Graduate School of Communication Persuasive Communication
Master’s Thesis
Marketing and producing around an actor’s private role
The influence of the “award-winning” tag and scandals around actors on moviegoers’ purchase intentions and attitudes
Lotte Frencken
11626410
Supervisor: Dr. S.F. Bernritter
Abstract
The purpose of this study is to investigate the extent to which the reputation and involvement of an actor influences moviegoers’ intentions to purchase a film ticket, their attitude toward the film, and their attitude toward the actor. Accordingly, this study investigates the impact of an award-winning versus a scandalous actor on moviegoing intentions and attitudes, as well as the differential impact of an actor who is part of the film versus an actor who solely endorsers a film. An experiment involving manipulated articles from the Hollywood Reporter generated the results among 204 moviegoers. The results suggest that the “award-winning” tag used to describe an actor signals quality, thereby increasing moviegoers’ purchase intentions. The prestigious prize also increased the evaluation of the film and the actor, while a scandal proved to significantly worsen these evaluations. In fact, moviegoers exposed to a scandalous actor predominantly evaluated the movie as worse than neutral and indicated they were not interested in purchasing a ticket. This suggests that moviegoers tend to view the actor and his role as intertwined. Furthermore, an interaction effect between an actor’s reputation and involvement on purchase intention demonstrated that an Oscar-winning actor led to greater purchase intentions when the actor starred in the movie than when the actor endorsed the movie, while a scandalous actor who starred in the movie also brought more risks than an endorser. This was not the case for moviegoers’ evaluations of the film and the actor. Taken together, these findings extend prior research on star power and celebrity endorsement, and offer insight into how film producers can strategically market and protect their films.
Keywords: actor, endorser, reputation, star power, celebrity endorsement, brand attitude, film marketing
Hollywood actors have dominated popular culture since films have become one of the prime sources of entertainment. At the same time, movie stars also prevail in advertising, as approximately 19% of televised ads featured a celebrity endorser in the U.S. in the past few years (Elberse, 2012). However, while it has often been researched how the reputation of celebrities can help or hurt the brand they endorse, the question how impactful celebrities are on a brand they do not merely endorse but are actually part of, such as a movie, has remained largely overlooked. Although both use their star power by having their name and face appear on the product (Desai & Basuroy, 2005), actors who are part of the film also co-create the product, whereas endorsers merely take part in the marketing effort once the product is finished (Wadhera & Chawla, 2015). In an industry as risky as the movie business with billions of dollars at stake, it is relevant for producers and marketers worldwide to gain new insights in the benefits and risks that come with signing actors.
Research on the impact of an actor’s reputation has predominantly focused on celebrity endorsement and found an actor can both help and hurt the brand they endorse. Celebrities with a good reputation have been found one of the most powerful tools of advertising (Atkin & Block, 1983), as their presence helps a persuasive message stand out (Muda, Musa, & Putit, 2012). However, research also emphasized that celebrity endorsement backfires when endorsers get caught in a scandal, as many brands have lost significantly in market value once their endorser’s scandal got publicized (Amos, Holmes, & Strutton, 2008). Nevertheless, academic research has not investigated whether these benefits and risks also apply for actors who are part of a film. In fact, no research has yet made a distinction between the two.
Rather, research that focuses on actors’ influence on films has found that high profile actors have a box office draw (Karniouchina, 2011). The Academy Awards – the most prestigious and recognized film prize in the world – can further influence box office revenue,
as the award ceremony gives worldwide publicity (Deutchert, Adjamah, & Pauly, 2005). However, while academia has substantially focused on the impact of the Oscars on a film’s success and made a distinction between Oscar-winning leading actors and supporting actors, it has simultaneously eclipsed the impact that an Oscar winner has years after his or her win, and overlooked the distinction between an actor and an endorser. Nevertheless, once an actor wins an Academy Award, this will be exploited in most films’ advertising, as trailers, posters, synopses, and articles use the tagline “Academy Award winner.” Despite this common use, no research has been conducted to investigate to what extent these taglines influence the
moviegoing public, and whether there is a difference between those who co-create the product and those who merely endorser a film.
Similarly, the influence of scandalous actors in Hollywood’s billion-dollar industry has largely remained overlooked, even though producers regularly have to determine whether their scandalous actor has the potential to damage their product. This was recently
demonstrated with the Hollywood film All the Money in the World. After sexual assault allegations were made against Kevin Spacey, it was decided that all his scenes would be reshot with a different actor, leading to nine days of reshooting and additional costs of approximately 10 million dollars (Nordine, 2017). Though this decision is unprecedented for Hollywood, research on whether this action is needed lacks still. Despite the current media environment’s growing interest to cover celebrities’ private lives, no studies have yet investigated whether a scandalous actor is as detrimental to a film’s success as an endorser (Muda et al., 2012).
Therefore, this research aims to investigate whether the benefits and risks of celebrity endorsers also hold for actors who are part of a film by taking both the reputation and
involvement of the actor into account. These findings can extend existing literature on celebrity endorsement by building on association theory (Spry, Pappu, & Cornwell, 2011),
match-up theory (Treme, 2010), and spillover effects (Votola & Unnava, 2006), as it adds the dimension of an actor’s involvement. Simultaneously, the focus on moviegoers’ intentions and attitudes rather than a film’s revenue allows this study to refute alternative explanations common in research on star power (Treme, 2010). Thus, this study aims to extend existing literature with the following research question: To what extent does framing a film actor as either an award-winner or a scandalous actor influence the image of the film? And does this depend on the level of involvement of the actor?
Theoretical Background The impact of an actor on a film’s image
With only a few weeks to make profit, constantly changing competitors, and a product that is difficult to observe prior to purchasing, the motion picture industry is a risky business (De Vany & Walls, 1999). Therefore, scholars have aimed to reduce the risks film studios face by examining potential determinants of revenue, such as ratings (Litman, 1983), budgets (Basuroy, Chatterjee, & Ravid, 2003), word-of-mouth (Hennig-Thurau, Wiertz, & Feldhaus, 2015), and reviews (Eliashberg & Shugan, 1997). Furthermore, a substantial amount of studies has investigated the impact A list stars – often operationalized as actors listed in the top 100 most powerful people in Hollywood – can have on a film’s success (De Vany & Walls, 1999). Although two of the highest-grossing films of all time – Avatar and Titanic – did not include any A list stars to drive the success of these motion pictures at the time (Elberse, 2007), many scholars do assign a critical role to a film’s cast. For instance, Austin (1989) demonstrated that moviegoers look for star presence when exposed to information about a movie. Therefore, replacing an actor with an A-list star can increase box office revenue with approximately 17 million dollars (Nelson & Glotfelty, 2012).
buzz being the first and marking power being the second. According to Karniouchina (2011), the mere presence of a star generates buzz and consequently breaks through the clutter. This buzz has been linked to enhanced box office results, as well as enhanced attitudes toward the film (Karniouchina, 2011). Besides drawing power, stars can also influence the success of a film by their “marking power” (Albert, 1998, p. 251). While every film has endless variables that create its quality, such as themes and dialogues, it is difficult for moviegoers to establish what an upcoming movie needs to have to meet their quality standard (Albert, 1998). A notable and objective constant between films is the cast, which means that actors can mark a certain film type (Albert, 1998). Stars’ marking power does not merely impact moviegoers’ purchase intention, but even enhances attitudes toward the film (Albert, 1998). Nevertheless, scholars have also acknowledged that this influence depends on numerous factors, such as the number of negative reviews (Basuroy et al., 2003). Therefore, variations in A list stars’ bankability may also rely on award-winning performances or reputation tainting scandals (Nelson & Glotfelty, 2012).
