Quality signals in the creative industries : the effect of directors’ and leading actors’ reputation on commercial performance in Dutch cinemas.

(1)

QUALITY SIGNALS IN THE CREATIVE INDUSTRIES:

THE EFFECT OF DIRECTORS’ AND LEADING ACTORS’ REPUTATION ON COMMERCIAL PERFORMANCE IN DUTCH CINEMAS

BY WOUTER DE RUITER BSC

WOUTERDERUITER@GMAIL.COM STUDENT NUMBER 6126839

THESIS SUPERVISOR DR.FREDERIK SITUMEANG F.B.SITUMEANG@UVA.NL

SECOND READER DRS.MONIKA KACKOVIC M.KACKOVIC@UVA.NL

(2)

(3)

Abstract and Contents 3

Abstract

and Contents

Author: Wouter de Ruiter

The motion picture industry produces experience goods, which generate highly variable revenue streams and are of cultural importance. The uncertainty revolving around a movie’s financial potential incentivized scholars to study motion picture’s properties and thereby marked its importance in academic research. One of these properties regarded the inclusion of certain star-actors or –directors, which could play a role in creating awareness and subsequent attraction of consumers for the future release of a movie. However, past-research to this concept of star-power prevailed to be inconclusive and incomplete. The research at hand seeks to provide a more in-depth longitudinal analysis by relating various past-performance measures of involved movie-personnel to their commercial success in future projects. The concept of the personal brand and quality signaling is introduced, which reflects on the way the career and image of involved director(s) and leading cast is fed back to the consumer. In what way this eventually leads to movie going behavior is covered by the concept of consumer persuasion, which entails the (heuristic) processing of information. Results show that the actual commercial- past-performance of directors and leading cast members in some cases result in commercial performance for their future projects, but that these relations did not apply when artistic past-performance was considered. All in all, this study connects to a vast stream of literature on the concept of star-power within markets for experience goods and points at several academic- as well as managerial implications.

(4)

Forrest Gump’s mother always said, “Life was like a box of chocolates. You never know what you’re gonna get”. The rational consumer gets to select a specific cholate of which they expect the contents to be in line with their taste. Hereby, their decision-making is influenced by past experiences with cholate boxes and personal preferences. However, while the appearance of the chocolate hints an appetizing continuation, the consumer is never fully able to assess whether he or she made the best choice. In other words, the consumer makes decisions based on incomplete information.

In a way, this still metaphor still reflects on consumers’ current experiences in the movie theatre. Upfront consumption, moviegoers can put off available information on a new motion picture, a chocolate, against their current movie preferences and past-experience with movies they perceive as being similar. For example, consumes could consider the involved personnel with the new picture (Elberse, 2007), teasing visual material (Finsterwalder, Kuppelwieser & de Villiers, 2012) and early-stage critical acclaim (Eliashberg & Shugan, 1997), which might result in them being pulled to the ‘silver screen’.

Hence, communicated movie-characteristics add up to the consumer’s broad idea of the movie’s quality, resulting in expectancies which could act as a trigger for behavioral intentions whenever this perceived quality meets the consumer’s demanded degree of ‘enjoyment’ (Basuroy, Desai & Talukdar, 2006). However, it sticks that, in the end, moviegoers can withhold from any consumption at all and moreover that they can only accurately assess their real enjoyment after actual movie consumption. Following from there, theatrically released motion picture only allow consumers to justify their behavior if their actual enjoyment meets upfront expectations. This characteristic leads ‘product-quality information asymmetries’ (Sawhney & Eliashberg 1996) to remain in this industry and from a consumers’ perspective this causes two types of markets errors to emerge. Firstly consumers could commit to consuming the wrong product based on their preferences; secondly, consumers might refrain of consuming a product which they would actually have enjoyed.On the supply side of the market, producers and distributors try hard to communicate the mentioned quality of their new projects. However, also these parties remain with an important missing piece of information, namely whether their (target) audience will actually like the movie(s) in question.

All in all, uncertainty thus prevails in the whole market. Some scholars therefore claim that a paradigm of ‘nobody knows’ reigns this industry (Caves, 2003; Eliashberg , Hui & Zhang, 2007), where risky and highly variable revenues are accepted and where a single motion picture could turn losses into profits (de Vany & Walls, 1999; Simonoff & Sparrow, 2000). Not surprisingly, industry experts and scholars have sought for methods to assist in understanding consumer preferences and -behavior that might improve producers’ and distributors’ ability in ameliorating their connecting with

(7)

Introduction 7 the target audience. In other words, for methods that could improve overall success rate of the industry (Eliashberg, Elberse & Leenders, 2006; Gazley, Clark & Sinha, 2011).

As mentioned, Informational dynamics on a motion picture’s quality could excite consumer’s interest, lead to certain expectations and might act as a cue for behavior (De Vany, 2004). Therefore, starting point for a more complete understanding of consumer behavior in the industry lies with decomposing the nature of the communicated movie characteristics. Logically, research has been conducted to the effects of these communicated ‘informational cues’ or ‘quality signals’ on consumer behavior. For example, Sochay (1994) studied the effects of acclaimed actor/actress involvement on box-office performance and Simonton (2009) the effects of awarding. Furthermore did Hennig-Thurau, Houston and Heitjans (2009) research the effects of sequels as brand-extensions, Prag and Cassavant (2004) the effect of the communication itself (the effect of marketing as an exposal mechanisms).

However, most of these studies neglect the fact that movie projects are interrelated. Next to a sequel relation, projects can be connected by prominent personnel. Involved directors, actors and actress control their character and are able to build reputation on basis of previous projects, both artistically and commercially, which is transferred from project to project. Subsequently, one could argue that research studies like that of Sochay (1994) should decompose the quality signal more in-depth. In this case, it would be valuable to consider actual past-performance of involved movie-personnel next to their mere involvement in the project. Hence, longitudinal effects of reputational mechanisms could thereby be featured more precisely.

This study therefore sees value in engaging this research gap. Mainly on basis of signaling- and persuasion-theory (Spence, 1974); Chaiken, 1980; Chaiken, 1987) it tries to clarify in what way the past-performance of directors, actors and actresses respectively affect the consumer’s movie going behavior. In other words:

In what way is the commercial success of a motion picture affected by the artistic- and commercial- past-performance of its involved director and leading cast?

The setting for this study will be the Dutch Motion Picture Industry, which will be featured in the next section. Subsequently, this study’s theoretical framework will be made more explicit and clear hypotheses will be drafted. Following from there, the used research method and data is described. This will be continued by presenting the conducting data analyses, a comprehensively discussion of the results, directions for future research and a final conclusion.

