What Makes a Super Bowl Ad Super? Five-Act Dramatic Form Affects Consumer Super Bowl Advertising Ratings

Abstract

Grounded in dramatic theory, this study presents an expanded model of dramatic form for empirical investigation of television commercials. A two-year content analysis of 108 commercials found significant association of dramatic form with consumer favorability in Super Bowl advertising rating polls. Results demonstrated that average consumer ratings were higher for commercials that followed a five-act dramatic form and a positive association of the number of acts in commercials with consumer favorability ratings. The paper discusses the relationship of five-act plot development with advertising function as well as the theoretical implications of narrative theory and its impact on consumer response.

Full text

This is an Accepted Manuscript of an article published by Taylor & Francis in [Journal of
Marketing Theory and Practice] on [December 7, 2014], available online:
https://doi.org/10.2753/MTP1069-6679220406
What Makes a Super Bowl Ad Super? Five-Act Dramatic
Form Affects Consumer Super Bowl Advertising Ratings
ABSTRACT
Grounded in dramatic theory, this study presents an expanded model of dramatic form for
empirical investigation of television commercials. A two-year content analysis of 108
commercials found significant association of dramatic form with consumer favorability in Super
Bowl advertising rating polls. Results demonstrated that average consumer ratings were higher
for commercials that followed a five-act dramatic form and a positive association of the number
of acts in commercials with consumer favorability ratings. The paper discusses the relationship
of five-act plot development with advertising function as well as the theoretical implications of
narrative theory and its impact on consumer response.
Keith A. Quesenberry (M.S. IMC, West Virginia University), Lecturer in the Center for
Leadership Education, Johns Hopkins University, Baltimore, kquesenberry@jhu.edu
Michael K. Coolsen (Ph.D., University of North Carolina at Chapel Hill), Associate Professor of
Marketing, Shippensburg University, Shippensburg, mkcool@ship.edu
The authors wish to express their appreciation to Heather Bilinski and Katherine Freeman for
their assistance in coding the data in this study.

“If the television commercial could be shown to be drama, it would be among the
most ubiquitous and the most influential of its forms and hence deserve the
attention of the serious critics and theoreticians of that art (Esslin 1979 p. 96).”
Barbara Stern (1994) appropriately began her study on drama in television advertising
with the Martin Esslin quote above. Drama is defined as a composition intended to portray life or
character or to tell a story usually involving conflicts and emotions through action and dialogue
(Merriam-Webster 2013). The presence of drama has been noted and studied in a variety of
approaches as applied to marketing, advertising, and consumer research. Esslin (1979) first
considered the television commercial as a form of drama by demonstrating that television
commercials exhibit the basic characteristics of drama including: authority figures, allegorical
characters, and reliance on character and imagery. The study of drama in television commercials
is rooted in narrative theory, which seeks to explain the essential role that stories play in our
lives. This study sought to continue this valuable investigation into the value of narrative theory
in its ability to understand, describe, explain, and predict consumer behavior (Grayson 1997).
Wells (1988) and Deighton, Romer, and McQueen (1989) began studying commercial
forms based on a dichotomy between drama and lecture, highlighting and defining a three-point
drama scale as a commercial that has character and a plot but does not have a narrator. From that
typology, the Deighton, Romer, and McQueen study (1989) coded 40 television commercials by
argument, drama, and hybrid. Later, much of the research examining commercials as dramas
followed the same dichotomy model of drama versus argument (see Boller 1990; Boller and
Olson 1991; Stern 1990, 1991, 1992). Differentiation was based upon dramatic variables such as
narration, plot, and character, and research demonstrated the importance of advertising form in
eliciting audience response.

In these various studies, the presence of drama in television commercials has been
associated with consumer response effects such as heightened emotion, diminished
counterargument, and enjoyment of the commercial (Deighton et al. 1989), increased inferencing
(Wells 1988), heightened mode of self-participatory processing (Boller 1990), and empathetic
and sympathetic response facilitating consumer favorability for the commercial (Stern 1994,
2003). Thus, advertising form has been shown to impact advertising function and consumer
favorability.
Although previous research has emphasized dramatic form, most advertising research has
identified drama based upon the three dramatic variables of narration, character, and plot (e.g.,
Boiler 1990; Deighton, Romer, and McQueen 1989; Stern and Escalas 2003). Deighton, Romer,
and McQueen (1989) first proposed the three-point “dramatization scale” to contrast argument
and drama saying that argument is plotless, characterless, and narrated. Thus, they restricted
drama to a commercial having character and plot but no narrator. This study accepts the
distinction between rational argument and emotional drama but contends that argument should
not be completely separated from character and plot. Argument is a discourse intended to
persuade that can just as easily be delivered by character within a coherent series of statements
leading to a conclusion that is directed by plot (Merriam-Webster 2013). In fact, character and
plot are used abundantly by marketers in commercials to make arguments and deliver the rational
benefits of their products or services. In a similar vein, narration does not exclusively deliver
rational argument. Narration is a tool that can just as easily be used to develop character and plot
and to evoke emotion and favorability.
Therefore, the aim of this study is to expand upon previous research by focusing more
precisely on plot, which encompasses the elements of narration and character as components to

tell a dramatic story. Many plot devices are delivered via character or narration in order to move
a story forward. Therefore, instead of following the Deighton, Romer, and McQueen (1989)
model of coding television commercials for narrated argument with no character and no plot
versus drama with character and plot, but no narration, this article proposes a new model. This
study instead treats narration and character as plot devices used to advance a story, which in turn
becomes the advertising form designed to develop an argument and to gain consumer
favorability.
Accordingly, this paper begins by tracing the roots of literary, drama, and film criticism
and advertising communication studies related to dramatic theory, subsequently presenting a
model for empirical analysis of advertising television commercials. Next, it discusses the specific
subset of television commercials that air within the environment of the NFL Super Bowl along
with related previous research. Then, it discusses the association of five-act commercial form
(plot development) with advertising function represented as favorable consumer response in
Super Bowl advertising ratings polls. Lastly, it discusses the theoretical implications of drama
theory and the use of narrative theory to understand the impact of narrative on consumer
behavior.
DRAMATIC THEORY
Aristotle’s Poetics, first written in 335 BC, is considered to be the earliest-surviving work
of dramatic theory – the basis of literary criticism. Aristotle considered plot to be the most
important element of drama – more important than character. In Poetics, he characterized plot as
having a beginning, middle, and end with events that must causally relate (Aristotle 2013). To
Aristotle, it is the plot that has the ability to arouse emotion and deliver drama. Character and
narration are devices used to advance the plot in a story. Thus, this research suggests that

character and narration should be not categorized as competing variables against plot as defined
by the Deighton, Romer, and, McQueen (1989) three-point “dramatization scale.” Alternatively,
this research suggests a new model that encompasses character and narration as development
tools under plot.
Classical Drama Model
To measure plot more precisely in television commercials and examine its association
with consumer favorability, this study turned to Freytag Gustav’s theory of dramatic structure.
This theory is derived from Gustav’s study of five-act dramatic structure first published as Die
Technik des Dramas in 1863 and frequently symbolized as “Freytag’s pyramid” (see Figure 1).
Freytag advanced Aristotle’s dramatic theory, based on a plot divided by a beginning, middle,
and end to a more precise five-act structure first developed by Renaissance dramatists such as
William Skakespeare. According to Freytag, a drama is divided into five parts called acts, and
these acts combine to form the following dramatic arc: exposition, rising action, climax, falling
action, and denouement.
Figure 1
Freytag’s Pyramid
Act 3: Climax
(Turning point)
Act 2: Complication Act 4: Reversal
(Rising action) (Falling action)
Act 1: Exposition Act 5: Denouement
(Inciting moment) (Moment of release)