The impact of an award-winning actor
Motion pictures are hedonic rather than utilitarian products, which means their consumers seek enjoyment (Clement, Proppe, & Rott, 2007). Simultaneously, movies are experience products, which makes it hard for consumers to determine its quality prior to consumption (Leenders & Wijnberg, 2008). Therefore, word-of-mouth and critics’ reviews are instrumental to a film’s success (Albert, 1998).1
Research suggests that critics have three effects: the promotion effect – initiating consumer attention, – the information effect – reducing moviegoers’ uncertainty with information, – and the opinion effect – signaling
1
The importance of critics’ reviews is emphasized by the finding that a moviegoer will solely purchase a ticket in case that
E (bn+1 | sn) > c,
in which the first term stands for the expected benefits that a certain film will bring a moviegoer, and sn stands
for the set of information one has received from word-of-mouth and reviews. The costs of the film are included as c, which stands for the ticket price (De Vany & Walls, 1996, p. 1499). A moviegoer will thus purchase a ticket when positive information outweighs the costs of a ticket.
quality with a review (Clement et al., 2007). Movie awards, and particularly the Academy Awards, have the same characteristics. First, the ceremony draws an annual global audience of several hundred million viewers, and the winners make headlines around the world. Second, studios already reduce consumer uncertainty by providing trailers, but the ‘Oscar buzz’ provides additional information (Clement et al., 2007). Third, attribution theory explains that moviegoers only take on information from opinion leaders (Leenders &
Wijnberg, 2008), but since the Oscars are considered the most recognized and prestigious film prize in the world as all voting members are film-industry professionals, it is found a credible source to signal quality. Indeed, scholars have predominantly found a positive relation
between an Academy Award and a film’s post-award revenue (Terry, Butler, & De’Armond, 2005). In fact, Litman (1983) demonstrated that an actor’s win in the best actor categories can add $16.3 million to a film’s box office revenue. This suggests that the recognition of an actor’s artistic reputation during such a global event (Elberse, 2007) does not only fuel positive communication, but even consumer purchase intentions (Krauss, 2008).
A literature review further implies that the widely used tagline “Academy Award winner” can have a persuasive effect and enhance moviegoers’ evaluation of the film (De Vany & Walls, 1996). The mere exposure effect explains that one unconsciously becomes more favorable toward a stimulus with repeated exposure (Matthes, Schemer, & Wirth, 2015), thus implying that one becomes more favorable toward a film during ‘Oscar buzz’. This also ties in with the uncertainty reduction theory, which explains that additional information limits the plausibility of alternatives and can enhance attitudes (Wyatt & Badger, 1990). This suggests that the “Academy Award winner” tag can decrease uncertainty and so enhance attitudes toward a film prior to experiencing it.
Moreover, scholars do not merely imply that the Oscars enhance attitudes toward a film, but also toward the award-winning actor. Research in various fields has found a positive
relationship between one’s status and perceived expertise, which has been explained with two stages (Kovács & Sharkey, 2014). The first stage describes that the boost of a star’s artistic reputation revealed through an award generates attention and allows an actor to be included in the consideration set of a moviegoer (Kovács & Sharkey, 2014). Once included in this set, it is likely for an award-winning actor to be evaluated positively in the second stage, since moviegoers tend to view an actor as the best in the business if others assign status to it as well (Kovács & Sharkey, 2014). However, this evaluation is not limited to professional expertise, as one’s consistency bias leads one to evaluate other attributes of that actor, such as
attractiveness, in accordance to their prior views (Kovács & Sharkey, 2014). Therefore, this consistency bias can spill over to other attributes, even when these are unrelated (Ahluwalia, Unnava, & Burnkrant, 2001). In contrast, however, scholars have also identified negative side effects of winning an award. Consumers value exclusivity and want to distinguish themselves from the masses, which can make the sudden increase in popularity “off-putting” (Kovács & Sharkey, 2014, p. 1). Thus, while most studies support the claim that winning an award enhances attitudes, some evidence for backfiring has been found.
The impact of a scandalous actor
Although many filmmakers argue that buzz is instrumental in gaining revenue, they also acknowledge that the statement ‘there is no such thing as bad publicity’ does not hold in the movie industry (Karniouchina, 2011). Nonetheless, scholars imply that a scandal is not necessarily detrimental to a film’s success. Rather, the star system that developed with the rise of Hollywood in the 1920s has made people interested in the lives of movie stars (Brants, Hermes, & Van Zoonen, 1997). In many cases, stars’ relationships, problems, and scandals have “intensified the public’s interest in the relationship between [an actor’s] private life and his films” (Girgus, 2002, p. 149).
implies that negativity bias can harm the evaluation of the film. Negative information around a film has been found to have a stronger influence than positive marketing, as negative news is different from the norm and catches attention (Hennig-Thurau et al., 2015). Although publicity around a scandalous actor does not necessarily include negative information about the movie itself, even mixed news has shown to worsen one’s attitudes (Wyatt & Badger, 1984). This is consistent with the associative network model, a framework that demonstrates that both an actor and a film are initially unconnected, but the actor’s involvement in the film creates a link between the two. Consequently, a strong connection between an actor and a film can occur, whereby information about the one influences the other (Spry et al., 2011).
Furthermore, a scandal can also negatively impact moviegoers’ attitudes toward the actor. According to Hu (2016), scandals can negatively impact audiences’ parasocial
interaction (PSI) with an actor. PSI is often defined as a pseudo-relationship between an actor and a spectator that makes the spectator see the actor as a friend (Moyer-Gusé, 2008). In fact, the stronger the parasocial relationship of the viewer with the performer, “the more parasocial breakup the participants experienced after the introduction of [a] scandal” (Hu, 2016, p. 217). In addition, spillover theory explains that people who are lowly committed to an actor will not solely like the person less when caught in a scandal, but will – unconsciously – also think he is less trustworthy or skilled, as negativity can spillover to other attributes (Ahluwalia, 2000). Fans of the actor, however, will compensate by evaluating the actor’s other attributes more positively (Ahluwalia, 2000).
Thus, based on academic literature, it is hypothesized that moviegoers’ purchase intention (H1a), attitude toward the film (H1b), and attitude toward the actor (H1c) will be higher when an actor is framed as an award-winner compared to when an actor is framed as a scandalous celebrity.
Distinguishing level of involvement
While an actor’s main job is to depict a character on screen, these celebrities are also in demand when it comes to endorsing a product. For instance, Jennifer Aniston is known as actress of films as Derailed and Cake but has also been hired to endorse brands as
SmartWater and Emirates. In fact, Aniston remains one of the highest-paid actresses due to all her endorsement deals, as she made over 10 million dollars by becoming the face of Aveeno Skincare alone (Vargas, 2017). According to scholars (Wadhera & Chawla, 2015), these big contracts can be justified by star power. Similar to their presence on the big screen, celebrity endorsers catch more attention than marketing messages without stars, while simultaneously increasing consumers’ purchase intentions and enhancing attitudes toward the brand they endorse (Wadhera & Chawla, 2015). This is consistent with classical conditioning that explains that one stimulus (the celebrity) benefits another stimulus (the film) as it builds an association between the two through repeated exposure (Spry et al., 2011). Moreover, the Elaboration Likelihood Model shows consumers can be persuaded with peripheral cues, such as a celebrity appearance, when one is not motivated or does not have the ability to process the persuasive message (Petty & Cacioppo, 1986). Thus, scholars agree that celebrity endorsers attract attention, reduce uncertainty, and encourage one to purchase a product or brand (Spry et al., 2011). Despite the similarities between A list actors and endorsers, however, there are some key differences.