(8)

Literature Review 8

Literature Review

2.1 Theatrical Release in the Netherlands

2.1.1 Theatrical Release Rights and Exhibition

With about 270 exhibitors across the country, the Dutch Theatrical Market for motion pictures is fairly small in comparison with for example the U.S.Market (NvB, 2014). Still, a slate of movies ends up being screened at either the more commercial cinemas or the smaller film theatres. The former cinema’s (amongst which cinema chains such as Pathé, JT, Wolff and Utopia/Utopolis) focus mainly on large Dutch or Hollywood productions and allocate movies to screens on a weekly basis such that the amount of visitors is maximized (Eliashberg et al., 2009). These multiple screen venues are equipped with the latest technology and have been established through horizontal integrations (Mergers and Acquisitions) and development of real estate (Eliashberg et al., 2006).

On the other hand, theatrical release thus takes place in smaller film theatres or ‘art-houses.’ This type of exhibition is based largely on non-commercial financing (Buisman, 2011; Eliashberg et al., 2006) and programming allows for new international production and concepts to be screened. These types of theatres present opportunities to its customers to distinguish themselves from ordinary moviegoers, since it carries an image of being more intellectual and high-culture (Wilinksy, 2001). However, as Wilinksy (2001) points out, the general art house theatre remains in constant negotiation with mainstream theatre, as it contributes to the process of staying commercially viable, while retaining artistic integrity (Caves, 2000).

Which movies both types of exhibitors will eventually screen, depends on their selection out of the available supply. Each week exhibitors could choose from as much as 3 to 5 movies for screening (Eliashberg et al. 2009), which are in this case supplied by Dutch film distributors. Hereby we distinguish broadly between two types of suppliers or film distributors, namely major- and independent distributors. In order to supply a picture to a theatre, both types of distributors need to be in possession of the so-called Theatrical Distribution Rights. Satorius (2003) defines these as the rights “that allow the distributor to exhibit new movies in theatres which are open to the public and

for which admission fees are charged”. In the Netherlands, distributors could try to buy these rights,

regardless of their current owner (NvF, 2011). However, major-distributors (e.g. Warner Bros.) usually have an advantage over their independent counterparts, since mentioned rights could be assigned to them free of charge by transfer from a foreign parent company.

(9)

Literature Review 9

2.2.2 Contracting

It becomes clear that distributors are intensively dealing with exhibitors as well as current theatrical release right owners, which requires contracting in two different directions. Firstly distributors sign contracts with the original film right holders. This transaction often takes place upfront production completion, meaning that it takes the form of an investment, thereby adding up to the production budget of a specific motion picture. Whenever theatrical release rights are assigned, it provides distributors with ‘a clear claim on residual profits’, since it shares net proceeds after distribution fees and expenses with actors, directors and other participants (Corts, 2004).

Secondly, distributors draft contracts with (non)-commercial theatres, which allows exhibitors to rent one of the movie prints and screen the picture in the theatre. In return, a minimum playing time is agreed as well as a clear weekly disposition of box-office receipts. Buisman (2011) denotes that during this process, the distributor basically covers the direct sales of a movie to the audience, which includes advertising expenses. In close collaboration with filmmakers, producers, exhibitors and other stakeholders, distributing parties deliver information about the film (of which they own the Theatrical Distribution Rights) to the final consumer. This happens via different media, with the eventual goal of bringing the movie under attention and to attract visitors (NvF, 2011).

From this thesis’ data-set it becomes clear that roughly 10% of all screened titles in the last decade (2002-2013) were domestically produced. Not surprisingly the Dutch Industry of Cinema, just like the European Industry as a whole, is characterized as suffering from structural weaknesses (Herold, 2004). This does not imply that domestically produced titles are not able to perform at the box office or to achieve positive critiques. It only points out that there is a more substantial supply of foreign titles (e.g. Hollywood related pictures).

2.2 Building Reputation

Following Delmestri et al. (2005) and Jantchi (2008), this thesis takes the point of view that a successful theatrical release either indicates box-office success or/and that the quality of its cultural contents is widely acknowledged. But to whom can this successful release be beneficial in the future? Moreover for this study’s interest: which contributors of a movie project are able to build (or tear down) reputation in the future, based on the performance of their current projects?

Previous literature (Faulkner & Anderson, 1987; Delmestri et al., 2005; Peretti & Negro, 2007; Jantchi, 2008) denotes that there are many contributors to a movie project. For example: multiple cameramen, a cast of actors/actresses, editors and technicians all provide their specialized inputs. However, it becomes clear that in a movie project, it is the producer, together with the director and screenwriter, who interconnect involved parties and are at the heart of the production.

(10)

Literature Review 10 Usually producers muster the majority of all personnel (Eliashberg et al., 2006), but this does not straightforwardly imply that he or she builds the most vivid reputation towards the movie going audience. In contrast, for example Chang and Ki (2005) speak of the director and involved actors/actresses as being mainly part of the brand-package that is being communicated towards the audience. Thus one could say that these parties have the most vivid reputation towards the movie going audience. Moreover, Kim (2013) puts this slightly different, namely that the main actor or actress and director are the most likely determinants of whether or not a movie will be a success at the box office.

These statements are not surprising, as the director is usually supervising the creative process and responsible for the creative content of the initial project idea. The director has the ability to combine all creative aspects of a movie into an attractive mix (Hennig-Thurau, Houston & Walsh, 2007). Secondly, mentioned actors and actresses are responsible for the actual creative interpretation of the guidelines set by their superiors and of utmost importance for the attractive creative mix. They could appeal greatly to the imagination of moviegoers, possibly resulting in the emerging of new star-actor s (or actresses) (Rosen, 1981). One could argue that thereby, from a consumer’s point of view, the eventual artistic- or/and commercial performance of a project can be contributed to the efforts of the involved director and main cast.

The Oxford Dictionary (2014) defines the above written “reputation” as being: “A widespread

belief that someone or something has a particular characteristic”. Thus whenever the success or

failure of a movie is contributed to a director, actor, actress or collaboration, this means that the widespread beliefs about these personnel’s capabilities are fed. Subsequently, consumers remain with quality expectations based on the level of perceived reputation, fed by past-performances (Shapiro, 1982).

2.3 Signaling

One could argue that a reputation is than a bundle of positive, negative or neutral information which is transferred over time. Thereby it forms the basic constructs of a signal (Connely et al., 2011), which has wide applicability in the motion picture industry. It becomes clear that it provides underlying informational mechanisms that connect the past- and future performances of the most prominent people involved with movie projects. This section will therefore elaborate on three topics, namely the signaling at the supply side of the industry, the conceptualization of the personal brand (extension) and furthermore this section will elaborate on the signaling process at the demand-side.