In this model, exposition sets the stage by introducing character and setting, followed by
a series of events that builds up the story to a climax where the story (plot) takes a turning point
for better (comedy) or for worse (tragedy). Falling action is the series of events that moves the
plot towards a resolution to the story. Freytag’s study and analysis were based on five-act plays
but have been applied to novels demonstrating that dramatic structure is also a literary element
(Freytag 1863). Gustav Freytag's pyramid is seen as a powerful story tool for the most successful
fiction (Stern 2000), has become an underlying element of classic Hollywood narratives (Trbic
2010), and has even been suggested as a solid structure for captivating and convincing public
speech in the form of sermons (Willobee 2009). The five-act model is further explained below:
Act 1 – Exposition
Act 1 introduces characters, establishes time and setting, and explains past events and
circumstances. The basic conflict is revealed leading to dramatic tension between opposing
forces such as a protagonist (principle character) and an antagonist (opposing character). The
first act ends with an inciting moment, which sets the story in motion beginning with the rising
action of complications.
Act 2 – Complication
Here the story becomes more complex. Interests clash and plans are made as action rises
and tension mounts. The protagonist's effort to reach his or her goal is complicated by additional
conflicts and obstacles. Frustration builds as opposing forces and circumstances intensify.
Act 3 – Climax
The development of the conflict reaches a high point. This turning point for the
protagonist could lead to victory or defeat. A change has happened that is for better or for worse.

In a comedy, things have gone badly for the protagonist and will now go well for him or her. If it
is a tragedy, things will go from good to bad for the protagonist.
Act 4 – Reversal
During Act 4, the consequences of the climatic turning point play out. This falling action
is a result of the reversal that occurred after the climax when the protagonist either won or lost
against the antagonist. Momentum slows as the characters head toward a final resolution. In a
comedy, it looks like all can be saved. In a tragedy, it looks like all may be lost.
Act 5 – Denouement
The conflict is resolved creating normalcy for the characters and a sense of release of
tension and anxiety. The comedy ends with the protagonist being better off than at the outset of
the story. On the other hand, the tragedy ends in catastrophe with the protagonist being worse
off.
Super Bowl of Advertising
To apply Freytag Gustav’s theory of dramatic structure in advertising television
commercials and investigate its association with consumer favorability, this research chose the
“stage” of the Super Bowl due to its movie-like viewing experience, its importance to marketers,
and the availability of data that directly measures consumer favorability in the form of national
Super Bowl advertising rating polls. Super Bowl television advertising has been studied before
(see Pavelchak, Antil, and Munch 1988; Xinshu 1997; Tao et al. 2001; Jin 2003/2004, Jin, Zhao,
and Soontae 2006; Nail 2007; Siefert et al. 2009), but research has not examined consumer
ratings of television commercials in Super Bowl advertising polls or correlated Super Bowl
advertising rating polls to dramatic theory.

How big is the “stage” of Super Bowl advertising? A record 106.5 million people
watched Super Bowl XLIV – one of the Super Bowl years of this study. An audience size that
large is rare in today’s fragmented media environment. Primetime broadcast TV viewership has
dropped by 50% since 1996 (Klaassen 2006). However, the costs of capturing a mass audience
this size are high. It cost $2.5 million to $3 million to purchase a 30-second commercial on Super
Bowl XLIV (Flint 2010). In comparison, a 30-second commercial on ABC's top-rated "Grey's
Anatomy" the same broadcast season only cost $228,851 to $240,462 (Steinberg 2009).
Additionally, production of 30-second national TV commercials costs an average of $350,000
(gaebler.com 2010) with elaborate productions costing up to $1 million or more (PR Leap 2010).
Nevertheless, for the high cost, Super Bowl commercials can generate considerable buzz
in traditional media and social media via word-of-mouth. In fact, 57% of U.S. adults who plan to
watch the Super Bowl tune in as much for the commercials as for the game (“Super Bowl Ads
Take Equal Billing” 2008). Studies also show that after a Super Bowl commercial airs, the
number of tweets about a brand jumps significantly (Kincaid 2009) along with brand-focused
blog posts (McClellan 2010). Because of the high marketing costs, the attention of Super Bowl
commercial ratings polls, Super Bowl commercial pre- and post- game Super Bowl commercial
TV show specials, and the proliferation of social media sharing, the advertising industry has
dubbed this event as the “Super Bowl of Advertising.”
Do consumers respond more favorably to Super Bowl television commercials that fully
develop plot or story based on Freytag Gustav’s theory of five-act dramatic structure? Drawing
from previous research that has linked the presence of drama in television commercials and
favorable consumer response effects (Boller 1990; Deighton et al. 1989; Stern 1994, 2003; Wells

1988), this study employed the advertising research practice of content analysis along with the
data from Super Bowl commercial rating polls to investigate this question.
RESEARCH QUESTIONS
An exploratory study was designed to investigate the potential association of five-act
form based on Freytag Gustav’s theory of dramatic structure with consumer favorability
measures in Super Bowl commercial ratings polls. Research questions were stated as follows:
RQ1: Do Super Bowl commercials that follow the full five-act dramatic form have
significantly higher consumer Super Bowl commercial poll ratings than commercials that do not
follow the full five-act form?
RQ2: Is there an additive effect of the number of individual acts in a commercial on the
consumer Super Bowl commercial poll ratings?
The first research question was a straight analysis of five-act commercials versus non-
five-act commercials related to their respective scores in consumer Super Bowl commercial
ratings polls. Beyond that, a subsequent analysis was conducted to determine if there was a linear
relationship between the number of dramatic form acts in a commercial and its corresponding
consumer Super Bowl commercial poll ratings. For example, do Super Bowl commercials that
have three of the five acts (e.g., the ad has an exposition, a complication, and a climax but not a
reversal or a denouement) rate higher than ads that have only two of the five-acts (e.g., an
exposition and a complication)? The second research question examined the first in finer detail,
expanding the categorization of commercials without a full five-act form. The measurement of
acts was designed to determine plot development or how complete a story was told in each
commercial. Would more acts correlate with higher ratings? Do well developed stories increase
favorability more than less developed stories?

Prior empirical evidence provided by Escalas and Stern (2003) sets the theoretical basis
for the two research questions mentioned above. Citing research that more developed stories
(compared to less developed stories) elicit higher levels of positive emotions (Edell and Berk
1987), Escalas and Stern reported that classical drama in advertisements, which is characterized
by a linear plot with resolution, causality, interaction, and change involving characters, elicits
increased favorability in consumer advertisement attitudes compared to advertisement vignettes,
which are characterized by unconnected episodes and characters that do not interact or change.
The authors offer a sympathy-empathy model in order to explain why classical dramas in
advertising increases consumer favorability: classical dramas are better than vignettes at
generating both sympathy because it takes well-developed characters and a linear plot to evoke
sympathy (“I understand your pain”) and empathy (“I feel your pain”). Such sympathy and
empathy of the characters subsequently transports viewers into the drama being played out in the
advertisement and ultimately elicits pleasure and enjoyment of the advertisement.
Using Freytag Gustav’s theory of dramatic structure with five-act form, the research
questions in this investigation seek to refine and extend previous research that used a three-point
“dramatization scale” to study the consumer response effects of drama in television commercials.
As previously mentioned, Gustav’s five-act structure provides a more comprehensive model for
achieving a linear plot, encompassing character and narration as development tools under plot.
Additionally, one can readily see the important development of emotional response in Gustav’s
five-act structure as well: conflict and tension rise in the first two acts, crest in the third act, and
is alleviated in the last two acts. Because classical drama has been associated with sympathy and
empathy regarding the character’s emotions (Escalas and Stern 2003), the emotional response