Although academic literature does not make an explicit distinction between movie stars who are part of a film and movie stars who merely endorse a film, the two differ in their involvement in the product. While actors use their star power by having their name and face appear on posters, trailers, and merchandise, they also have a substantial role in the final product, whereas endorsers merely take part in the marketing effort once the product is finished (Wadhera & Chawla, 2015). Accordingly, consumers only learn about Hugh
Jackman’s endorsement of the film Deadpool 2 in a commercial, while he sits down with newspaper journalists, walks the red carpet, and appears on international talk shows to discuss his latest film The Greatest Showman.
Actors versus endorsers in their impact on a film’s image
Similar to the case of actors, scholars suggest that the reputation of an endorser has consequences on consumer behavior and attitudes. For instance, the source credibility model indicates that an endorser should fit the product, as the association between stimuli grows stronger when a brand is endorsed by a credible endorser (Spry et al., 2011). Therefore, having won an award does not merely make the endorser stand out as one of the best in the business, but also adds credibility. Accordingly, news around Tiger Woods’ win of a golf tournament has directly been linked to increased sales and a more favorable perception of Nike (Farrell, Karels, Motfort, & McClatchey, 2000), while scandals have proven to damage an endorser’s credibility as well as the endorsed brand (Koo, Ruihley, & Dittmore, 2012). For instance, Tiger Woods costed Nike an estimated loss of “$1.7 million in sales and 105.000 customers” when his cheating scandal got publicized (Chadwick, Arthur, & Beech, 2017, p. 89).
Despite these results, the impact of winning an award or being caught in a scandal is suggested to be more profound for actors than for endorsers. While both actors and endorsers have drawing power as they can help a message break through the clutter, only actors
involved in a movie have marking power (Albert, 1998). Additionally, while both Oscar-winning actors and endorsers are deemed experts in the field, levels of expertise – and thus levels of persuasion – do differ (Erdogan, Baker, & Tagg, 2001). Since Oscar-winning actors have more knowledge about the film as they worked on it, and since they have had the ability to use their knowledge in the production phase, they can be deemed to have a higher level of expertise. In case of a scandal, however, research suggests that an actor’s trustworthiness
decreases and so harms moviegoers’ purchase intentions (Karniouchina, 2011), whereas research found that a celebrity’s trustworthiness does not directly impact purchase intentions when it involves an endorser (Ohanian, 1991).
Furthermore, research implies that the reputation of an actor has a greater impact on moviegoers’ evaluation of the film than the reputation of an endorser. Within the context of associative networks, celebrities and brands that appear together become part of the same set in consumers’ minds (Till, 1998). Although associations can be built with just one pairing, classical conditioning states repetition is instrumental for this mechanism (Till, 1998). Since actors are involved in both the marketing effort and the production, the associative network model suggests that there is a stronger unconscious link between a film and an actor than between a film and an endorser (Spry et al., 2011). Moreover, the spillover effect suggests that news about one attribute of an individual can transfer from the celebrity to the associated brand (Votola & Unnava, 2006). Nevertheless, there are degrees of spillover effects, which depends on how culpable consumers find the endorsed brand (Votola & Unnava, 2006). Since the actor is more involved, it can be expected that consumers find the brand of a scandalous endorser less culpable than those who spent months working on a film together with a scandalous actor.
Additionally, academic literature implies that level of involvement influences the effect a celebrities’ reputation has on moviegoers’ attitudes toward an actor. Publicity around an actor influences the popular appeal of the movie star (Treme, 2010). Namely, Adler (1985) noted that moviegoers’ attitudes toward an actor enhance over a wide range – from likability to expertise – as consumers learn more about the star. Since more news appears around an actor who stars in the film rather than endorses it, and since actors can show their personality during several media interactions as opposed to endorsers (Treme, 2010), it can be expected that the effect is more profound for actors. Moreover, films have a more dominant position in
an actor’s career, which implies articles on a scandalous actor will mention the actor’s past films over the films the actor endorsed. For instance, articles covering James Franco and his scandal predominantly mentioned him as actor of The Disaster Artist and 127 Hours, though he was rarely mentioned as endorser of Gucci. Although Treme (2010) noted that a star’s personality can act as a countervailing force against negative and recent publicity, studies predominantly found that the repetitive use of one frame can lead to enduring negative effects on consumer attitudes (Lecheler, Keer, Schuck, & Hänggli, 2015).
Thus, it is hypothesized that moviegoers’ purchase intention (H2a), attitude toward the film (H2b), and attitude toward the actor (H2c) will be higher when an actor is framed as an award-winner and lower when the actor is framed as a scandalous celebrity, but this effect will be more pronounced for actors who star in the film than actors who endorse the film.
Conceptual model
Methods Sample and design
Over a course of three weeks, 204 people participated in an online experiment (Mage =
25.5; 68.1% female, 29.9% male, 0.5% non-binary).2 Participants were randomly assigned to one of the conditions of a 2 (Reputation of the actor: award-winning vs. scandalous) X 2
2
Four responses had been excluded as all questions were answered with the middle option on the 7-point scales in under a minute.
(Involvement of the actor: actor vs. endorser) between-subjects design. The study
predominantly included students, which is in accordance with a report provided by the Motion Picture Association of America (2016) that showed people between the ages of 18 and 24 make up the biggest part of the moviegoing public. However, since participants were recruited through personal contacts and via Facebook groups where moviegoers of all nationalities and ages gather to discuss films, the study represents a large part of the moviegoing public as reported by film industry, with ages ranging from 19 to 67. Only people under the age of eighteen were excluded for ethical reasons. Since this research focuses on moviegoers specifically, another criterion was that participants watched at least one movie at the cinema in the past twelve months. In line with industry practice (Chakravarty, Liu, & Mazumdar, 2010), the participants predominantly included infrequent moviegoers (83.3% versus 16.7% frequent moviegoers) who were rather knowledgeable, interested, and informed in the film industry (M = 4.59, SD = 1.27).
Procedure
Participants were informed about the consent procedure and told the study was about film marketing before filling out the Qualtrics survey. The survey guaranteed anonymity and a chance of winning a movie ticket as an incentive. After being randomly assigned to one of the four experimental conditions, participants were asked about their purchase intentions for the film, as well as their attitude toward the film and the actor. Subsequently, participants were asked to answer four manipulation checks, and six questions about their personal situation, such as their moviegoing tendencies, interest in the film industry, and
demographics. Finally, all participants were debriefed and informed about the true purpose of the study and the fact that the articles did not reflect the truth.