(11)

Literature Review 11

2.3.1 Directors’ and Main Cast’s Participation in Future Projects

In deciding upon watching a movie in the cinema, consumers’ decision-making is driven by their perceived expectation of enjoyment. This perceived enjoyment was said to be constructed on basis of informational dynamics that allow consumers in approximating the degree to which the received information about the movie matches the preferences of the individual moviegoer.

It is evident that this information is limited, as it withholds an evaluation of actual enjoyment. On top of that price differentiation does not apply in this industry, since cinema’s and film theatres charge the same prize for movies of different quality. Spence (1974) mentions that the decreasing ability of the individual consumer to evaluate the product, requires increasing presence of signals that do not directly derive from the product itself. Hence, signaling is specifically usable for products whose quality is inaccurately assessable upfront consumption. Amongst these products are experience goods, like motion pictures. There is thus impetus for filmmakers to provide quality signals. One of these signals could be the reputation of involved personnel (Wang et al., 2010), but only if the reputation is perceived as being true (Connely et al., 2011).

This emphasizes that the process of signaling around a new projects commences when a producer and his or her financial backers decide upon the directors, actors and actresses suitability for a specific motion picture. In a sense this overlaps the example of Spence (1974), where job-applicants could signal quality to their superiors by having had a particular education, hinting at sufficient levels future performance. In this case directors, actors and actresses increase the filmmaker awareness and perceptions of quality, by means of their current reputation, which reflects their generated commercial- and artistic exposure respectively.

First of all, these quality signals thus project potential commercial- and artistic performance of the motion picture after its completion. However, they are also a necessity during the production, as sufficient funding (a satisfactory production budget) should be gathered in order for completion of the movie-project as it was sketched by the filmmakers. Lampel and Shamsie (2000, p.250) namely denote that in the motion picture industry, filmmakers “ are likely to find it easier to get backing for their movies if they contain elements that can readily be communicated and made attractive to a wide audience.

Large Hollywood productions are often reluctant on the backing of a studio, while independent film makers are often required to draw a comprehensive financing plan with e.g. distributors, TV Stations and other private Investors. In the Netherlands independent filmmakers can also appeal to the additional funding of the Dutch Film Fund. This is a not for profit governmental institutions that nowadays works with €27.32 million in subsidiaries to cover operations that cover participation in development, production, distribution, promotion and participation in shaping a good climate for the national film industry (Film Fund, 2014). Thus, if filmmakers would like to

(12)

Literature Review 12 increase their chances to be backed by studios and distributors or claim funds from e.g. the Dutch Film Fund, their projects should upfront signal a sufficient of quality.

At the same time, financial backing is also induced by an innovative idea or interesting character. Lampel and Shamsie (2000) point that: “any form of novelty that can help to set a movie apart from others is always an alluring prospect because of its potential to grab attention” (p.250). This is directly opposed to the aforementioned, which denoted that proven successful elements (such as specific actors, actresses and directors) indicate that an investment is worthwhile because of its low-risk character.

In short, signaling properties such as the past-performance of actors, actresses and directors could offset risks at the production and distribution stage of movie projects (Eliashberg et al., 2006). Hereby, the actual involvement of movie personnel on basis of quality signals, hinges on the filmmakers chosen balance strategy between convention and innovation with a new project. The early-stage suitability-decision of mustering personnel is an essential keystone for further persuasion of consumers as their signaled quality could play an essential role in future marketing- and distribution-strategies. Eventually this will reduces projected information product-quality asymmetries between multiple parties (Spence, 2002).

2.3.2 Personal Brand (Extensions)

The process by which filmmakers and their financial backers decide on the suitability of personnel and how their involvement will be signaled to target audiences are thus part of a wider communication strategy. In a sense this could be described as a process whereby filmmakers capitalize on human assets, amongst which human specific reputation, which was said to be constructed based on past-performances. One could argue that according to the Oxford Dictionary this reflects largely on the concept of branding: “the activity of giving a particular name and image to

goods and services so that people will be attracted to them.” Following from there, Wood (2000,

p.666) notions that a brand is the: “mechanism for achieving competitive advantage for firms,

through differentiation (purpose). The attributes that differentiate a brand provide the customer with satisfaction and benefits for which they are willing to pay.”

On basis of these definitions one could argue that directors’, actors’ or actresses’ reputation, when observed, creates a recognizable mini-brand with which the moviegoer could identify themselves and for which he or she is willing to pay. Lair, Sullivan and Cheney (2005) identified this phenomenon as “personal branding”: Practices of humans marketing themselves and/or their careers as brands. It entails that brandings also involves a unique (human) asset, by creating a memorable impression, based on exterior appearance, areas of knowledge or skill. Hence it follows that either filmmakers (or a personal brand itself) could capitalize on these human assets. As there is

(13)

Literature Review 13 impetus for filmmakers to signal quality to consumers, it follows that vast association with involved personal brands could create awareness, excitement and anticipation, which could help subsequent sales (Situmeang et al., 2013).

Research in the motion picture industry has already conceptualized the more renowned personnel brands as “stars”. Chang and Ki (2005) distinguish herein between the star-actor (Rosen, 1981) and the star-director (Walls, 2005). Subsequently, the result of filmmakers capitalizing on stars (or personal brands) in persuading consumers is conceptualized as “star power”. Hereby, stars could positively influence media- and consumer attention, which indirectly influences consumer’s movie going intentions (Wang et al., 2010; Basuroy, Chatterjee, and Ravid 2003).

Wang et al. (2010) mention that star-power could thus signal the role of a brand extension. On a conceptual level, movie sequels were usually mainly regarded as brand-extensions. Hereby the original concept left such a memorable footprint that it allowed for future capitalization by using the known concept in a different setting. The sequel transfers ‘a familiar product’ over time and in its turn this could reduce consumer’s uncertainties about its quality since it is a similar product. In a sense, the addition of a new motion picture to the oeuvre and reputation of a personal brand could thus also be conceptualized as a personal brand-extension. The familiar face, or known reputation, is extended from one project to another and allows filmmakers to capitalize on these personal brands that have demonstrated appeal in the market place.