developed in Gustav’s five-act structure would seem to be fully transferred to the viewer of the
advertisement.
This investigation can be illustrated by looking at two exemplar commercials. In “Fence”
from Anheuser-Bush (aired during the 2010 Super Bowl), act one occurs as the scene is set with
two fields next to each other. The characters are introduced as a young Clydesdale and a young
calf. The exposition is established in the second act when complications arise. Running together,
the two friends first realize a fence separates them. Act three is established when years later the
grown Clydesdale runs by the field pulling the Budweiser wagon. The grown steer starts running
and catches up to him. The turning point happens as he approaches the fence. The fourth act (the
reversal) comes to life when instead of stopping; the steer breaks through the fence. This leads to
the final act where tension is released and the old friendship is returned to normal. Following
Aristotle and Freytag Gustav’s dramatic theory, the plot is fully developed in five acts.
In contrast, “Ice Bottle” from Anheuser-Bush (also aired during the 2010 Super Bowl),
shows a bottle of Budweiser Select 55 spinning in the air as the voice-over describes how it is the
lightest beer in the world. The commercial does not introduce characters or conflict, conversely
act one is never established. No rising action, complications, turning point, falling action or
release of tension occurs. Freytag Gustav’s dramatic arc never forms and a story does not
develop. Some may say that the plot never “thickens.”
METHOD
Sample
This study involved a multi-step content analysis of Super Bowl XLIV and Super Bowl
XLV television commercials. Commercials were recorded from the 2010 and 2011 network
broadcast of the game: (1) commercials from the 2010 network broadcast of the game were

analyzed as a pretest and (2) commercials from the 2011 network broadcast of the game were
analyzed as a confirmatory test of 2010 commercial findings. The sample was limited to
commercials that paid premium media costs to be broadcast nationally during the official four
quarters of the game. This precondition followed the standards established by national Super
Bowl commercial ratings polls, which made up the data set. Pre-game, post-game and non-
national commercials that only aired locally during the affiliate station breaks were not included.
Movie trailer ads were also excluded from the study because there was no variance in their form.
They followed the same industry formula of movie clips, music and voiceover. This precondition
followed the standard of the SpotBowl.com ratings poll used in this study, which excludes movie
trailer ads for the same reason. A total of 59 commercials met the requirements for the 2010
Super Bowl and 49 commercials met the requirements in 2011. Subsequently 108 Super Bowl
commercials were analyzed in this study.
Consumer Favorability Polls
The consumer favorability ratings of the 2010 and 2011 Super Bowl television
commercials in the sample were collected from voting results of the SpotBowl.com and USA
Today Ad Meter national ratings polls. Active for ten years, SpotBowl.com is a live national
online Super Bowl commercial poll. SpotBowl.com drew over 35,891 votes in 2010 and 35,120
votes in 2011 from all 50 states. Participants were able to view and vote on their favorite
commercials from kickoff time (6:30 p.m. EST, on Sunday February 7 in 2010 and Sunday,
February 6 in 2011) until 3:00 p.m. the following Monday. They were given instructions on how
to vote by ranking ads “from one helmet (a fumble) to five helmets (a touchdown).” Thus, a five-
point categorical scale was employed (1 = “a fumble”; 5 = “a touchdown”). All voting occurred

on the Spotbowl.com website through computers or Web enabled phones where consumers could
watch the commercials again (Spotbowl.com website 2010).
The second poll utilized was the annual USA Today Ad Meter, which tracked live
consumer responses to commercials during the Super Bowl from focus group panels of: (1) 250
adult volunteers in San Diego, CA and McLean, VA in 2010, and (2) 282 adult volunteers in
Bakersfield, CA and McLean, VA in 2011. Ad Meter used handheld meters to register how much
participants liked each commercial using a zero-to-ten scale with “zero being the worst, ten the
best” (USA Today Ad Meter website 2010).
Coding
After reviewing literature on previous coding procedures used in advertising content
analysis research, it was determined that this study would employ a multi-step method for coding
the Super Bowl commercials. The multi-step coding process has been used successfully in
multiple published studies (e.g., Bailey 2006; Gagnard 1989; Kim, McMillian, and Hwang 2005;
Weinberger and Spots 1989). Two of these studies used a third judge/researcher. For example,
Weinberger and Spots (1989) studied humor in television commercials with coding by two
independent judges. Then discrepancies in coding were resolved by a third judge to improve
intercoder reliability. Similarly, Kim, McMillian, and Hwang (2005) studied Super Bowl
advertising with two of the authors acting as primary coders. Disagreements between the primary
coders were resolved with the third author acting as a judge to help them come to agreement.
This study followed closely these previously accepted methods. Two independent coders
(communications graduate students) were recruited. One had previous coding experience and the
other practical marketing experience. Both were familiar with dramatic or narrative form as the
one coder was a communications major (for both undergraduate and graduate studies) and the

other coder was both a communications major (for graduate studies) and a creative writing major
(for both undergraduate and graduate studies).
The two research assistants met with the authors and were trained in the definitions and
guidelines for coding five-act dramatic form using seven Super Bowl commercials from 2009
(see Appendix I and Appendix II for coding form and coder guidelines). The coders viewed each
commercial several times and used the coding form to mark the existence of each of the five acts
following Freytag Gustav’s dramatic theory commonly expressed as “Freytag’s Pyramid.” To
reduce subjectivity, the judges were instructed not to indicate whether they personally liked the
commercials, but to only code for dramatic form characteristics. The authors and coders then
discussed the 2009 codings to produce final agreement for the set of training commercials.
After the training meeting, the two coders independently coded 59 Super Bowl XLIV and
49 Super Bowl XLV television commercials (i.e., the pretest and confirmatory test samples
analyzed in this study). Upon completion of these sets of independent codings, significant
discrepancies were clarified and resolved by a third author judge to ensure reliable coding data
and acceptable intercoder reliability numbers. One of the authors was used as this third judge
because of his familiarity and professional experience in advertising dramatic form. Coding and
analysis of the 59 Super Bowl commercials from 2010 was treated as a pretest, and the coding
and analysis of the 49 Super Bowl commercials from 2011 were conducted separately as a
confirmatory test. In all, 108 Super Bowl commercials were coded and analyzed.
Measurement scales (0-8) on the coding form (see Appendix I) were translated into
dichotomous variables by indicating anything above a level of 0 as development of that act and
thus being present. In other words, if anything was indicated above a level of 0, then that act was
considered to be present. The data for each act was in binary form – the number of acts for each

ad was added to form a number of acts score for each ad. Our training procedure for the coders
focused exclusively on the presence versus absence of each act, but did not attend to the
establishment score on the coding form. The 0-8 scale was created simply to be an additional
coder check for the presence versus absence of an act after the coding sheets were completed.
Intercoder Reliability
Kang et al (1993) studied the various methods of assessing intercoder agreement used in
advertising content analysis. The authors identified the major issues in the choice of a measure of
intercoder agreement as: 1) sensitivity to systematic coding errors, 2) correction for chance
agreements, 3) ability to handle multiple coders, and 4) the level of measurement represented by
the stimuli or objects of interest. Based on the authors’ analysis, Cohen’s Kappa, Scott's Pi and
Krippendorff's Alpha were the top three recommended intercoder agreement indices because
they correct for chance agreement and handle a nominal scale appropriate for assessing
categorical variables. All three of the recommended indices were chosen and used for this study.
After coding for plot development by determining the existence or absence of each of the
five acts for each commercial, intercoder reliability estimates were generated using Cohen’s
Kappa, Scott's Pi and Krippendorff's Alpha. Reliabilities were calculated across all acts and by
individual act (1 – 5) for both the 2010 pretest commercials and the 2011 confirmatory test
commercials. The levels of agreement are reported in Table 1 with estimates generally within
accepted levels for coding content analysis. Landis and Koch (1977) suggest that values of
agreement .41 to .60 are moderate, .61 to .80 are substantial and .81 to 1.00 are almost perfect
agreement. Fleiss (1981) indicates that Cohen’s Kappa’s from .40 to .75 are fair to good and over
.75 are excellent. Thus, intercoder reliability was determined to be acceptable for data analysis.