Stimulus material
emphasis of this research was placed on the comparability across the experimental conditions, and, consequently, only the title and the last sentences differed per article. The
“award-winning” conditions emphasized that the actor had previously won an Oscar, while the
“scandal” conditions emphasized recent sexual harassment allegations made against the actor. The articles also emphasized the actor’s level of involvement to distinguish between an actor starring in the film and an actor endorsing the film. Furthermore, the articles were based on real news reports to decrease the artificiality of the stimuli. In fact, the articles were created by merging various segments that have appeared on Casey Affleck – a recent Oscar winner and accused actor – from sources as Variety, Deadline, and Time, while copying the exact style of an online article from the Hollywood Reporter (see Appendix I).
All treatments centered around Jake Gyllenhaal, an American actor who is known for films as Brokeback Mountain and Nightcrawler. Gyllenhaal was nominated for an Academy Award in 2006 but never won, and he had not been involved in a scandal of any kind at the time of the study. Furthermore, although Gyllenhaal is a household name in Hollywood, it was believed that he was not as famous as Leonardo DiCaprio or Brad Pitt, which means that framing him in a scandal might be perceived more realistically than these other A list actors. Additionally, all treatments discussed a real film – the upcoming drama Backseat – to enhance the authenticity of the stimulus material. Still, however, only few was known about Backseat at the time of conducting the study, which allowed this research to frame Gyllenhaal both as an actor of the film as well as an endorser, though he is neither. Nevertheless, a pretest was conducted to test the appropriateness of Jake Gyllenhaal and the film Backseat.
Pretest
Since participants may become sensitized to the purpose of the experiment, only people who were comparable to members of the target group but would not be part of the study were recruited to participate (Bryman, 2015). Therefore, 40 students from the
University of Amsterdam were exposed to one of the four conditions. The results proved that the stimulus worked as intended (see Appendix II), and no one questioned Jake Gyllenhaal as an endorser, an Oscar-winner, nor a scandalous actor of the film Backseat. Nevertheless, small changes to the manipulation were made to emphasize the most important information and to make the distinctions between the conditions even stronger.
Measurements Dependent variables
Purchase intention. Moviegoers’ willingness to buy a ticket for the film Backseat was measured by a 7-point bipolar scale with three items (Holzwarth, Janiszewski, &
Neumann, 2006; see Appendix III for an overview of the items). The overall purchase intention was assessed by averaging the items, resulting in a Cronbach’s alpha of .91 (M = 4.22, SD = 1.61).
Attitude toward the film. Moviegoers’ evaluation of the film Backseat before watching was measured with a 7-point bipolar scale that consisted of four items (D’Astous & Touil, 1999; see Appendix III), resulting in a Cronbach’s alpha of .89 (M = 4.26, SD = 1.27). Attitude toward the actor. The source attractiveness model – which is said to include likability, similarity, and familiarity (Roy, 2006) – was used as foundation to assess moviegoers’ attitude toward the actor. Therefore, six items measured likability (Yilmaz, Telci, Bodur, & Iscioglu, 2011), resulting in a Cronbach’s alpha of .94 (M = 4.70, SD = 1.31). Similarity was assessed by three items on a 7-point Likert scale (Feick & Higie, 1992) with a Cronbach’s alpha of .90 (M = 3.24, SD = 1.24). Familiarity was assessed with three items on a 7-point bipolar scale (Erfgen, Sattler, & Villeda, 2016), resulting in a Cronbach’s alpha of .91 (M = 4.88, SD = 2.04). Furthermore, the actor’s expertise was measured on a 7-point bipolar adjective scale with 6 items (Taylor, Halstead, & Haynes, 2010; Ohanian, 1990; see Appendix III), resulting in a Cronbach’s alpha of .93 (M = 4.99, SD = 1.23).
Manipulation checks
Four manipulation checks were implemented. The first question investigated the actor’s reputation and asked: “How do you perceive the current reputation of Jake Gyllenhaal as described in the article?” with answer options ranging from 1 (very negative) to 5 (very positive). Second, participants were asked to rate the involvement of the actor with the question: “How do you perceive the role of Jake Gyllenhaal in the upcoming film Backseat?” with answer options ranging from 1 (very uninvolved) to 5 (very involved). Finally,
participants were asked to respond to statements stating Jake Gyllenhaal has won an Oscar and had been in a scandal, both with a 5-point Likert answer scale ranging from 1 (strongly disagree) to 5 (strongly agree).
Control variables
Moviegoing frequency. The frequency with which one goes out to see a movie was operationalized by means of the question: “In the past two months, I have watched a movie in the cinema…” with answer options ranging from 1 (once or twice) to 4 (10 or more times), based on a recent study (Chakravarty et al., 2010). According to industry practice (MPAA, 2016), this question identifies two groups of moviegoers3: frequent moviegoers who go to the movies at least three times in two months, and infrequent moviegoers who go less than three times in two months. The distinction between the groups was made as frequent moviegoers can be expected to have higher purchase intentions than people who rarely go out to see a movie. Furthermore, infrequent moviegoers may rely more on an article’s chosen frame as they are less familiar with actors or genre expectations (Chakravarty et al., 2010).
Movie interest. How knowledgeable, interested, and informed one would say he or she is with regards to actors, movies, and the film industry in general. The answer options ranged from 1 (not at all) to 7 (extremely). The overall movie interest was assessed by
averaging the items, resulting in a Cronbach’s alpha of .87 (M = 4.59, SD = 1.27; see Appendix IV). This control variable was included as people who are interested in the film industry may know more about the Academy Awards and thus attach more weight to the prize than those who are less interested.
Results
Randomization and manipulation checks
Randomization checks indicated there were no issues with randomization for the relevant variables Moviegoing frequency, Movie interest, and Age, Gender, and Nationality (see Appendix V), while the manipulation checks proved participants had perceived the stimulus materials as intended (see Appendix VI). Furthermore, the control variable Movie frequency was not significantly correlated to the outcome variables Purchase intention, r = .08, p = .284, Attitude toward the film, r = .05, p = .444, nor Attitude toward the actor, r = .06, p = .426. Analysis revealed, however, that there was a significant small to medium positive correlation between Movie frequency and one’s purchase intention, r = .23, p = .044, though not between the control variable and one’s attitude toward the film, r = .09, p = .422, nor attitude toward the actor, r = .13, p = .241. Therefore, the control variable of movie frequency was included in the single MANOVA that was used to test the hypotheses.
Normality tests
The normality of the data was tested as part of the preparatory analyses and demonstrated that purchase intention (condition 1 and 4) and attitude toward the film
(condition 1 and 4) significantly differed from normal distribution (see Table 1). The sample size of this research could have influenced the test and made even small effects significant (Field, 2015); a view that was further confirmed by looking at the P-P plots and Q-Q plots. Therefore, no additional steps were taken.
Table 1
Normal Distribution (Shapiro-Wilk test)
Note: * p < .05.