With respect to star-power, past research is inconclusive. Mixed results have been found and as Eliashberg et al. (2005) mentions: it is unclear to what extent stars actually contribute to the success of a movie and benefit themselves from the success of a movie. Bagella and Bechetti (1999) find a positive relation between the past ‘performance’ of the director and box office success. However, most studies (e.g. Litman, 1983; Sochay, 1994) find no significant relation. As for actors and actresses, Litman and Kohl (1989), Wallace, Seigerman and Holbrook (1993), Bagella and Bechetti (1999), and Sochay (1994) find evidence of a positive relation between popularity of the cast of actors preceding the movie’s release on its subsequent box office performance. Furthermore, Pokorny and Segwick (1999) find that this positive relation between stars and commercial success only holds for big budget films and Karniouchina (2011) adds that the star buzz can enhance box-office receipts mainly during the opening weekend and contribute to the pre-release anticipation of a movie. However, Litman (1983), Prag and Casavant (1994), Ravid (1999), Brewer (2009) and de Vany and Walls (1999) find no significant relation of the star power. The latter proposes that only the audience makes a movie a hit ‘and no amount of star power or marketing can alter that.’ (de Vany & Walls, 1999, p.1). Lastly, research also elaborated on the effects of awarded personal brands on subsequent receipts at the box office. Nelson et al.’s (2001) findings suggest that specifically an award nomination (for top prizes such as academy awards) or won award positively impacts a firm’s

(14)

Literature Review 14 probability of survival, market share of screens and average revenue per screen. Moreover, Prag and Cassavant (1994) find that awards are positively related to marketing expenses, and marketing expenses in their turn positively related to movie revenue

With respect to (personal) brand-extensions, past research seems more conclusive. Smith and Park (1992) note that the extension of a brand allows for cost savings, descending from avoided introduction costs of a new concept. This eventually results in increased advertising efficiency because target audiences are familiar with the concept (Kelller, 2013). Secondly, brand extensions could appeal to an existing fan-base, thereby requiring a lower degree of awareness, as the fan-base is more easily activated. Moreover could it possibly influence renewed interest in the original brand. In this case this would imply renewed interest in either the preceding motion picture of the sequel or earlier projects of the involved director and leading cast.

As a result, Litman and Kohl (1989), Prag and Cassavant (1994) and Ravid (1999) found that there are significant relations between the previously conceptualized brand extensions (sequels and/or returning collaborations) and future theatrical earnings. However, Sochay (1994) does not find the relations reported in the study of Prag and Cassavant (1994), but he used a different sample. Basuroy and Chatterjee (2008) find that although a sequel is an efficient brand extension, in reality these projects do not perform as well as their originators in terms of box-office revenues. Empirical analysis showed that sequels, which have a smaller time-lag with their originals, perform significantly better at the office than sequels released further in the future. This implies that a =brand loses commercial value over time due to lack of exposure, which reduces the chances of profitable exploitation of a possible extension. The same might thus imply for the personal brand-extensions, where recent performance could have significant effects on future performance, while an increase in time length between the current and future project could antagonize this effect.

2.3.2 Quality Perceptions and Evaluative Consensus

For filmmakers there is impetus to provide quality signals, but for consumers there is impetus to look for these signals, as they try to form perceptions of the quality, which might result in perceived expectations of enjoyment (behavioral intentions) and movie going behavior (Basuory et al., 2006). Thus an important mechanism emerges, which reflects on how awareness and perceptions of the receipt signals, leading to the formation of consumer’s behavioral intentions. However, this will be comprehensively discussed in the next section. For no, it should be denoted that the consumer’s behavioral intentions at the exhibition stage in the motion picture industry hinges on information about the quality of the product from a variety of sources (Reinstein and Snyder, 2000).

(15)

Literature Review 15 Initially, quality signals (e.g. on involved personal brands) are directly received by producers, distributors and exhibitors via different media. Posters, TV-commercials, or notions inside the movie theatres themselves mediate the transfer of information, which influences consumer’s product awareness, resulting in a certain perception of quality and expectation of enjoyment.

Secondly, professional critics are favored to judge the product qualities before moviegoers do and therefore they are able to interfere during this awareness process. Critic’s opinions are able to stir curiosity and to decreases target audience’s uncertainties regarding the movie’s quality (Litman, 1983), which in its turn could accelerate or slow down demand for the upcoming release. Hence, as the perceived quality of the picture is either confirmed or rejected, this predicts (spurious correlation between reviews and induced demand by mutual correlation with quality) or influences (causal effect of reviews on demand, holding quality constant) eventual consumers behavior (Eliashberg & Shugan, 1997; Gemser, van Oostrum & Leenders (2007).

Most important is arguably the word-of-mouth process on basis of quality perceptions, which takes place amongst consumers (Wang et al., 2010; Basuroy et al., 2006; De Vany & Walls, 1999). Independent moviegoers could make other potential viewers aware of a new project by forming evaluative information based on their perception of actual creative performance by personal brands in the consumed motion picture. This might take the form of visible evaluations (magazines, internet), or invisible evaluations (friend’s advice) and could lower evaluative uncertainty even more. Therefore these consumer evaluations could next to critical evaluations be of influence on movie-goers behavioral intentions.

However, notions should be made on the concept of retrievability and consensus. In order for (the past- and/or current personal brand’s project) critical- and consumer evaluations to lower quality uncertainty, they should be retrievable in the first place. Hereby consumers can set off these evaluations against their own expectations, based on their perception quality. Secondly, it requires both types of evaluation to have a high degree of consensus (either positive or negative).

The effect of variability in evaluations on consumer’s movie going behavior in the sphere of brand-extensions is for example highlighted in the study of Situmeang et al. (2013). In their research they conclude that in a series of sequels (brand-extensions), every edition of this series matters. If critics conclude there is a poor quality edition amongst the series, it will hurt the commercial success of possible latter sequels. Not only is there a carry-over effect distinguished from one edition to another, there also follows negative effects on performance as the variation in evaluations creates uncertainty at the demand-side. Following earlier reasoning, such findings could be generalized to the concept of the personal brand extension.

(16)

2.4 Information Processing: Signals, Cues and Effects on Planned Behavior

From previous section it should become clear how personal branding serves as a tool for signaling quality. On top of that, it could reduce uncertainty in the consumer’s process of forming expectancies and eventual movie going behavioral intentions. Herein, consumer, critics, filmmakers and the personal brands themselves have a clear role. This section in its turn describes a more in depth discussion on the mental processes of signal interpretation and hence clarifies how mentioned uncertainty on product quality is actually reduced. In other words, how the interpretation of signals persuades consumers and their subsequent behavior.

2.4.1 Persuasion, Systematic- and Heuristic Information Processing

Previous sections described the way in which personal brands were to be featured in new movie projects and how their brands (and other signals) were communicated to consumers in order to attract them. The overarching concept of “Persuasion” thus seemed to reflect on this process rather well. Surely, when the Business Dictionary defines persuasion as: “A process aimed at changing a person's (or a group's) attitude or behavior toward some event, idea, object, or other person(s), by using written or spoken words to convey information, feelings, or reasoning, or a combination of them.”