Table 1
Intercoder Reliabilities
Acts
Cohen’s Kappa
2010 2011
Scott’s Pi
2010 2011
Krippendorff’s Alpha
2010 2011
All Acts
.88 .93
.88 .93
.88 .93
Act 1
.67 .79
.67 .79
.67 .79
Act 2
.71 .95
.70 .95
.71 .95
Act 3
.93 .96
.93 .96
.93 .96
Act 4
1.00 1.00
1.00 1.00
1.00 1.00
Act 5
.92 .83
.92 .83
.92 .83
FINDINGS
Pretest and confirmatory test analysis was conducted for each of the two consumer Super
Bowl favorability ratings polls as well as for a composite score that combined the two consumer
Super Bowl favorability ratings polls to form an aggregate measure – to compute this aggregate
poll measure, scores from each consumer favorability poll were standardized and then averaged.
Dramatic Form Act Codings
Combining each set of codings from the two independent coders, the combined dataset
contained 38 five-act ads, 6 four-act ads, 8 three-act ads, 24 two-act ads, 20 one-act ads and 22
zero-act ads in 2010. In 2011, the combined dataset contained 38 five-act ads, 10 four-act ads, 13
three-act ads, 20 two-act ads, 12 one-act ads and 17 zero-act ads. Each ad was placed exclusively
in one of six categories from having zero acts to having all five acts represented in the ad;
therefore, each category represents the last act executed in the ad based on all acts before that
being necessarily present in that ad.

RQ1 Analysis: Pretest on 2010 Commercials
The RQ1 analysis tested the difference in consumer 2010 Super Bowl commercial poll
ratings between commercials that were coded as having the full five-act dramatic form and
commercials that were coded as not having the full five-act dramatic form. Figure 2 displays the
results of independent samples t-tests conducted to investigate this first research question. In all
figures, the average consumer favorability rating for the ads with a five-act form was significantly
higher than the average rating for the ads without a five-act form (p < .01 for all t-tests).
Figure 2
RQ1 Analysis: Pretest (2010 Ads)
2010 SpotBowl.com Mean Ad Favorability Rating was Significantly Higher
in Ads with Full Five-Act Form Compared to Ads without a Full Five-Act Form
Note: t = -8.34, p < .01.
2010 USA Today Mean Ad Favorability Rating was Significantly Higher
in Ads with Full Five-Act Form Compared to Ads without a Full Five-Act Form
Note: t = -5.46, p < .01.
2.44
3.4
1
2
3
4
5
Ads Without Five-Act Structure Ads With Five-Act Structure
5.88 6.83
0
2
4
6
8
10
Ads Without Five-Act Structure Ads With Five-Act Structure

2010 Aggregate Measure Mean Ad Favorability Rating was Significantly Higher
in Ads with Full Five-Act Form Compared to Ads without a Full Five-Act Form
Note: t = -8.21, p < .01.
RQ1 Analysis: Confirmatory Test on 2011 Commercials
The findings for the RQ1 analysis of 2010 pretest commercials were replicated in RQ1
analysis of the 2011 confirmatory test commercials: the average consumer favorability rating for
the commercials with a five-act form was significantly higher than the average rating for the
commercials without a five-act form (p < .01 for all t-tests). Figure 3 displays the results of
independent samples t-tests.
Figure 3
RQ1 Analysis: Confirmatory Test (2011 Ads)
2011 SpotBowl.com Mean Ad Favorability Rating was Significantly Higher
in Ads with Full Five-Act Form Compared to Ads without a Full Five-Act Form
Note: t = -5.84, p < .01.
-0.53
0.68
-1
-0.5
0
0.5
1
Ads With Five-Act Structure
2.62
3.27
1
1.5
2
2.5
3
3.5
4
4.5
5
Ads Without Five-Act Structure Ads With Five-Act Structure
Ads Without Five-Act Structure

2011 USA Today Mean Ad Favorability Rating was Significantly Higher
in Ads with Full Five-Act Form Compared to Ads without a Full Five-Act Form
Note: t = -7.13, p < .01.
2011 Aggregate Measure Mean Ad Favorability Rating was Significantly Higher
in Ads with Full Five-Act Form Compared to Ads without a Full Five-Act Form
Note: t = -7.38, p < .01.
RQ2 Analysis: Pretest on 2010 Commercials
The preliminary analysis was admittedly simplistic. Do Super Bowl commercials with a
five-act dramatic form rate higher in consumer favorability than commercials without a five-act
form? Significant support in the pretest and confirmatory test was demonstrated for RQ1. Now
the additive effect of each act on consumer favorability poll ratings was investigated. The two
independent judges coded commercials by each act. Was there was a linear relationship between
the total number of acts in a Super Bowl commercial and the corresponding consumer Super
6.07
7.2
0
1
2
3
4
5
6
7
8
9
10
Ads Without Five-Act Structure Ads With Five-Act Structure
-0.28
0.78
-1
-0.5
0
0.5
1
Ads With Five-Act Structure
Ads Without Five-Act Structure

Bowl commercial poll rating? To conduct the RQ2 analysis, a variable was created to represent
the sum total number of acts that were coded to exist in each Super Bowl commercial.
Consumer Super Bowl commercial poll ratings were regressed on the total number of
acts in each commercial. Linear regression analysis was conducted using the total number of acts
variable to predict consumer Super Bowl commercial poll ratings. The linear regression models
using the 2010 pretest commercials accounted for significant proportions of the variance in all
three sets of consumer Super Bowl commercial poll ratings: R2 = .42, p < .01 for SpotBowl.com;
R2 = .40, p < .01 for USA Today; and R2 = .51, p < .01 for the aggregate measure. This revealed
significantly strong positive associations of total number of acts in each commercial with
consumer Super Bowl commercial poll ratings: β = .65, p < .01 for SpotBowl.com; β = .63, p <
.01 for USA Today; and β = .71, p < .01 for the aggregate measure.
An alternative test of RQ2 was then conducted by running a one-way ANOVA among the
2010 pretest commercials with the total number of acts in each commercial as the predictor
variable and the consumer favorability rating as the outcome variable. Each model proved to be
significant across the two 2010 pretest consumer favorability rating indices (p < .01 for all F-
tests). Figure 4 displays these findings. It should be noted that there is a slight deviation in the
2010 USA Today ratings (Figure 4) where there is a slight dip from a higher mean in four acts
(m=6.88) to five acts (m=6.82), but the rising tendency is still apparent overall in both ratings
polls over two years. Figure 4 displays these findings.