Method of analysis
To test the hypotheses, one multivariate analysis of variance (MANOVA) was used. This way, the dependent variables Purchase intention, Attitude toward the film, and Attitude toward the actor could be tested, while investigating an interaction effect between the independent variables Reputation and Involvement simultaneously (Field, 2015). The control variable Movie frequency was included as covariate, but the MANOVA revealed that the frequency with which one sees a movie did not significantly influence purchase intention, F (1, 197) = 0.14, p = .708, attitudes toward the film, F (1, 197) = 0.09, p = .767, nor attitudes toward the actor, F (1, 197) = 0.01, p = .918. Furthermore, the Levene’s test proved
nonsignificant for purchase intention (p = .084), attitude toward the film (p = .502), and
Dependent variable Condition Statistic df
Purchase intention Actor, Oscar .946* 53
Actor, scandal .961 54
Attitude toward the film
Attitude toward the actor
Endorser, Oscar Endorser, scandal Actor, Oscar Actor, scandal Endorser, Oscar Endorser, scandal Actor, Oscar Actor, scandal Endorser, Oscar Endorser, scandal .962 .937* .944* .973 .965 .947* .958 .972 .958 .972 49 48 53 54 49 48 53 54 49 48
attitude toward the actor (p = .668), meaning the assumption of homogeneity of variance was met.
The role of an actor’s reputation
Results from the MANOVA revealed that the reputation of an actor had a significant effect on all three dependent variables. Moviegoers that were exposed to an article on an Oscar-winning actor (M = 5.04, SD = 1.41) had significantly higher purchase intentions for the mentioned film than those exposed to an article that framed the actor as scandalous (M = 3.45, SD = 1.54), F (1, 197) = 23.85, p < .001, and this represented a medium-sized effect, ƞ² = .11. The same applied for moviegoers’ attitude toward the film, as those exposed to an Oscar-winning actor (M = 4.79, SD = 1.26) rated the film significantly more favorable than those exposed to a scandalous actor (M = 3.67, SD = 1.08), F (1, 197) = 33.59, p < .001, and this represented a medium-sized effect, ƞ² = .15. Similarly, results also revealed that the reputation of an actor had a significant effect on moviegoers’ attitude toward the star, F (1, 197) = 54.11, p < .001, and represented a large-sized effect, ƞ² = .22. The group exposed to a scandalous actor (M = 3.83, SD = 1.01) perceived the actor significantly worse than those exposed to an Oscar winner (M = 5.09, SD = 1.04). Therefore, the first hypotheses (H1a, H1b, and H1c) were supported.
The role of an endorser’s reputation
Further analysis with the MANOVA demonstrated that Oscar-winning endorsers did not enhance intentions and attitudes in comparison to a scandalous endorser. Namely, moviegoers who were exposed to an Oscar-winning endorser (M = 4.47, SD = 1.38) rather than a scandalous endorser (M = 3.96, SD = 1.68) did not have significantly higher purchase intentions than the latter, F (1, 197) = .02, p = .889, and this represented a small-sized effect, ƞ² < .01. The same applied for moviegoers’ attitude toward the film, as those exposed to an Oscar-winning endorser (M = 4.69, SD = 1.11) rated the film not significantly more favorable
than those exposed to a scandalous endorser (M = 3.86, SD = 1.27), F (1, 197) = .10, p = .788, and this represented a small-sized effect, ƞ² < .01. Results further revealed that the reputation of an endorser did not have a significant effect on moviegoers’ attitude toward the star, F (1, 197) < .01, p = .950, and represented a small-sized effect, ƞ² < .01. The group exposed to a scandalous endorser (M = 4.07, SD = 1.27) did not perceive the celebrity significantly worse than those exposed to an Oscar-winning endorser (M = 4.85, SD = 0.87).
Moderating role of involvement
The MANOVA further showed that there was an interaction effect between an actor’s reputation and involvement on purchase intention (see Figure 1), as that effect yielded a F-ration of F (1, 197) = 13.48, p = .016, and represented a small-sized effect, ƞ² = .03.
Figure 1. Consumers’ purchase intention as a function of reputation and involvement
Accordingly, an actor’s favorable reputation established through an Oscar led to moviegoers’ greater purchase intentions when the actor starred in the movie (M = 5.04, SD = 1.41) than when the actor merely endorsed the movie (M = 4.47, SD = 1.38). An independent t-test showed this difference was significant, t (105) = 5.51, p < .001, and represented a small-sized effect, d = 0.37. A scandalous actor who stars in the movie also brings more risks, as purchase intentions were significantly lower for moviegoers when it involved an actor who starred in
the film (M = 3.45, SD = 1.54) than an endorser (M = 3.96, SD = 1.68). Indeed, a second independent t-test4 showed this difference was also significant, t (95) = 4.18, p = .044, and represented a small-sized effect, d = 0.32. Therefore, hypothesis H2a was supported. No significant interaction effect was found on moviegoers’ attitude toward the film, F (1, 197) = 0.94, p = .413, ƞ² < .01 nor attitude toward the actor, F (1, 197) = 2.75, p = .093, ƞ² = .01. Accordingly, hypotheses H2b and H2c were rejected.
Conclusion
Research on film marketing has shown that star power is not a hallmark of a film’s success but does play an important role in attracting an audience for these hedonic products. The purpose of this study was to build on this existing literature and investigate the extent to which an award-winning or scandalous actor affects moviegoers’ purchase intentions, as well as their evaluation of the film and the actor. This study further investigated the differential impact of an actor who is part of the film versus an actor who solely endorsers a film, as previous studies have not yet taken a celebrity’s level of involvement into account. The results demonstrated that an actor’s reputation does influence moviegoers’ intentions and attitudes, as consistent with the first hypotheses. This suggests that the Academy Awards can give salient signals of quality by awarding Oscars for outstanding performances, thereby drawing moviegoers and enhancing the evaluation of the film and actor. In contrast, negative news around an actor in combination with an upcoming film proved to negatively impact moviegoers’ purchase intentions and could even worsen a film’s evaluation. In fact, most moviegoers exposed to a scandalous actor evaluated the movie as worse than neutral and indicated they were not interested in purchasing a ticket. This suggests that moviegoers tend to view the artist and the artwork as intertwined and are unable to
4
The Levene’s test for this independent t-test was found significant, meaning the assumption of homogeneity of variance had been violated.
distance an actor’s role from his or her personal life. Consistent with the hypothesis, a scandal also showed to spill over to the actor, thus worsening the actor’s perceived attractiveness and professional expertise.
The interaction effect between an actor’s reputation and involvement on moviegoers’ purchase intentions showed that the effect of an actor’s reputation is especially profound when an actor starred in the movie rather than endorsed it. An Oscar-winning actor led to greater purchase intentions when the actor starred in the movie than when the actor endorsed the movie, while a scandalous actor who stars in the movie also brings more risks than an endorser. Although it was expected that this interaction also applied for moviegoers’ attitudes towards the film and the actor, results did not support these hypotheses.
Theoretical implications
The motion picture industry is still an uncertain business in which it is difficult to predict the success or failure of a film. Therefore, many scholars have investigated potential determinants of a film’s success by focusing on star power, but lack knowledge about the influence of an actor’s reputation. In contrast to previous claims (Austin, 1989), this research demonstrates that it is not solely star presence that influences moviegoers, but this influence depends on the reputation of the actor. Although the impact of Oscar winners on a film’s success has been the subject of numerous studies in the field of film marketing research (Terry et al., 2005; Litman, 1983), these studies have predominantly focused on box office revenue rather than taking the perspective of moviegoers to draw conclusions. Nevertheless, previous research (Treme, 2010) has already hinted at an alternative explanation for this financial success: the fact that Oscar winners have the opportunity to pick the best projects. Therefore, by taking the perspective of the moviegoer, this study is the first to support the notion that the “award-winning” tag increases a film’s success with experimental evidence. Furthermore, the start of the #MeToo movement in January has led to accusations
against more than a hundred high profile celebrities, and once again emphasized the need for more literature on the impact of scandalous actors on a film. In contrast to previous research (Girgus, 2002), this study found that scandals do not increase purchase intentions, and supports filmmakers’ claim that there is such thing as bad publicity in the movie industry (Karniouchina, 2011). By investigating moviegoers’ intentions and attitudes, this study builds on previous findings (Hennig-Thurau et al., 2015; Ahluwalia, 2000) and shows negativity around an actor can spill over to a film.