Within the field of persuasion-research, multiple theories have been put forward, which elaborate on signal-processing or informational-cues themselves. Stiff (1986) acknowledges the Kahneman’s (1973) Elastic Capacity Model (ECM) and the Elaboration Likelihood Model (ELM) of Petty and Cacioppo (1986) as being one of the more prominent models. Both dwell on the effects of source and message cues in persuasion and concern the moderating effect of message recipient “involvement” on the effectiveness of central messages and peripheral source cues on attitudes. Meanwhile researcher Shelly Chaiken (1980, 1987) developed the Heuristic-Systematic Model (HSM) and Eagly and Chaiken (1993) conclude that in future research for understanding a variety of social influence phenomena, ELM and HSM should be treated as complementary models to create a ‘dual-processing framework’. This is because both models recognize a host of variables conceptually independent of message quality that influence people and moreover could they trigger qualitatively different information processing.

However there are some subtle different approaches in both models. On the one hand ELM (Petty and Cacioppo, 1986) discusses two main routes of persuasion processing, unlike HSM: a central route and peripheral route for information processing. Hereby both routes could act as a device for altering attitude, in this case towards movie going behavior. Griffin (2006) explains that the central route is reflective and requires the processor with willingness to process- and think about

(17)

Literature Review 17 the message (or signal). In contrast, the peripheral route occurs when behavior or attitude is formed without extensive thought, but more from mental shortcuts, appearance cues. As Lord and Maher (1990, p.13) mention, this hints that signal- or cue-recipients could “rely on already developed knowledge structures to supplement simplified means of processing information.” All in all, Griffin (2006), mentions that the route of persuasion processing depends on the level of involvement in the topic or issue. High involvement or elaboration increases central route processing especially when motivation and ability in the message exists. Therefore, low involvement increases peripheral route processing when motivation and ability conditions of persuasion do not exist.

HSM, on the other hand, specifically emphasizes “validity seeking” persuasion settings concerning people’s motivations within the social environment (Eagly and Chaiken, 1993). The model states that individuals can process messages in one of two ways: namely heuristically or systematically, which can co-occur (Eagly and Chaiken , 1993). Heuristic processing occurs when individuals simply take decisions and conclusions based on the information available. It allows individuals to formulate decisions (validate their behavior) based on the opinion of experts and what the consensus believes. On the other hand, Eagly and Chaiken (1993) explain that systematic processing requires analytical processing of information, which requires more cognition. The latter approach thus values signal-reliability and the processing of message contents in its entirety.

2.4.2 Triggering Moviegoers

In Section 2.2 it was described how consumers at the exhibition stage of the motion picture industry receive quality indicators from a variety of sources. As mentioned, signals (or informational cues) are received from producers, distributors, exhibitors, critics and from other consumers.

How the moviegoer uses the signals depends on their retrievability (in this case it is assumed that all signals are equally retrievable) and their diagnosticiyy (ability to be assigned to a specific category or quality) (Basuroy et al., 2006). “Cue diagnosticity theory suggests that consumers first retrieve any cue and, if it is non-diagnostic, retrieve a second cue, and so on, until the retrieved cue is found to be diagnostic (Basuroy et al., 2006, p.289). Hence, consumers need less cues when for example the personal brand is already known (their past-performance) or when it follows that signal-consensus is high, thereby triggering product (brand) awareness and movie going intentions.

It seems to reflect on the view that potential audiences, unlike critics, judge the received signals with respect to the personal brand heuristically. Todorov, Chaiken and Henderson (2002) explain this as being a “brand name heuristic”. Hereby the brand name, as a memory structure, can provide a basis for decision-making. Meaning that in this case, if the personal brand carries a positive evaluation within the consumer’s mind, than the moviegoer’s movie going intentions might be triggered, merely based on this reputational construct. One could even argue that mere exposure to

(18)

Literature Review 18 a known personal brand is a favorable cue over an unknown personal brand. Thus, the persuasion cue of a personal brand’s quality is diagnostic and easily matched with individual consumers existing memory structures with respect to their past-experience with the brand. Re-Citing: Situmeang et al., (2013): “It awakes awareness, excitement and anticipation, which will help subsequent sales”

However, Situmeang et al. (2013) notes that according to the theory of reasoned action (Fishbein & Ajzen, 1975), behavioral intentions are not only determined by individual attitudes (resulting from persuasion cues), but also by subjective norms. A normal individual could firstly process and value available information and the likelihood of particular outcomes. Thereafter it could arrive at the eventual behavioral intention, when considering and valuing also the subjective norms, the beliefs of others. In this case, an individual would not only consider its own memory of the personal brand, but also the evaluations and expert evaluation as subjective norms, leading to a more systematic processing of information. However, as mentioned, this could co-occur with the heuristic processing.

In a different way, Chaiken (1987) describes this as the bandwagon effect: a possible type of heuristic processing, where people are most likely to consider an opinion as credible if the opinion is supported collectively. Remal and Real (2005) describes these descriptive norm as ‘perceptions of what most people do’. Whenever the perceived prevalence (consensus) of a particular behavior (or attitude) is greater, the more likely people are to engage in the mentioned behavior, as it is perceived as normative. In short, both the bandwagon heuristic and descriptive norms denote a mental shortcut that favors collective sources over individual sources, thereby influencing behavioral intentions.

(19)

2.5. Wrap-Up and Hypotheses

This section will wrap up the discussed literature from previous sections. Thereafter hypotheses will be drawn, while guided by the theoretical framework from Figure 1.

Figure 2.1: Theoretical Framework

Source: Author’s adaptation of discussed literature in Paragraph 2

Theoretical Framework and Hypotheses

Signaling is specifically usable for products whose quality is inaccurately assessable upfront purchase, such as motion pictures, which are experience goods. At the exhibition stage of the motion picture industry various signals are received by consumers which descent from movie producers, distributors, professional critics, or other critics. This results in product awareness and a certain expectation of a new motion picture’s quality. Whenever the consensus in most of these signals is great, this decreases uncertainties regarding the quality, which could accelerate or slow down demand for the release. One of the discussed signals revolved around the past-performance of involved personnel, which was conceptualized as the personal brand. Hereby the moviegoer would be willing to pay for the director’s, actors’, or actress’ reputation.