Figure 4
RQ2 Analysis: Pretest (2010 Ads)
2010 SpotBowl.com Mean Ad Favorability Rating Increased
as the Number of Acts Present in The Ad Increased
Note: F [5, 112] = 18.70, p < .01.
2010 USA Today Mean Ad Favorability Rating Increased
as the Number of Acts Present in The Ad Increased
Note: F [5, 106] = 15.78, p < .01.
2.08 2.45 2.57 2.64 2.94
3.4
1
1.5
2
2.5
3
3.5
4
4.5
5
Zero Acts One Act Two Acts Three Acts Four Acts Five Acts
5.19 5.63 6.02 6.78 6.88 6.82
0
2
4
6
8
10
Zero Acts One Act Two Acts Three Acts Four Acts Five Acts

2010 Aggregate Measure of Mean Ad Favorability Rating Increased
as the Number of Acts Present in The Ad Increased
Note: F [5, 112] = 25.29, p < .01.
RQ2 Analysis: Confirmatory Test on 2011 Commercials
The findings for the RQ2 analysis of 2010 pretest commercials were replicated in RQ2
analysis of 2011 confirmatory test commercials: the linear regression models using 2011
confirmatory commercials accounted for significant proportions of the variance in all three sets
of consumer Super Bowl commercial poll ratings: R2 = .40, p < .01 for SpotBowl.com; R2 = .37,
p < .01 for USA Today; and R2 = .45, p < .01 for the aggregate measure. These models revealed
significantly strong positive associations of total number of acts in each commercial with
consumer Super Bowl commercial poll ratings: β = .63, p < .01 for SpotBowl.com; β = .61, p <
.01 for USA Today; and β = .67, p < .01 for the aggregate measure. In addition, all alternative
one-way ANOVA tests of this second research question among the 2011 confirmatory test ads
proved to be significant. Figure 5 displays these findings. It should be noted that anomalies occur
in 2011 Spotbowl.com ratings (Figure 5) where one act (m=2.44) is the same as two acts
(m=2.44) and in USA Today ratings (Figure 5) where there is a slight drop from one act
-1.11
-0.67
-0.34
0.11
0.38
0.68
-1.2
-1
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
Three Acts Four Acts Five Acts
Zero Acts
One Act
Two Acts

(m=6.01) to two acts (m=5.69), but again the overall rising tendency is still apparent in both
ratings polls over two years. Figure 5 displays these findings.
Figure 5
RQ2 Analysis: Confirmatory Test (2011 Ads)
2011 SpotBowl.com Mean Ad Favorability Rating Increased
as the Number of Acts Present in The Ad Increased
Note: F [5, 92] = 13.33, p < .01.
2011 USA Today Mean Ad Favorability Rating Increased
as the Number of Acts Present in The Ad Increased
Note: F [5, 92] = 13.34, p < .01.
2.33 2.44 2.44 2.77
3.23 3.27
1
1.5
2
2.5
3
3.5
4
4.5
5
Zero Acts One Act Two Acts Three Acts Four Acts Five Acts
5.81 6.01 5.69 6.29 6.64 7.2
0
2
4
6
8
10
Zero Acts One Act Two Acts Three Acts Four Acts Five Acts

2011 Aggregate Measure of Mean Ad Favorability Rating Increased
as the Number of Acts Present in The Ad Increased
Note: F [5, 92] = 17.72, p < .01.
Accounting for Time Duration of Commercial as a Covariate
As with any research question worthy of investigation, a vast amount of covariate
variables can be offered as alternative influences that put the validity of a tested relationship in
doubt. Many such variables come to mind when investigating various relationships with
consumer favorability of commercials. Type of ad appeal (informational vs. emotional), type of
execution format (e.g., comparison, demonstration, testimonial), ad tonality (e.g., humor,
romance, sex appeal), creative tactics (e.g., animation, celebrities, jingles), advertising medium
(e.g., internet, TV, radio), and any previous exposure to or repetition of the advertisement are
just some possible covariates in advertising research.
One important covariate in this investigation concerns the duration of time for each
commercial—that is, dramatic five-act commercials in this study’s dataset may, in fact, be longer
in duration than commercials with less than five dramatic acts due to the simple fact that more
acts may need significantly more time to develop in a commercial.
The theoretical basis for the research questions offered here involves the elicitation of
sympathy and empathy in a dramatic form commercial—that is, a developed five-act linear plot
-0.62
-0.44 -0.6
-0.05
0.46
0.78
-1.2
-1
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
Four Acts Five Acts
Zero Acts
One Act
Two Acts
Three Acts

seems necessary to create sympathy and empathy in the viewer, which subsequently increases
the viewer’s pleasure and favorability for the commercial. Since full dramatic form involves the
accumulation of five acts, it is entirely plausible that the simple duration of time for the
commercial could be accounting for more unique variance in (and thus be more predictive of)
consumer Super Bowl commercial poll ratings compared to how many dramatic acts appear in
the commercial. Therefore, a simultaneous regression analysis was performed to assess the
unique contributions of both the total number of dramatic acts variable and the commercial time
duration variable in predicting consumer Super Bowl commercial poll ratings.
This simultaneous regression model using the 2010 pretest commercials accounted for
significant proportions of the variance in all three sets of consumer Super Bowl commercial poll
ratings: R2 = .50, p < .01 for SpotBowl.com; R2 = .43, p < .01 for USA Today; and R2 = .57, p <
.01 for the aggregate measure. While this model revealed significant positive associations of
commercial time duration with consumer Super Bowl commercial poll ratings (β = .30, p < .01
for SpotBowl.com; β = .18, p < .01 for USA Today; and β = .27, p < .01 for the aggregate
measure), the model revealed much stronger significant positive associations of the total number
of acts in each commercial with consumer Super Bowl commercial poll ratings: β = .58, p < .01
for SpotBowl.com; β = .59, p < .01 for USA Today; and β = .65, p < .01 for the aggregate
measure.
The simultaneous regression model using 2011 confirmatory commercials also accounted
for significant proportions of the variance in all three sets of consumer Super Bowl commercial
poll ratings: R2 = .40, p < .01 for SpotBowl.com; R2 = .40, p < .01 for USA Today; and R2 = .46,
p < .01 for the aggregate measure. However, unlike the model using 2010 pretest commercials,
this 2011 confirmatory commercials model revealed two nonsignificant associations and one

significant negative association of commercial time duration with consumer Super Bowl
commercial poll ratings (β = .09, ns for SpotBowl.com; β = -.20, p < .05 for USA Today; and β =
-.07, ns for the aggregate measure). Alternatively, and similar to the model using 2010 pretest
commercials, this 2011 confirmatory commercials model revealed very strong significant
positive associations of the total number of acts in each commercial with consumer Super Bowl
commercial poll ratings: β = .61, p < .01 for SpotBowl.com; β = .65, p < .01 for USA Today; and
β = .69, p < .01 for the aggregate measure.
At the very least (given the 2010 pretest commercials), the number of acts in a
commercial seems to have a much stronger unique effect size regarding its association with
consumer Super Bowl commercial poll ratings compared to the moderate effect size of the time
duration of the commercial. Interestingly enough, the stronger effect size for the number of acts
in a commercial was retained in the 2011 confirmatory commercial, whereas the smaller effect
sizes of the time duration of the commercial dropped down to nonsignificant levels or, in one
poll, flipped from a significant positive association to a significant negative association with
consumer Super Bowl commercial poll ratings. Taken altogether, these results seem to support
the validity of our investigated relationship of fully developed dramatic form structure in Super
Bowl commercials with consumer favorability in this study.
DISCUSSION
Overall, the results suggest that presence of a five-act dramatic form or plot development
expressed through Freytag Gustav’s dramatic theory does drive consumer response in the form
higher consumer favorability ratings across two separate Super Bowl commercial rating polls.
Results were consistent despite the two polls using different methods and samples (e.g., open
online voting vs. controlled geographic limited focus groups). Results indicated that Super Bowl

XLIV and Super Bowl XLV commercials with full plot development (full five-act dramatic
form) were rated higher on average by consumers than commercials with less story development
(less than five-acts). This effect of commercial form was shown to be significant on consumer
response (see Table 2 for an example).
Table 2
Examples of 2010 Super Bowl Commercial Classifications and Ratings
Acts
Sponsor – Title
SpotBowl.com
USA Today
5
Anheuser-Bush – “Fence”
4.07
7.82
4
HomeAway – “Griswold’s”
2.87
7.07
3
Denny’s – “Chicken Warning”
2.73
6.31
2
Dockers – “No Pants”
2.10
4.85
1
GoDaddy.com – “Talk Show”
1.64
4.82
Consumer favorability ratings for Super Bowl commercials also experienced a significant
linear increase with the number of acts in a commercial. The higher the number of acts a Super
Bowl commercial had, the higher the average consumer ratings were. Thus, five-act dramatic
form plays an important role in this high profile advertising atmosphere where heightened
attention and consumer focus invites comment and criticism. One factor that may separate the
winners from the losers of “The Super Bowl of Advertising” is the telling of a story following
Freytag Gustav’s five-act dramatic form. As commercials made more use of the number of acts,
their respective ratings in the Super Bowl commercial polls increased on average. On this stage,
five-act dramatic form (i.e. a more developed plot) has a direct connection to consumer response.
Going back to Aristotle, it seems plot has the ability to arouse emotional response.