Besides advancing film marketing literature, this study also fills a gap in celebrity endorsement theory by extending the theoretical understanding of the phenomena. While literature has made no distinction, this study found a difference between those who solely endorse a product and those who are part of a product. The latter has a greater impact on moviegoers’ purchase intentions, both positive as well as negative. This is an important finding since this is not solely applicable to the film industry, but might also be generalizable to the music or fashion industry where celebrity artists co-create products. Furthermore, this research is the first to demonstrate that the reputation of an actor who is part of the product can help or hurt brand attitudes similar to endorsers, which is in line with celebrity
endorsement studies (Amos et al., 2008; Muda et al., 2012). Accordingly, by matching an endorser with the target audience, an endorser can improve the evaluations of a film to the same extent as the actors, though also bringing the same risks. Thus, this study builds on match-up theory (Treme, 2010), association theory (Spry et al., 2011) and spillover effects (Votola & Unnava, 2006) by adding the dimension of an actor’s involvement.
Limitations and directions for future studies
Despite this study’s scientific relevance, there are some limitations to the study. First, participants in this experiment solely learned about the upcoming film from one article. Although literature points at critic reviews and experiences from friends and family as vital in
one’s decision process, participants of this study could not base their intentions and
evaluations on other sources (De Vany & Walls, 1996). Second, this experiment did not allow multiple exposures to either the same frame or different frames. This is especially relevant for the interaction effect, as the difference between actors and endorsers could also be affected by the associative network (Till, 1998), in which repeated exposure is of importance (Lecheler et al., 2015). Lastly, this study used an Academy Award to signal a positive reputation and a sexual misconduct allegation as scandal, which raises the question whether other sources of positive and negative news around an actor lead to different effects.
Therefore, future research should investigate to what extent the results of this research hold for other forms of publicity around an actor. Positive news could also be an actor’s environmental work, while negative news can be a scandal that is more out of the hands of the actor, such as an iCloud hack. Additionally, future research could also focus on other
dimensions of hedonic products, such as books, comedy shows, and music. In all cases, awards play an important role and scandals have a worldwide reach. Moreover, since many scholars have mentioned the importance of critic reviews and word-of-mouth on moviegoers’ film evaluations (De Vany & Walls, 1996; Clement et al., 2007), future studies could also investigate how an audience’s film evaluation is impacted by their viewing experience as well as by the conversations they have about the film afterwards.
Managerial implications
This research is one of the first to investigate and find that the “award-winning” tag has a positive impact on moviegoers’ intentions and attitudes, even in an international sample. Therefore, this study confirms the important role that is attributed to Oscar winners, and proposes that film marketers continue to emphasize Oscar winners in their promotional campaigns. Furthermore, by uncovering an interaction effect, this research also points out that endorsers with a positive reputation can help push a movie further as well. This might be of
importance when a film’s cast does not produce sufficient favorable publicity themselves. Moreover, in an industry as risky as the movie business with billions of dollars at stake, this research can help practitioners make some thoughtful choices in the production phase. While it is already common for brands to add a moral clause to an endorser’s contract, these results point out that studios should undertake the same action to protect their projects when it involves actors. Film producers bring together hundreds of people as cast and crew, but it is essential that they acknowledge the detrimental impact one actor can have on the final product. For instance, Mel Gibson made a comeback as actor in Hollywood after his domestic abuse scandal, but he suffered one failed movie after another, such as Edge of Darkness and The Beaver. Although this study offers only initial results, these movies may not have flopped so badly had the film studios received more information on this topic, and not casted a
troubled but familiar face.
Finally, this research also ties in with recent discussions on whether moviegoers can distance the artist from the artwork. These ongoing debates have recently gained a lot of attention in international media, as a response to Hollywood’s decision to cut accused actors in the #MeToo tide from films and series. For instance, Netflix was soon to announce that Kevin Spacey would not return as President Frank Underwood in the highly popular series House of Cards, which led to many negative comments on Netflix’s social media accounts. Dutch newspapers such as De Volkskrant, and talk shows as De Wereld Draait Door and Pauw, as well as The New York Times and the Independent Spirit Awards, have all openly questioned whether a films’ audience can see a great actor as independent from his or her public persona. Although these conclusions are preliminary, it does demonstrate that
moviegoers do not see an actor as independent from his or her artwork. Therefore, the study backs up the unprecedented move from the producers of All the Money in the World and House of Cards that replaced or airbrushed their scandalous leading actor out of the story, and
may encourage producers to do the same in a similar situation in the future.
Adler, M. (1985). Stardom and talent. American Economic Review, 75, 208-212. Ahluwalia, R. (2000). Examination of psychological processes underlying resistance to persuasion. Journal of Consumer Research, 27(2), 217-232.
Ahluwalia, R., Unnava, H., & Burnkrant, R. (2001). The moderating role of commitment on the spillover effect on marketing communications. Journal of Marketing Research, 38(4), 458-470.
Albert, S. (1998). Movie stars and the distribution of financially successful films in the motion picture industry. Journal of Cultural Economics, 22, 249-270.
Amos, C., Holmes, G., & Strutton, D. (2008). Exploring the relationship between celebrity endorser effects and advertising effectiveness. International Journal of Advertising, 27(2), 209-234.
Atkin, C. & Block, M. (1983). Effectiveness of celebrity endorsers. Journal of Advertising Research, 23(1), 57-61.
Austin, B. (1989). Immediate seating: A look at movie audiences. Belmont, CA: Wadsworth. Basuroy, S., Chatterjee, S., & Ravid, S.A. (2003). How critical are critical reviews? The box office effects of film critics, star power, and budgets. Journal of Marketing, 67(4), 103-117.
Brants, K., Hermes, J., & Van Zoonen, L. (1997). The Media in Question: Popular Cultures and Public Interests. Thousand Oaks, CA: Sage Publications.
Bryman, A. (2015). Social Research Methods. Oxford, UK: Oxford University Press. Chadwick, S., Arthur, D., & Beech, J. (2017). International cases in the business of sport. Abingdon, UK: Taylor & Francis.
Chakravarty, A., Liu, Y., & Mazumdar, T. (2010). The differential effects of online word-of- mouth and critics’ reviews on pre-release movie evaluation. Journal of Interactive Marketing, 24, 185-197.
Clement, M., Proppe, D., & Rott, A. (2007). Do critics make bestsellers? Opinion leaders and the success of books. Journal of Media Economics, 20(2), 77-105.
D’Astous, A. & Touil, N. (1999). Consumer evaluations of movies on the basis of critics’ judgments. Psychology & Marketing, 16(8), 677-694.
Desai, K. & Basuroy, S. (2005). Interactive influence of genre familiarity, star power, and critics’ reviews in the cultural goods industry: The case of motion pictures.
Psychology & Marketing, 22(3), 203-223.