The answer to the question how the final consumer processes this signal is captured by persuasion theory. One argumentation explains that if the personal brand carries a positive

(20)

Literature Review 20 evaluation within the consumer’s mind, than the consumers movie-going intentions might be triggered, merely based on this reputational construct. This positive evaluation might not only be derived from memory of past experience with the personal brand, but also on basis what the consensus believes is true about the quality of the personal brand (bandwagon-effect). In a way this connects to reasoned action theory, which explains that behavioral intentions are not only determined by individual attitudes (own memory and valuation of available information), but also by subjective norms (other’s memory and valuation of available information). However, the latter mechanism projects a more systematic processing of signal, but this poses no problems as it can co-occur with heuristic processing.

In this study the view is taken that mere exposure to reputational signals of personal brands will heuristically trigger behavioral intentions. This is supported by cue-diagnosticity theory, which states that consumers need less cues when for example the personal brand is already known (their past-performance) or when it follows that signal-consensus is high, thereby triggering more product awareness and movie going intentions.

For the individual brand this means that excellence in: commercial performance of past-projects (Herafter CPP), critical acclaim of past-past-projects (Hereafter CAP) and consumer evaluation of past-projects (Hereafter CEP) could create awareness, excitement and anticipation, which will help subsequent sales. Thereby the personal brand could be positively being carried-over from one project to another. All in all leads to the following hypotheses:

H1a: CPP of a personal brand positively influences commercial performance in future projects H1b: CAP of a personal brand positively influences commercial performance in future projects H1c: CEP of a personal brand positively influences commercial performance in future projects

Furthermore, considering the fact that when consensus in signals is great, this decreased uncertainties regarding a motion pictures’ quality, this could accelerate or slow down demand for the release. Whenever the CPP, CAP and CEP of personal brands show an increasing trend, this means that consensus regarding the personal brands’ quality is acknowledged, leading to the following hypotheses:

H2a: Significant positive developments in CPP of a personal brand positively influences commercial performance in future projects

H2b: Significant positive developments in CAP of a personal brand positively influences commercial performance in future projects

(21)

H2c: Significant positive developments in CEP of a personal brand positively influences commercial performance in future projects

Past research on brand-extensions, showed that sequels with a smaller inter-project time-gap with their originator, outperformed sequels with a larger time-gap. One could hereby argue that either the earlier released sequels profit to a larger degree from the original movie’s exposure and consumer awareness or that the reputational signals decrease in strength over-time. A future project of an actor, actress or director was conceptualized as a personal brand-extension, which were to be affected by the forces written above. Therefore one could argue that a personal brand’s reputational quality signal decreases in strength over time, such that it moderates the personal brand’s past-performance. Hence:

H3 The effects mentioned in H1a, H1b & H1c are moderated by the time-lengt (‘t’) between the past- and future projects of the personal brand

The motion picture industry is shaped by the existence of a variety of products, which are connected by personal brands. However the uniqueness of project could also allow a movie to stand on its own. Logically, past research already focused on the quality signals due with individual motion pictures. For example: Litman (1983); Litman and Kohl (1989); Wallace, Seigerman and Holbrook (1993); Eliashberg and Shugan (1997) and King (2007) clarified the influence that critics have on current movie projects. Moreover, Gemser, van Oostrum and Leenders (2007) clarified that mainstream consumers are not led by critical reviews, but that a prediction effect of reviews of present. On the other hand, art house moviegoers are influenced in their movie going behavior, which is enforced by findings of e.g. Dellarocas et al. (2007).

Furthermore, the word-of-mouth (WOM) process has been discussed as an ingredient linked a movie’s performance (De Vany and Walls, 1999). Similar effects are found for off- (survey’s, inference) as well for online platforms (discussion groups; online reviews; blogs (Abel, 2010; Dellarocas et al., 2007), where results did not seem to differ between mainstream and non-mainstream movies (Yang, Kim, Kim and Amblee, 2009).

All in all leads to the following final hypotheses:

H4a: The critical evaluation of a personal brand’s current project (Hereafter CEC) has a positive effect on its commercial performance

H4b: The consumer generated Word-of-Mouth of a personal brand’s current project (Hereafter CGW) has a positive effect on its commercial performance

(22)

Methods 22

Methods

3.1. Empirical Setting

In order to test for the drafted hypotheses, analysis of secondary data has been conducted. It was possible to obtain data on all theatrically released movies in the Netherlands from the past 12 years (2002-2013). This large time frame allowed for longitudinal observation of the development in the careers of actors, actresses and directors (the personal brands). On top of that, rejection data was obtained from the same time period, which was operationalized as being theatrically released movie titles, although not at the Dutch Cinemas and Art Houses. It was assumed that these rejected movies, could be released in a market with larger potential audiences more easily. Therefore the U.S. market was selected and hence a release list from this market was extracted covering the same time-frame. This list was set off against the Dutch release list, which allowed for the elimination of double cases. Lastly, re-releases on the rejection list were omitted, as they do not reflect on the current reputation of actors, but re-releases on the Dutch list were kept as they were needed in order to calculate a competition parameter, but more about that later in this paragraph. The initial procedures resulted in a sample of 6630 movie titles.

3.2 Data Collection and Variable Operationalization

3.2.1 Release Lists and Commercial performance

The Dutch release list was obtained from http://boxofficenl.net/ , which is a website that tracks weekly data on box-office performances published by the Dutch Association for Film Distributors (NvF). These box-office performances are for this study the operationalization of a movie’s commercial performance. The data of BoxofficeNL included movies’ date of release and thus the box-office performance of feature films as well as theatrically released documentaries. Documentaries which attracted more than 8.000 visitors (+/- 50.000€ in receipts) were included in the database, while others were neglected. The reasoning behind this is that due to the mentioned bandwagon-cue, popular documentaries are in fact seen as substitute-products for feature films. Therefore it would affect the competition mechanism in the market and should be included for later analyses.

Thus, hereafter the information of BoxofficeNL was extracted and information from the Internet Movie Database (IMdB, http://www.imdb.com) supplemented data for missing values It was possible to extract all release dates and to find box-office data on 3427 of 3673 titles (93%), whereby in some cases, gross receipts were estimated by multiplying the estimated amounts of visitors times an average ticket price of six euros.

(23)

Methods 23 In order to correct for inflation, the release date was converted to a release month, which allowed for setting off the box-office performance against monthly inflation rates obtained from the Dutch “Central Bureau of Statistics”. Hereby, the base year for inflation calculation was set at 2004. Thereafter, the conversion of the release date to a specific release week allowed for setting off the box-office performance against manipulated daily Euro-Dollar exchange rates obtained from a Dutch exchange rate website (www.wisselkoers.nl), which was weekly averaged.