Theoretical Implications
For years advertising research has focused on the identification of variables in television
commercials that influence consumer response. Previous research has studied variables such as
humor, music, sex appeal, testimony, rational messaging, emotional messaging, executional
style, product focus and consumer focus. Yet, narrative theory represents a rich new area for
consumer advertising research. Narrative form investigates the intrinsic role stories hold in our
lives and the theory has been applied to various disciplines from anthropology, linguistics and
psychoanalysis to social psychology, literary criticism, cognitive psychology, and philosophy
(see Rayfield 1972; Gee 1985; Cohler 1982; Sarbin 1986; Jameson 1981, Black and Wilensky
1979; Ellos 1994). Kent Grayson (1997) held a special session on narrative theory and consumer
research in which he emphasized the value of narrative to help researchers understand, describe,
explain, and predict consumer behavior. To that end, Deighton (1992) and Arnould and Price
(1993) found that appreciation of consumption experiences depends on how well the experience
can be captured and remember as a story. Hirschman (1988) and Stern (1995) showed how
stories told in television commercials can provide models or templates for consumers’ lives.
Deighton, Romer, and McQueen (1989) and Peracchio (1993) demonstrated how consumers
process and store narratives differently than other kinds of information. This study furthers this
valuable line of research rooted in narrative theory and interest in understanding the impact of
narratives on consumer behavior.
Yet, those early narrative theory studies focused on the presence of drama in television
commercials in the forms of lecture, argument, classical drama and vignette drama. Much
advertising research has also studied Super Bowl advertising due to the high cost, high profile
and mass audience. Previous drama advertising research compared classical drama type and

vignette drama type to lectures and argument (Boller and Olson 1991; Deighton et al 1989; Stern
1994). The results of this study indicate there is additional value in categorizing advertising
drama based on a five-point scale model of an inciting moment, rising action, turning point,
falling action and moment of release derived from Freytag Gustav’s theory of drama and
Aristotle’s emphasis on plot development.
This study was the first to find a correlation of plot development based on Freytag
Gustav’s five-act dramatic form with consumer response in Super Bowl commercial rating polls.
Many storylines in literature, theater and film follow this form, and now it could be an important
factor in predicting consumer favorability of television commercials airing on the high stakes
stage of the Super Bowl. In this highly publicized event, Pre- and post-game consumer buzz is
now seen as an integral part of any Super Bowl commercial effort. The commercial’s ultimate
success is dependent on much more than the 30 seconds that it airs, indicating that the increased
understanding of consumer preference in Super Bowl commercial ratings polls could be an
important factor in eliciting online buzz already proven to increase with airing of a Super Bowl
commercial (Kincaid 2009; McClellan 2010).
Managerial Implications
For marketing managers and advertising practitioners, this research indicates that use of a
five-act dramatic form or story arc could help predict improved performance in the all-important
Super Bowl advertising rating polls. It also indicates that television commercial scripts that
include the drama of a fully developed five-act plot could increase consumer engagement (i.e.
they like the commercial enough to vote for it). Super Bowl Commercial Polls also increase the
likely hood of a particular commercial being shared over social networks such as YouTube and
Facebook via free social media “word-of-mouth” advertising. The top Super Bowl commercial

poll winners also receive positive news media attention that generates valuable earned media
coverage and increased online buzz. Vivek, Beatty, and Morgan (2012) found that involvement
with current and potential customers lead to consequences of trust, value, commitment, word-of-
mouth, and loyalty.
Previous Super Bowl advertising research has demonstrated that online buzz increases
emotional response (Siefert et al 2009) and emotional response increases ad recall (Pavelchak,
Antil and Munch 1988). Super Bowl advertising publicity has also been shown to have a direct
positive impact on the effectiveness of commercials for both recall and recognition (See Jin
2003/2004; Jin, Zhao and Soontae 2006). This study identifies a television commercial form that
drives advertising function through consumer response in higher Super Bowl commercial poll
ratings. This positive consumer response can be seen as generating publicity, online buzz and
consumer engagement – all important factors to commercial success in today’s fragmented
media environment. A study by Cheung, Anitsal, and Anitsal (2007) urged marketers to find
ways to induce positive word-of-mouth about their product/brands through advertising
campaigns that demonstrate strong emotional connection with their consumers. This
acknowledges the persuasiveness and importance of word-of-mouth in shaping consumer’s
attitudes and behaviors. Creating a well-liked Super Bowl commercial can serve as a strong
motivator for this positive communication sharing.
Five-act dramatic form can also provide guidance to pressured marketers and advertising
creatives writing for an audience of millions with a large portion of a marketing budget riding on
a single commercial’s outcome. Imagine facing a blank 30-second television script when
millions and millions of dollars and months of work are on the line. The use of a five-act
dramatic form gives content creators a framework on which to build a favorable Super Bowl

commercial that can transfer its goodwill to the product, service and overall brand while
increasing the possibility of online engagement and buzz. Delivering a top-rated Super Bowl
commercial can generate permanent online content through Super Bowl commercial rating
websites, YouTube, Corporate websites, Super Bowl special shows appearing on networks, Hulu
and on-demand. This Internet content can then impact consumers for years to come. Barber
(2013) found that the Internet has become a new consumer socializing agent that influences
consumer related knowledge and attitudes.
Limitations and Future Research
There are limitations to our research. The sample was based on the results of two Super
Bowl commercial rating polls for two years. Would including previous or subsequent years
affect the results? SpotBowl.com had a large number of votes (over 35,000 each year) and a
broad base of participants from across the country, but researchers did not recruit participants.
Respondents sought out and voted on the commercials on their own accord and could vote as
many times as they wished. USA Today’s Ad Meter was based on smaller focus groups of
recruited participants in an unnatural group setting using handheld meters (Schmalz et al 2010).
Did these circumstances influence results?
Because one of the included polls was online, there may be a bias towards people who
have online access. The sample was not random and included the entire population of both polls
(minus movie ads). Results may also be biased because data measured only people who were
motivated enough to participate in a Super Bowl commercial ratings poll or participate in a
research study. Research on a broader demographic sample may be warranted.
Five-act dramatic form is a broader base of classification that may be a useful evaluation
tool for further consumer response research. Future research into the five-act dramatic form

could test direct effects on consumer emotional response, recall and recognition. Research could
also analyze five-act dramatic form in non-Super Bowl commercials to see if it predicts
consumer response in a less-engaging environment. There is also opportunity to examine the
influence of five-act dramatic form in other advertising mediums such an online brand
experiences.