Deutchert, E., Adjamah, K., & Pauly, F. (2005). For Oscar glory or Oscar money? Journal of Cultural Economics, 29, 159-176.
De Vany, A. & Walls, W. (1996). Bose-Einstein dynamics and adaptive contracting in the motion picture industry. The Economic Journal, 106(439), 1493-1514.
De Vany, A. & Walls, D. (1999). Uncertainty in the movie industry: does star power reduce the terror of the box office? Journal of Cultural Economics, 23(4), 285-318.
Elberse, A. (2007). The power of stars: do star actors drive the success of movies? Journal of Marketing, 71(4), 102-120.
Elberse, A. (2012). The economic value of celebrity endorsements. Journal of Advertising Research, 52(2), 149-165.
Eliashberg, J. & Shugan, S. (1997). Film critics: influencers or predictors? Journal of Marketing, 61(2), 68-78.
Erdogan, B., Baker, M., & Tagg, S. (2001). Selecting celebrity endorsers: the practitioner’s perspective. Journal of Advertising Research, 41(3), 39-46.
Erfgen, C., Sattler, H., & Villeda, I. (2016). Effects of celebrity endorsers for multiple brands on attitudes and purchase intentions. Schmalenbach Business Review, 17(1), 49-67. Farrell, K., Karels, G., Motfort, K., & McClatchey, C. (2000). Celebrity performance and endorsement value: the case of Tiger Woods. Managerial Finance, 26(7), 1-15.
Feick, L. & Higie, R. A. (1992). The effects of preference heterogeneity and source characteristics on ad processing and judgments about endorsers. Journal of Advertising, 21(2), 9-24.
Field, A. (2015). Discovering statistics using IBM SPSS statistics (4th ed.). London, UK: SAGE Publications Ltd.
Girgus, S. (2002). The films of Woody Allen. Cambridge, UK: Cambridge University Press. Hennig-Thurau, T., Wiertz, C., & Feldhaus, F. (2015). Does Twitter matter? The impact of microblogging word of mouth on consumers’ adoption of new movies. Journal of the Academy of Marketing Science, 43(3), 375-394.
Holzwarth, M., Janiszewski, C., & Neumann, M. (2006). The influence of avatars on online consumer shopping behavior. Journal of Marketing, 70(4), 19-36.
Hu, M. (2016). The influence of a scandal on parasocial relationship, parasocial interaction, and parasocial breakup. Psychology of Popular Media Culture, 5(3), 217-231. Karniouchina, E. (2011). Impact of star and movie buzz on motion picture distribution and box office revenue. International Journal of Research in Marketing, 28, 62-74. Koo, G., Ruihley, B., & Dittmore, S. (2012). Impact of perceived on-field performance on sport celebrity source credibility. Sport Marketing Quarterly, 21, 147-158.
Kovács, B. & Sharkey, A. (2014). The paradox of publicity: how awards can negatively affect the evaluation of quality. Administrative Science Quarterly, 59(1), 1-33.
Krauss, J., Nann, S., Simon, D., & Fischbach, K. (2008). Predicting movie success and Academy Awards through sentiment and social network analysis. 16th European Conference on Information Systems, 116.
Lecheler, S., Keer, M., Schuck, A., & Hänggli, R. (2015). The effects of repetitive news framing on political opinions over time. Communication Monographs, 82(3), 339-358. Leenders, M. & Wijnberg, N. (2008). Why some awards are more effective signals of quality
than others: a study of movie awards. Journal of Management, 34(1), 25-54. Litman, B. (1983). Predicting success of theatrical movies: an empirical study. Journal of Popular Culture, 16(4), 159-175.
Matthes, J., Schemer, C., & Wirth, W. (2015). More than meets the eye: investigating the hidden impact of brand placements in television magazines. International Journal of Advertising, 26(4), 477-503.
Motion Picture Association of America. (2016). Theatrical market statistics. Retrieved from https://www.mpaa.org/wp-content/uploads/2017/03/MPAA-Theatrical-Market- Statistics-2016_Final-1.pdf
Moyer-Gusé, E. (2008). Toward a theory of entertainment persuasion: explaining the persuasive effects of entertainment-education messages. Communication Theory, 18(3), 407-425.
Muda, M., Musa, R., & Putit, L. (2012). Breaking through the clutter in media environment: how do celebrities help? Procedia – Social and Behavioral Sciences, 42, 374-382. Nelson, R. & Glotfelty, R. (2012). Movie stars and box office revenues: an empirical analysis. Journal of Cultural Economics, 36(2), 141-166.
Nordine, M. (2017). ‘All the Money in the World’ reshoots will probably cost more
than $10 million. Retrieved from http://www.indiewire.com/2017/11/all-the-money- in-the-world-reshoots-kevin-spacey-1201896732/
Ohanian, R. (1990). Construction and validation of a scale to measure celebrity endorsers’ perceived expertise, trustworthiness, and attractiveness. Journal of Advertising, 19(3), 39-52.
Ohanian, R. (1991). The impact of celebrity spokesperson’s perceived image on consumer’s intention to purchase. Journal of Advertising Research, 31(1), 46-52.
Consumer Research, 11(1), 673-675.
Roy, S. (2006). An exploratory study in celebrity endorsements. Journal of Creative Communications, 1(2), 139-153.
Spry, A., Pappu, R., & Cornwell, T. (2011). Celebrity endorsement, brand credibility and brand equity. European Journal of Marketing, 45(6), 882-909.
Taylor, V., Halstead, D., & Haynes, J. (2010). Consumer responses to Christian religious symbols in advertising. Journal of Advertising, 39(2), 79-92.
Terry, N., Butler, M., & De’Armond, D. (2005). The determinants of domestic box office performance in the motion picture industry. Southwestern Economic Review, 32(1), 137-148.
Till, B. (1998). Using celebrity endorsers effectively: lessons from associative learning. Journal of Product & Brand Management, 7, 400-409.
Treme, J. (2010). Effect of celebrity media exposure on box-office performance. Journal of Media Economics, 23(1), 5-16.
Vargas, K. (2017). How much is Jennifer Aniston’s net worth? Here’s how she made her millions. Retrieved from http://www.entitymag.com/jennifer-aniston-net-worth/ Votola, N. & Unnava, H. (2006). Spillover of negative information on brand alliances. Journal of Consumer Psychology, 16(2), 196-202.
Wadhera, R. & Chawla, N. (2015). Celebrity endorsement and attitude: a study to assess the impact of celebrity endorsement on attitude of consumers. International Journal of Scientific & Technology Research, 4(8), 111-114.
Wyatt, R. & Badger, D. (1984). How reviews affect interest in and evaluations of films. Journalism Quarterly, 61, 874-878.
Wyatt, R. & Badger, D. (1990). Effects of information and evaluation in film criticism. Journalism Quarterly, 67(2), 359-368.
Yilmaz, C., Telci, E., Bodur, M., & Iscioglu, T. E. (2011). Source characteristics and advertising effectiveness. International Journal of Advertising, 30(5), 889-914.
Appendix I: Experimental stimuli Condition 1. Oscar-winning actor
Condition 2. Scandalous actor
Appendix II: Pretest Table 2
Descriptive Statistics of the Pretest
Note: the variables were measured on a 5-point scale in which the answer options ranged from 1 (very negative) to 5 (very positive), 1 (very uninvolved) to 5 (very involved), 1 (strongly disagree) to 5 (strongly agree), 1 (strongly disagree) to 5 (strongly agree).