Subsequently, the U.S. release list was obtained from http://boxofficemojo.com/yearly/, which is a website (owned by Amazon) that published box-office performance published by American film distributors. The website provided release-date and box-office data for all included 3101 U.S.-only releases. Commercial performance for these rejected titles was set at 0, since no receipts were grossed at the Dutch Box-Office. For database purposes, U.S. box-office data was included and corrected similarly for inflation by using inflation data from the “U.S. Inflation Rate Calculator”.

3.2.2 Seasonality and Competition Measures

There is a tradeoff for filmmakers and distributors between trying to maximize revenue capturing and trying to avoid competition. Rescheduling of releases by distributors because of announced competition is common and competition is said to influence performance outcomes (Einav, 2007).

Therefore this study engaged in the process of calculating two competition scores, which could control for the influence of competition on commercial performance outcomes. Table 3.1.(below) presents the procedure of this calculation.

Table 3.1: Competition Score Calculation

The Dutch Competition score reflects the amount of movies the consumer could choose from and how this competition might have had influence on the commercial performance of movie “x” in week “t”. On the other hand, the International Competition score reflects the amount of movies the distributor and exhibitors could have possibly choose to screen, therefore relating the actual screening of products by exhibitors to a construct of competition before the exhibition stage.

Dutch Competition Amount of Movies Released in Week “t-2” Amount of Movies Released in Week “t-2” Amount of Movies Released in the same week “t” as movie “x” Amount of Movies Released in week “t+1” Amount of Movies Released in week “t+2” International Competition Amount of Present Movies in Week “t-2” Amount of Present Movies in Week “t-1” Amount of Movies Present in the same week “t” as movie “x”

Amount of Present Movies in Week “t+1”

Amount of Present Movies in Week “t+2”

(24)

Methods 24 After calculating the competition scores, some titles were removed from the simple as they did not resemble any importance to this study. Firstly, these titles included re-releases in the Netherlands, since these are not regarded as new projects by directors, actors and actresses, but merely a re-release of an often successful picture in the past. Secondly, it included animation titles, in which eventual success cannot easily be attributed to actors and actresses’ performance, as they are only heard by voice. The resulting sample included 6374 movies.

Thereafter, on basis of the release dates of the remaining titles, control dummies were created in order to control for the effect of seasonal patterns. The importance for this inclusion lies mainly with the study of Radas and Shugan (1998), who estimated for seasonal patterns and found that high season (Summer peak, Christmas peak) releases show huge potential for box office success. However, it should be noted that this effect was moderated by the degree of competition at these periods. The studies of Brewer, Kelley and Jozefowicz (2009) and Simonoff and Sparrow (2000) find similar results

3.2.3 Data Gathering

Multiple sources were used to supplement information on the remaining 6374 movie titles. This helped offsetting the problem of having to deal with incomplete cases due to lack of information. Table 3.2 (below) summarizes which type of data was collected from what specific source

Table 3.2: Data Collection by Source

Source Data

http://boxofficenl.net/ -Dutch Box Office Earnings -Week of Release

-Age Restriction (AL, 6, 12 or 16) -Length of Movie

-Spoken Language - Distributor Name

http://boxofficemojo.com/ - U.S. Box Office Earnings - Movie Budgets

- Distributor Name

http://www.imdb.com/ - Supplemented Dutch/ U.S. Box Office Data -Name of Director (s)

-Name of Main-Cast (Three Leading Actors/Actresses) -Genre

-Movie Budgets

(25)

Methods 25 This study regards both expert and consumer evaluations, which is consistent with the approach of Hennig-Thurau, Houston and Sridhar (2006), who consider Holbrook’s (2005) finding that “consumer and expert quality judgments appear to be created with the same norms” (p.206) . In order to operationalize the consumer evaluation construct, two measures were averaged, namely the average rating by IMDb users as published on IMDb.com (given that the amount of votes “n” exceeds n=30) and the average rating by Rotten Tomatoes users (given that the amount of votes “n” exceeds n=30. The latter measure had to be re-scaled from a 1-5 score to a 1-10 score in order to enable for averaging out the measures. The expert evaluation construct was operationalized by averaging out the MetaCritic- and Tomatometer-score, respectively extracted from IMDb.com and Rottentomatoes.com. Herein is the former measure an average rating of up to 40 leading U.S. film critics, which is calculated by (normalized and weighted) metacritic.com and the latter an average rating of critics connected to the renowned Tomato-meter.

There is enough evidence to assume that the evaluations that stern from mentioned (U.S.) websites can be generalized to this study’s setting, namely the Dutch Theatrical Release Market. Firstly, Dutch audiences are familiar with both websites and are their evaluation scores are included on many Dutch movie fan websites. Secondly, the debate on globalization (Bird & Stevens, 2003) led some authors to point to an increasing convergence on specific forms of musical, culinary or – in this case- artistic culture. Thereby one could assume that the American scores could reflect on Dutch opinions and that the effect that either of the constructs could have on western audience’s movie going behavior is becoming more similarly.

All in all, the view that western artistic culture is converging supports the major use of U.S. critic- and consumer evaluations, which would thereby represent the Dutch audiences’ opinion. Table I (Appendix) clarifies the operationalization of other initial control variables mentioned in Table 3.2, amongst which for example the movie’s budget which is used in past research as a(n) (indirect) commercial success predictor (Litman, 1983).

http://www.imdb.com/ -Metascore Rating (MetaCritic Evaluation) -Sequels

-Gouden Kalf / Oscar / Rembrandt Award / Golden Globe Award Nominees and Winners

http://www.film1.nl/ -Critic Scores

http://www.rottentomatoes.com/ -Tomatoscore (Critic Evaluation) -Audience Score (Consumer Evaluation)

(26)

Methods 26

3.2.4 Past Performance

Next step in the database construction was sorting all movie titles on basis of release date, which resulted in a longitudinal overview. For each case (movie “i”), the Dutch Box office earnings and the consumer- and critic evaluation were retrieved from the previous three projects of all involved personal brands. This was done using various Microsoft Excel formulas. Using these retrieved values and new third party data on movie awards, new past-performance variables were created, which are comprehensively explain in Table II (Appendix).

In order to capture significant developments in the performance of involved personal brands (Related to hypotheses 2), multiple HOT-ratios were calculated. These are operationalized as special dummy variables where for example the HOT-Commercial-Director-Ratio took a value of 1 whenever (one of) the director(s) had an extremely successful past-project or showed significant positive developments in either his (their) last three commercial performances. The specifics and details around this procedure are explained in Table IV in the Appendix.