REFERENCES
Aristotle (2013) Poetics by Aristotle Written 350 B.C.E Translated by S. H. Butcher. The Internet
Classics Archive. Retrieved from http://classics.mit.edu/Aristotle/poetics.1.1.html
Amould, Eric and Linda Price (1993), "River Magic: Extraordinary Experience and the Extended Service
Encounter," Journal of Consumer Research, (June), 24-45.
Bailey, Ainsworth A. (2006). “A Year In The Life Of The African-American Male In
Advertising.” The Journal of Advertising, 35 (Spring), 83-104.
Barber, Nelson A. (2013). “Investigating The Potential Influence of The Internet As A Socialization
Agent In Context With Other Traditional Socialization Agents. Journal of Marketing Theory and
Practice, 21 (Spring), 179-193.
Black, J. G. and R. Wilensky (1979), "An Evaluation of Story Grammars," Cognitive Science, 3, 213-
230.
Boller, Gregory W. (1990). "The Vicissitudes of Product Experience: ‘Songs of Our Consuming Selves’
in Drama Ads,” Advances in Consumer Research, 17, 621-626.
------ and Jerry Olson. (1991). “Experiencing Ad Meanings: Crucial Aspects of Narrative/Drama
Processing.” Advances in Consumer Research, 18, 172-175.
Cheung, Mee-Shew, Anitsal, Meral M., and Anitsal, Ismet (2007). “Revisiting Word-of-Mouth
Communications: A Cross-National Exploration.” Journal of Marketing Theory and Practice, 15
(Summer), 235-249.
Cohler, Bertram J. (1982), "Personal Narrative and Life Course," Life Span Development and Behavior,
vol. 4, Paul B. Baltes and Orille G. Brim, Jr. eds.. New York, NY: Academic Press.
Deighton, John (1992), "The Consumption of Performance," Journal of Consumer Research, 19
(December), 362-372.
Deighton, John, Daniel Romer, and Josh McQueen (1989), "Using Drama to Persuade," Journal of
Consumer Research, 16 (December), 335-343.
Edell, J. A., & Burke, M. (1987). “The Power of Feelings in Understanding Advertising Effects.” Journal
of Consumer Research, 14(3), 421-433
Ellos, William J. (1994), Narrative Ethics (Sydney: Avebury).
Esslin, Martin. (1979). “Aristotle and the Advertisers: The Television Commercial Considered as a Form
of Drama,” Kenyon Review, 1 (Fall), 96-108.

Fleiss, Joseph L. (1981) Statistical methods for rates and proportions. 2nd Ed., New York: John Wiley,
38–46.
Flint, Joe. (2010, February 8). “Saints’ Super Bowl win nips ‘MASH’ finale for most-watched show
ever,” Los Angeles Times,
http://latimesblogs.latimes.com/entertainmentnewsbuzz/2010/02/saints-super-bowl-win-is-.html
Freytag, Gustav. ([1863] 1895), Technique of the Drama: An Exposition of Dramatic Composition and
Art, trans. Elias J. MacEwan from the 6th German Ed., Chicago: Griggs.
Gabler.com. (2010). “How Much Do Television Ads Cost?,” Gabler.com.
http://www.gaebler.com/Television-Advertising-Costs.htm.
Gagnard, Alice. (1989). “Elements of Timing and Repetition In Award-Winning TV Commercials.”
Journalism Quarterly. 66 (Winter), 965-969.
Grayson, Kent (1997). “Narrative Theory and Consumer Research: Theoretical and Methodological
Perspectives.” Advances in Consumer Research, 24, 67-70.
Gee, James P. (1985), "The Narrativization of Experience in the Oral Style," Journal of Education, 167,
9-35.
Hirschman, Elizabeth (1988), "The Ideology of Consumption: A Structural-Syntactical Analysis of
'Dallas' and 'Dynasty,' Journal of Consumer Research, 15 (December), 344-359.
Jameson, Fredric (1981), The Political Unconscious: Narrative as a Socially Symbolic Act, Ithaca, NY:
Comell University Press.
Jin, Hyun Seung. (2003/2004). “Compounding Consumer Interest,” Journal of Advertising, 32 (Winter),
29-41.
------ , Xinshu Zhao, and Soontae. (2006). “Examining Effects of Advertising Campaign Publicity in a
Field Study,” Journal of Advertising Research, 46 (June), 171-182.
Kang, Namjun, Ali Kara, H. A. Laskey, and F. B. Seaton. (1993). “A SAS MACRO for Calculating
Intercoder Agreement in Content Analysis,” The Journal of Advertising, 22 (June), 17-28.
Kim, Juran, Sally J. McMillian, Jang-Sun Hwang. (2005). “Strategies for the Super Bowl of Advertising:
An Analysis of How the Web is Integrated into Campaigns.” The Journal of Interactive
Advertising. 6 (Fall), 59-78.
Kincaid, Jason. (2009, February 2). “Tweet Bowl Taps Into Twitter to Find The Super Bowl’s Real
Winners (The Ads),” TechCrunch. http://techcrunch.com/2009/02/02/tweetbowl-taps-into-twitter-
to-find-the-super-bowls-real-winners-the-ads/.

Klaassen, Abbey. (2006, August 6). “McKinsey Study Predicts Continuing Decline in TV Selling
Power,” Advertising Age. http://adage.com/abstract.php?article_id=110899.
Landis, Richard J. & Koch, Gary G. (1977)."The measurement of observer agreement for categorical
data," Biometrics 33 (March), 159–174.
Merriam-Webster (2013). “Drama.” http://www.merriam-webster.com/dictionary/drama
McClellan, Steve (2010, February 8). “Super Bowl TV Advertisers See Online Gains,” Adweek.
http://www.adweek.com/aw/content_display/news/e3ib96053a9e47796d74a9a1046314f3720.
Nail, Jim. (2007). “Visibility versus Surprise: Which Drives the Greatest Discussion of Super Bowl
Advertisements?,” Journal of Advertising Research, 47 (December), 412-419.
Pavelchak, Mark A., John H. Antil, James M. Munch. (1988). “Super Bowl: An Investigation into the
Relationship Among Program Context, Emotional Experience, and Ad Recall,” Journal of
Consumer Research, 15 (December), 360-367.
Peracchio, Laura (1993), "Young Children's Processing of a Televised Narrative: Is A Picture Really
Worth a Thousand Words?" Journal of Consumer Research, (September), 281-293.
PR Leap. (2010, January 14). “Pepsi Backs Crowdsourcing and Social Media. No Super Bowl
Commercial Will Yield Stronger ROI, Says AdHack,” PR Leap.
http://www.prleap.com/pr/146232/.
Rayfield, J. R. (1972), "What is a Story?" American Anthropologist, 74,1084-1106.
Sarbin, Theodore R. ed. (1986), Narrative Psychology: The Storied Nature of Human Conduct, New
York, NY: Praeger.
Schmalz, Julia, Ron Coddington, Courtney Peksens, Christopher Kamsler, David Carrig, Leslie Smith Jr.
and Rene Alston. (2010, February 15). “2010 USA Today Ad Meter tracks Super Bowl XLIV
ads,” USA Today http://www.usatoday.com/money/advertising/admeter/2010admeter.htm.
Siefert, Caleb J., Ravi Kothuri, Davra B. Jacobs, Brian Levine, Joseph Plummer, Carl D. Marci. (2009).
“Winning the Super “Buzz” Bowl,” Journal of Advertising Research, 49 (September), 293-30.
SpotBowl.com website. (2010). http://www.spotbowl.com/Leaderboard.aspx.
Steinberg, Brian. (2009, October 26). “‘Sunday Night Football’ Remains Costliest TV Show,”
Advertising Age. http://adage.com/article?article_id=139923.
Stern, Barbara B. (1990). “Marketing as Drama: Theatre of the Absurd,” Research in Consumer
Behavior, 4, 195-215.

