curiosity for Effective Web Advertising Strategies,”Journal of
Advertising, 31 (3), 1–14.
Nelson, Leif D., Tom Meyvis, and Jeff Galak (2009), “Enhancing
the Television-Viewing Experience Through Commercial In-
terruptions,”Journal of Consumer Research, 36 (2), 160–72.
Newman, Lucile F. (1990), Hunger in History: Food Shortage,
Poverty, and Deprivation. Cambridge, MA: Blackwell.
Noordewier, Marret K., and Eric van Dijk (2015), “Curiosity and
Time: From Not Knowing to Almost Knowing,”Cognition and
Emotion, 31 (3), 1–11.
Nowlis, Stephen M., Naomi Mandel, and Deborah Brown McCabe
(2004), “The Effect of a Delay Between Choice and Consumption
on Consumption Enjoyment,”Journal of Consumer Research,
31 (3), 502–10.
Ogilvy, David (1983), Ogilvy on Advertising. New York: Crown.
Perry, Samuel W., Lawrence B. Jacobsberg, Baruch Fishman,
Pamela H. Weiler, Jonathan W.M. Gold, and Allen J. Frances
(1990), “Psychological Responses to Serological Testing for
HIV,”AIDS, 4 (2), 145–52.
Pham, Michel Tuan, Maggie Geuens, and Patrick De Pelsmacker
(2013), “The Influence of Ad-Evoked Feelings on Brand Eval-
uations: Empirical Generalizations from Consumer Responses to
More Than 1000 TV Commercials,”International Journal of
Research in Marketing, 30 (4), 383–94.
Polman, Evan, Rachel L. Ruttan, and Joann Peck (2017), “Curiosity
and Want/Should Conflict,”working paper, University of
Wisconsin–Madison.
Seaborn, Katie, and Deborah I. Fels (2015), “Gamification in Theory
and Action: A Survey,”International Journal of Human-Computer
Studies,74,14–31.
The Teasing Effect 569

Shani, Yaniv, Orit E. Tykocinski, and Marcel Zeelenberg (2008),
“When Ignorance Is Not Bliss: How Distressing Uncertainty
Drives the Search of Negative Information,”Journal of Economic
Psychology, 29 (5), 643–53.
Shani, Yaniv, and M. Zeelenberg (2007), “When and Why Do We
Want to Know? How Experienced Regret Promotes Post-
Decision Information Search,”Journal of Behavioral Decision
Making, 20 (3), 207–22.
Shen, Luxi, Ayelet Fishbach, and Christopher K. Hsee (2015), “The
Motivating-Uncertainty Effect: Uncertainty Increases Resource
Investment in the Process of Reward Pursuit,”Journal of Con-
sumer Research, 41 (5), 1301–15.
Shen, Luxi, Christopher K. Hsee, and Joachin Talloen (2018), “The
Fun and Function of Uncertainty: Uncertain Incentives Reinforce
Behavioral Repetition,”Journal of Consumer Research, pub-
lished electronically, https://doi.org/10.1093/jcr/ucy062.
Sieff, Elaine M., Robyn M. Dawes, and George Loewenstein (1999),
“Anticipated Versus Actual Reaction to HIV Test Results,”
American Journal of Psychology, 112 (2), 297–311.
Silvia, Paul J. (2008), “Interest: The Curious Emotion,”Current
Directions in Psychological Science, 17 (1), 57–60.
Silvia, Paul J., and Todd B. Kashdan (2009), “Interesting Things and
Curious People: Exploration and Engagement as Transient States
and Enduring Strengths,”Social and Personality Psychology
Compass, 3 (5), 785–97.
Simonsohn, Uri, Leif D. Nelson, and Joseph P. Simmons (2014),
“P-Curve: A Key to the File-Drawer,”Journal of Experimental
Psychology. General, 143 (2), 534–47.
Spielberger, Charles D., and Laura M. Starr (1994), “Curiosity and
Exploratory Behavior,”in Motivation: Theory and Research,
H.F. O’Neil and M. Drillings Jr., eds. Hillsdale, NJ: Lawrence
Erlbaum Associates, 221–43.
Stewart, David W., and David H. Furse (1986), Effective Television
Advertising. Lexington, MA: Lexington Books.
Strotz, Robert Henry (1955), “Myopia and Inconsistency in Dynamic
Utility Maximization,”Review of Economic Studies,23(3),165–80.
Swasy, John L., and Arno J. Rethans (1986), “Knowledge Effects on
Curiosity and New Product Advertising,”Journal of Advertising,
15 (4), 28–34.
Thaler, Richard H., Amos Tversky, Daniel Kahneman, and Alan
Schwartz (1997), “The Effect of Myopia and Loss Aversion on
RiskTaking:AnExperimentalTest,”Quarterly Journal of Economics,
112 (2), 647–61.
Tu, Yanping, and Christopher K. Hsee (2018), “Hedonomics:
On Subtle yet Significant Determinants of Happiness,”in
E-Handbook of Subjective Well-Being, E. Diener, S. Oishi, and
L. Tay, eds. NobaScholar, https://www.nobascholar.com/books/1.
Van den Bos, Kees (2001), “Uncertainty Management: The Influ-
ence of Uncertainty Salience on Reactions to Perceived Pro-
cedural Fairness,”Journal of Personality and Social Psychology,
80 (6), 931–41.
Van Dijk, Eric, and Marcel Zeelenberg (2007), “When Curiosity
Killed Regret: Avoiding or Seeking the Unknown in Decision-
Making Under Uncertainty,”Journal of Experimental Social
Psychology, 43 (4), 656–62.
Welker, W.I. (1956), “Some Determinants of Play and Exploration
in Chimpanzees,”Journal of Comparative and Physiological
Psychology, 49 (1), 84–89.
Whitchurch, Erin R., Timothy D. Wilson, and Daniel T. Gilbert
(2011), “‘He Loves Me, He Loves Me Not...’Uncertainty Can
Increase Romantic Attraction,”Psychological Science,22(2),
172–75.
Wilson, Timothy D., David B. Centerbar, Deborah A. Kermer, and
Daniel T. Gilbert (2005), “The Pleasures of Uncertainty: Pro-
longing Positive Moods in Ways People Do Not Anticipate,”
Journal of Personality and Social Psychology, 88 (1), 5–21.
Wu, George (1999), “Anxiety and Decision Making with Delayed
Resolution of Uncertainty,”Theory and Decision, 46 (2), 159–98.
Yang, Yang, Yangjie Gu, and Jeff Galak (2017), “When It Could
Have Been Worse, It Gets Better: How Favorable Uncertainty
Resolution Slows Hedonic Adaptation,”Journal of Consumer
Research, 43 (5), 747–68.
570 JOURNAL OF MARKETING RESEARCH, AUGUST 2018