3.3. Description of Final Sample and Regression Models

3.3.1 Final Sample and Early Stage Problems

Mentioned procedures led to a sample consisting of 6374 movie titles, including 2953 rejected and 3421 released titles. Table 3.3 (next-page) shows how much of these released and rejected projects were related to past-projects in the same time-frame. It becomes clear that for a slate of motion pictures, no past-performance signals were present, wherein the percentage of missing signals was significantly higher for Art House Titles than for ‘Mainstream’ Titles. These missing signals are partly due to the construction of the time-frame, where titles released or rejected in the first half of the timeframe could have included quality signals based on past-projects that precede this study’s given time frame. Secondly, it sticks that for 1196 (18,76%) and 238 (3,7%) of the movie titles no information on critical evaluation and box office, respectively could be extracted from mentioned sources. However, this study’s author is convinced that the overall sample is sufficient to eventually draw conclusions on all categories, except for a separate category of Dutch Language titles. For example, past-critical evaluations are only fully present for n=11 cases, which is evidently too small for drawing conclusion on basis of statistical procedures.

(27)

Methods 27 Unreleased Titles Released Titels w/ Foreign Language Released Titels w/ Dutch Language

Released Art House Titles

Released Mainstream Titles

Both the Director & Main Cast 165 684 72 221 535

Director 187 577 81 320 338

Main Cast 549 607 61 339 329

Without Past Performance N/A 2052 1220 119 922 417

TOTAL 2953 3088 333 1802 1619

Director 350 598 24 322 300

Main Cast 682 582 33 313 302

TOTAL 2953 3088 333 1802 1619

Director 386 626 81 357 350

Main Cast 714 574 62 332 304

TOTAL 2953 3088 333 1802 1619

Past Performance (Weighted Last Three Projects) With Commercial Past

Performance (Previously Released in NL)

With Past Critical Evaluation

with Past User Evaluation

Table 3.3: Description of Final Sample

From Table V (Appendix) it is firstly learned that all measures of user evaluation scores correlate highly with all measures of critic’s evaluations. On top of that, from Table V & Table VI (Appendix), it becomes clear that the variables that were to capture moderating effects of time-gaps between projects and signals pose further possible multicollinearity problems. Therefore, in order to have a healthy functioning predicting model, either the evaluation scores have to be compounded in a single “perceived quality construct” or one of the two constructs has to be omitted from further analysis. Due to limited time at hand, the latter option is preferred to the former. Furthermore, from Table VI (Appendix) it becomes clear that continuous “length” variables and their interaction values with past performances interact. Therefore they were subsequently mean-centered in order to deflect future multicollinearity problems (a procedure suggested by Cronbach (1987)).

3.3.2 Overview of Variables and Multiple Regression Models

The hypotheses are tested by estimating comprehensive model reflecting on the past-performances of involved movie personnel. The statistical software SPSS 20 will be used in order to conduct binary- and multiple linear regressions. A detailed overview of the used variables in these analyses is presented in Table 3.4 (next page). For clarity purposes: a few variables present in Table I & II (Appendix) are renamed differently, but the operationalization remains the same. Furthermore, only the Christmas-season variable was included to control for seasonality, since other dummies proved to provide no significant added explanatory value. On top of that the dummy variables for age restrictions were omitted, since they included too many cases to be included in the probit as well as multiple regression models.

(28)

Methods 28 The eventual model takes the Dutch box-office performance, (past) expert-evaluations of the director and main cast into account as well as some control variables. Amongst these control variables, is a variable (Prob_Screened) which allows controlling for the fact that some movies end up distributed to the theatres and subsequently screened, while others are rejected. This variable is obtained by estimating a preceding probit model, and including the resulting probabilities in this second stage model. This allows the variable (Prob_Screened) to capture endogenous factors which would otherwise be embedded in the error term of the main model.

Table 3.4: Overview of Variables Variable Name Description

Log(BO)i Log value (base 10) of Gross Box Office receipts for movie i in The Netherlands

SEQi Dummy Variable controlling for whether movie i was a sequel

ARTi Dummy Variable controlling for whether movie i was an Art House movie

EXPMi Dummy Variable controlling for whether movie i was expected to be intensively marketed

MAJORi Dummy Variable controlling for whether movie i was released by a Major Distributor

CMi Dummy Variable controlling for whether movie i was released around the Christmas holidays

COMEDYi Dummy Variable controlling for whether movie i belonged to the popular Comedy Genre

ROMANCEi Dummy Variable controlling for whether movie i belonged to the popular Romance Genre

DRAMAi Dummy Variable controlling for whether movie i belonged to the popular Drama Genre

AWDACTi Highlights whether movie i included at least one awarded actor

AWDDIRi Highlights whether movie i included at least one awarded director

log(BUD) i Log value (base 10) of movie i ‘s production budget

NLCOMPi Competition intensity at time of release for movie i

CRITEVi Critical Evaluation (normalized to a 1-10 score) of movie i

log(WPP_BO_D) i Log value (base 10) of the Director ('s average) past Box Office Performance (Either the weighted average of the past _{three projects ('W') or merely the past project ('P'))}

log(WPP_BO_A) i

Log value (base 10) of the Main Cast's past Box Office Performance. Herein does 'W' imply the average of the best 2 main cast members' weighted averages of the Past Three Projects ('W') and 'P' implies merely the average of the best 2 main cast members' latest project's

WPP_CRITEV_Di Director('s average) of CritEV_('P') weighted average of the past three projects ('W') ,or merely CritEV of the past project

WPP_CRITEV_Ai Average of the best 2 main cast members' weighted average CritEV of the past three projects ('W') ,or average of the _{best 2 main cast member's CritEVi of their latest projects ('P')}

LEN_Di (Average) Length (in years) between the Director('s) latest project(s) and current movie i

LEN_Ai Average length (in years) between the Main Cast('s) two latest projects and current movie i

HOTDCOMi Dummy Variable highlighting whether the movie involves a director who is HOT in past commercial performance (See _{Appendix Table IV)}

HOTACOMi Dummy Variable highlighting whether the movie involves an actor or actress who is HOT in past commercial

performance (See Appendix Table IV)

HOTDCRITi Dummy Variable highlighting whether the movie involves a director who is HOT in past critical evaluations (See

Appendix Table IV)

HOTACRITi Dummy Variable highlighting whether the movie involves an actor or actress who is HOT in past critical evaluations

(See Appendix Table IV)

(29)

Methods 29 Hence, the proposed hypotheses will be tested with the following statistical models.

Model 1: Probit Model for whole sample

( ) (

( ( ) ) - ( ( ) ) - ( ) - ( ) -

Wherein (.) represents the cumulative distribution function of the standard normal distribution.

Model 2: Multiple Regression Model for exhibited movies

( ) ( ) ( ( ) ) ( ( ) ) ( ) ( ) ( ( ) ) ( ( ) ) ( ( ( ) ) ( ( ( ) )