------ (1991). “Who Talks Advertising? Literary Theory and Narrative Point of View,” Journal of
Advertising, 20 (September), 9-22.
------ (1992). “’What’s in a Name?’Aristotelian Criticism and Drama Research,” Advances in Consumer
Research, 19, 452-454.
------ (1994). “Classical and Vignette Television Advertising Dramas,” Journal of Consumer Research,
20 (March), 601-615.
------ (1995), "Consumer Myths: Frye's Taxonomy and the Structural Analysis of a Consumption Text,
Journal of Consumer Research, 22 (September), 165-185.
------, Jenniger Edson Escalas. (2003). “Sympathy and Empathy: Emotional Response to Advertising
Dramas,” Journal of Consumer Research, 29 (March), 566-578.
Stern, Jerome (2000). Making Shapely Fiction New York: W.W. Norton & Company.
“Super Bowl Ads Take Equal Billing to Game in 2008 According to Hanon McKendry Survey.” (2008,
January 31). PR Newswire. http://www2.prnewswire.com/cgi-
bin/stories.pl?ACCT=KFSNSVY.story&STORY=/www/story/01-31-
2008/0004747337&EDATE=THU+Jan+31+2008,+05:30+PM.
Tao Sun, William D. Wells, Zhao Xinshu, Youn Seounmi. (2001). “Commercial Liking and Memory:
Moderating Effects of Product Categories,” Journal of Advertising Research, 41 (May/June), 7-
13.
Trbic, Boris. (2010). “Scripting Short Films.” Screen Education, 58 (Winter), 72-77.
USA Today Ad Meter website. (2010).
http://www.usatoday.com/money/advertising/admeter/2010admeter.htm.
Vivek, Shiri D., Beatty, Sharon E., and Morgan, Robert, M. (2012) “Customer Engagement: Exploring
Relationships Beyond Purchase.” Journal of Marketing Theory and Practice, 20 (Spring), 127-
145.
Weinberger, Marc G. & Harlan E. Spots. (1989) “Humor in U.S. Versus U.K. TV Commercials: A
Comparison.” The Journal of Advertising, 18 (2), 39-44.
Wells, William D. (1988), "Lectures and Dramas," in Cognitive and Affective Responses to Advertising,
Eds. Pat Cafferata and Alice M. Tybout, Lexington, MA: Lexington, 13-20.
Willobee, Sondra B. (2009). The Write Stuff: Crafting Sermons That Capture and Convince. Louisville,
KY: Westminster John Knox Press.
Xinshu, Zhao. (1997). “Clutter and Serial Order Redefined and Retested,” The Journal of Advertising
Research, 37 (September/October), 57-73.

APPENDIX I – CODING FORM
Dramatic Structure Content Analysis Coding Sheet
TITLE: __________________________________ ADVERTISER: ______________________________
DESCRIPTION:
_____________________________________________________________________________________
_____________________________________________________________________________________
Describe the scenes in the commercial that correspond with each act.
NO ACTS – WHY NOT?:
_____________________________________________________________________________________
ACT 1 – INCITEMENT:
_____________________________________________________________________________________
EXPOSITION IS: NOT AT ALL ESTABLISHED 0 1 2 3 4 5 6 7 8 FULLY ESTABLISHED
ACT 2 – RISING ACTION:
_____________________________________________________________________________________
COMPLICATIONS ARE: NOT AT ALL ESTABLISHED 0 1 2 3 4 5 6 7 8 FULLY ESTABLISHED
ACT 3 -- CLIMAX OF ACTION:
_____________________________________________________________________________________
CLIMAX IS: NOT AT ALL ESTABLISHED 0 1 2 3 4 5 6 7 8 FULLY ESTABLISHED
ACT 4 -- FALLING ACTION:
_____________________________________________________________________________________
FALLING ACTION IS: NOT AT ALL ESTABLISHED 0 1 2 3 4 5 6 7 8 FULLY ESTABLISHED
ACT 5 – RELEASE:
_____________________________________________________________________________________
CATASTOPHE IS: NOT AT ALL ESTABLISHED 0 1 2 3 4 5 6 7 8 FULLY ESTABLISHED
Briefly describe how the commercial corresponds with the structure of a comedy (protagonist is better off
at the end) versus a tragedy (the protagonist is worse off at the end). Leave the opposing genre blank.
COMEDY: __________________________________________________________________________
____________________________________________________________________________________
TRAGEDY: _________________________________________________________________________
____________________________________________________________________________________

APPENDIX II – CODER GUIDE
Dramatic Structure Content Analysis Coding Sheet
INSTRUCTIONS: The purpose of this sheet is to code television commercial content for the presence of
a dramatic structure. In particular, indicate whether or not each of the five stages or acts of a dramatic
structure are present in each commercial. In addition, each commercial should be classified as either a
traditional comedy (protagonist becomes better off) or tragedy (protagonist becomes worse off).
DEFINITIONS:
Act 1 – Exposition (Inciting moment). The characters in the story are introduced and time and place are
established. Antecedent information is given. The origin of impending conflict and dramatic tension is
established. The exposition provides the background information needed to properly understand the story,
such as the protagonist (principle character), the antagonist (opposing character), the basic conflict, and the
setting. It ends with an inciting moment, which is the incident without which there would be no story. The
inciting moment sets the remainder of the story in motion beginning with the second act, the rising action.
Act 2 – Complication (Rising action). The course of action becomes more complicated, the "tying of knots"
takes place. Interests clash, intrigues are spawned, and events accelerate in a definite direction. Tension
mounts, and momentum builds up. During rising action, the basic internal conflict is complicated by the
introduction of related secondary conflicts, including various obstacles that frustrate the protagonist's attempt
to reach his goal. Secondary conflicts can include adversaries of lesser importance than the story’s
antagonist, who may work with the antagonist or separately, by and for themselves or for actions unknown.
Act 3 – Climax (Turning point). The development of conflict reaches its high point, the Hero stands at
the crossroads, leading to victory or defeat, crashing or soaring. There is a turning point, which marks a
change, for the better or the worse, in the protagonist’s affairs. If the story is a comedy, things will have
gone badly for the protagonist up to this point; now, the tide, so to speak, will turn, and things will begin
to go well for him or her. If the story is a tragedy, the opposite state of affairs will ensue, with things
going from good to bad for the protagonist.
Act 4 – Reversal (Falling action). The consequences of Act 3 play out, momentum slows, and tension is
heightened by false hopes/fears. If it's a tragedy, it looks like the Hero can be saved. If it’s a comedy, then
it looks like all may be lost. During the falling action, or resolution, which is the moment of reversal after
the climax, the conflict between the protagonist and the antagonist unravels, with the protagonist winning
or losing against the antagonist. The falling action might contain a moment of final suspense, during
which the final outcome is in doubt.
Act 5 – Denouement (Moment of release). The conflict is resolved creating normality for the characters
and a sense of release of tension and anxiety. The comedy ends with a conclusion in which the
protagonist is better off than at the story’s outset. The tragedy ends with a catastrophe in which the
protagonist is worse off than at the beginning of the narrative.
Comedy. A comedy is drama of a light and amusing character that typically has a happy ending. It deals
with the comic or with the serious in a light or satirical manner. If the story is a comedy, things will have
gone badly for the protagonist up to Act 3 or the climax. Just when it looks like all may be lost the tide
turns and things begin to go well for him or her. In a comedy Act 5 concludes with the protagonist being
better off than at the beginning.
Tragedy. A tragedy is a drama that typically describes the downfall of a protagonist. It usually deals with
a conflict between the protagonist and a superior force (such as an antagonist or simply destiny) that leads
to a sorrowful or disastrous conclusion eliciting feelings of pity or fear. If the story is a tragedy, things
will go from good to bad for the protagonist. Act 3 or the climax is accentuated by a false sense of hope
that the protagonist can be saved. In a tragedy Act 5 concludes with the protagonist being worse off than
at the beginning of the story.