Appendix III: Measurements Table 3
Measurements of the Latent Constructs Variable Condition 1 (Oscar & actor) Condition 2 (scandal & actor) Condition 3 (Oscar & endorser) Condition 4 (scandal & endorser) Sig. Perceived reputation of Jake
Gyllenhaal in the article
M = 4.11 SD = 0.33 M = 2.55 SD = 0.88 M = 4.00 SD = 0.87 M = 2.56 SD = 1.01 .000 Perceived role of Jake
Gyllenhaal in the film
M = 3.67 SD = 0.71 M = 4.11 SD = 0.33 M = 2.33 SD = 1.12 M = 2.56 SD = 1.24 .000 “Jake Gyllenhaal is an Oscar
winner” M = 3.78 SD = 0.67 M = 2.22 SD = 0.83 M = 3.33 SD = 1.66 M = 2.67 SD = 1.00 .026 “Jake Gyllenhaal has been
involved in a recent scandal”
M = 2.00 SD = 0.71 M = 3.67 SD = 0.87 M = 2.44 SD = 1.13 M = 4.11 SD = 1.05 .000 Latent
construct Items Reliability Source
Purchase I cannot imagine buying a ticket / I can imagine α = .95 Holzwarth et intention I would not consider this film / I would consider al., 2006
Attitude toward Backseat is a movie that… α = .91 d’Astous &
the film Is poor / Is good Touil, 1999
I would not go out and see / I would go out and see I would not recommend / I would recommend Does not interest me / Interests me
Attitude toward I find Jake Gyllenhaal… α = .91 Yilmaz et al.,
the actor: Unfriendly / Friendly 2011
Likability Insincere / Sincere Unlikable / Likable Cold / Warm
Uncharismatic / Charismatic Unappealing / Appealing
Attitude toward I find Jake Gyllenhaal… α = .98 Taylor et al.,
the actor: Not credible / Credible 2010; Ohanian,
Expertise No expert / An expert 1990
Unexperienced / Experienced Unknowledgeable / Knowledgeable Unqualified / Qualified
Unskilled / Skilled
Attitude toward * α = .86 Feick & Higie,
the actor Jake Gyllenhaal and I probably have similar 1992 Similarity values and beliefs
Attitude toward
Jake Gyllenhaal is quite a bit like me It's likely that Jake Gyllenhaal and I have similar tastes and preferences
With regards to Jake Gyllenhaal, I would say… α = .88 Erfgen et al., the actor I am not familiar with him / I am familiar with him 2016
Familiarity I would not recognize him / I would recognize him I have not heard of him / I have heard of him
Appendix IV: Construction of the variable Movie interest
A principal components factor analysis with three items that measured one’s interest in the movie industry demonstrated the Kaiser-Meyer-Olkin measure was .73, which verified the sample size adequacy for this factor analysis. The Barlett’s Test of Sphericity was significant (p < .001), thereby indicating there was enough correlation between the items. According to the Eigenvalue-criterion (2.39) and Scree plot, one variable explained 79.48% of the variance. All items correlated positively with the first component, and the variable “How informed would you classify yourself” had the strongest association (see Table 4). After Varimax rotation, the factor was labelled Movie interest.
Table 4
Factor Loadings for Movie Interest
Factor 1 With regards to actors, movies, and the film industry in general, how
informed would you classify yourself?
With regards to actors, movies, and the film industry in general, how knowledgeable would you classify yourself?
.88
.86 With regards to actors, movies, and the film industry in general, how
interested would you classify yourself? .76
The reliability of the scale was measured by means of Cronbach’s alpha, which reported the coefficient was .87 and could not be further improved by deleting an item. Thus, Movie interest was proven to be a unidimensional and reliable construct.
scale from 1 (strongly disagree) to 7 (strongly agree). All other items were measured on a bipolar scale, anchored by 1 for the for the first answer option and 7 for the last answer option.
Appendix V: Randomization checks
In order to check whether randomization of the participants was successful, a one-way analysis of variance (ANOVA) was conducted with the Experimental condition as independent variable and Moviegoing frequency, Movie interest, and Age as dependent variables. The ANOVA demonstrated that participants’ moviegoing frequency F (3, 198) = 1.18, p = .320, movie interest F (3, 198) = 2.20, p = .089, nor age F (3, 197) = .22, p = .882 significantly differed over the conditions.
Furthermore, to check if participants’ gender, and nationality were comparable over the four conditions, a Chi-Square was conducted with the Experimental condition as independent variable, and Gender and Nationality as dependent variables. The Chi-Square revealed that participants’ gender (p = .346) and nationality (p = .176) did not vary
significantly per condition (see Table 5).
Table 5
Appendix VI: Manipulation checks Table 6
Descriptive Statistics of the Manipulation Checks Variable Condition 1 (Oscar & actor) Condition 2 (scandal & actor) Condition 3 (Oscar & endorser) Condition 4 (scandal &
endorser) Total sample Movie frequency M = 1.26 SD = 0.45 M = 1.11 SD = 0.38 M = 1.15 SD = 0.51 M = 1.21 SD = 0.46 M = 1.18 SD = 0.45 Movie interest M = 4.92 SD = 1.30 M = 4.30 SD = 1.32 M = 4.60 SD = 1.16 M = 4.53 SD = 1.23 M = 4.59 SD = 1.27 Age M = 24.92 SD = 8.25 M = 25.92 SD = 9.99 M = 24.98 SD = 6.51 M = 26.00 SD = 9.97 M = 25.46 SD = 8.77 Gender Nationality Male: 21 Female: 31 Dutch: 18 Other: 34 Male: 14 Female: 39 Dutch: 22 Other: 31 Male: 12 Female: 36 Dutch: 19 Other: 28 Male: 14 Female: 33 Non-binary: 1 Dutch: 27 Other: 21 Male: 61 Female: 139 Non-binary: 1 Dutch: 86 Other: 114
Note: the variables were measured on a 5-point scale in which the answer options ranged from 1 (very negative) to 5 (very positive), 1 (very uninvolved) to 5 (very involved), 1 (strongly disagree) to 5 (strongly agree), 1 (strongly disagree) to 5 (strongly agree).
Variable Condition 1 (Oscar & actor) Condition 2 (scandal & actor) Condition 3 (Oscar & endorser) Condition 4 (scandal & endorser) Sig. Perceived reputation of Jake
Gyllenhaal in the article
M = 4.30 SD = 0.70 M = 2.28 SD = 1.01 M = 3.94 SD = 0.72 M = 2.29 SD = 0.99 .000 Perceived role of Jake
Gyllenhaal in the film
M = 4.30 SD = 0.80 M = 3.68 SD = 1.11 M = 2.63 SD = 1.41 M = 2.19 SD = 1.00 .000 “Jake Gyllenhaal is an Oscar
winner” M = 4.04 SD = 1.14 M = 2.85 SD = 0.99 M = 4.22 SD = 0.99 M = 2.71 SD = 1.05 .000 “Jake Gyllenhaal has been
involved in a recent scandal”
M = 2.25 SD = 1.09 M = 3.79 SD = 1.25 M = 2.41 SD = 1.15 M = 4.08 SD = 0.96 .000
