This is an open access article distributed under the terms of the Creative Commons
Attribution License (Recent studies suggest that celebrity culture can be perceived as repulsive and disgusting (Shabahang et al., 2023; Shabahang et al., 2021). People who hate celebrities often feel contempt, anger, and fear. As a result, some individuals seek distance from celebrities who may be regarded as menacing. However, others appear motivated to seek out celebrities to express hate-motivated behaviors (e.g., real and online celebrity bashing; see Shabahang et al., 2023). In recent years, celebrities have often been targets of aggression and hate (Ouvrein et al., 2018). For example, celebrity hate websites focus on negative information about celebrities (Soukup, 2006). Ouvrein and colleagues (2017) suggest that negative and scandal-driven criticisms of celebrities are increasing, as evidenced by the large number of aggressive comments in the context of celebrity news (Ouvrein et al., 2018). These negative behavioral reactions are relatively common and extreme (Ouvrein et al., 2019), highlighting the importance of examining this intergroup (celebrity class-community) conflict.
Intergroup conflicts are the “problem of the century” (Fiske, 2002). Such conflicts can be particularly menacing (Condor & Brown, 1988) and bring about substantial societal changes (Cárdenas & de la Sablonnière, 2020). Intergroup conflicts fuel adverse emotions (e.g., fear), cognitions (e.g., stereotyping), and behaviors (e.g., aggression) in individuals involved (Böhm et al., 2020).
Negative in-group and out-group consequences of intergroup conflict have been documented in the in-group and the out-group. For example, Silva and Mace (2015) examined the impact of intergroup conflict on in-group altruism and out-group hostility by investigating trends in charitable donations before, during, and after violent sectarian riots between Catholics and Protestants in Belfast, Northern Ireland. Results revealed that the conflict was linked to a decline of philanthropic cooperation in both the in-group and the out-group. Cooperation increased only after the conflict had been resolved. Indeed, in-group challenges associated with inter-group conflict can occur within the lifetime of individuals or extend across generations (De Dreu & Triki, 2022).
Accordingly, widespread public hatred of celebrities and their culture could lead to intergroup conflict between the community and celebrity class. This may encourage hypervigilant focus on the conflict itself, resulting in troublesome distractions and decline of functionality of both groups, which may fuel the conflict. Furthermore, disputes between the public and the celebrity class are likely to dilute the positive impacts of celebrity support for communities (e.g., promoting health-related knowledge using endorsements, advice, and medical guidance; see Hoffman et al., 2017 for review), and reduce celebrities' desire for engagement in humanitarian and charity issues. A decline in celebrity giving tends to decrease community philanthropic contribution (the celebrity-lift hypothesis; Harris & Ruth, 2015). Therefore, investigating the nature of celebrity culture hate represents a step toward harmonizing relations between community and celebrity groups, which could result in reduced intergroup conflict. The present study aims to provide a comprehensive understanding of celebrity culture hate by dissecting the links between skepticism toward celebrities, perceived celebrity attributes (i.e., perceived deception and dark triad characteristics), and celebrity culture hate.
Diverse media content may be perceived as deceiving. Hancock (2009) explains the notion of the Digital Deception, or perceived deception that emerges in the context of information and communication technology. Distribution of fabricated news has accelerated in contemporary media (e.g., Khan et al., 2021), and a subjective feeling of mistrust toward the mainstream news media has increased (known as media skepticism; see Tsfati & Cappella, 2003). Media skeptics often believe that journalists sacrifice precision in reporting for personal or commercial benefits (Tsfati & Cappella, 2003). Along with journalists, advertisements, and news, celebrities are often perceived as deceptive and manipulative (e.g., Bailey, 2007; Ouvrein et al., 2023).
Interpersonal deception is perceived in both face-to-face and distant communications (e.g., online interactions and media; Sharabi & Caughlin, 2019). Misleading and deceptive endorsements by celebrities and social influencers have been reported (e.g., Balasubramanian et al., 2016; Harris, 2018). Interpersonal Deception Theory (Buller & Burgoon, 1996) details the reciprocal relationship between social interactions and deception. Drawing on this theoretical model, perceived deception in the celebrity class (as deceiver) can breed intergroup conflict, as individuals (the deceived) become motivated to aggress against the deceiver (“celebrity bashing”). The main purpose of this study is to unravel some of the mysteries surrounding the process of the celebrity culture hate by introducing the Perceived Celebrity Deception Inventory (PCDI). To date, studies in the field of perceived media deception have mainly focused on media content such as advertisements (e.g., Chaouachi & Ben Rached, 2012; Darke & Ritchie, 2007; Lim et al., 2020), news (e.g., Vaccari & Chadwick, 2020), and shopping websites (e.g., Riquelme et al., 2016; Román, 2010). However, little attention has been paid to perceived celebrity deception. Furthermore, there is no reliable and validated measure of perceived celebrity deception. The PCDI developed in this study will provide a new tool for investigating the processes underlying intergroup conflict between the community and celebrity classes.
Hypothesis 1: The Perceived Celebrity Deception Inventory (PCDI) is a robust measure of the perceived deception from celebrities.
Drawing on previous evidence concerning age and gender variations in the ability to detect deception (e.g., female deceivers are more apt to be detected; see Sweeney & Ceci, 2014), and hatred toward celebrities (e.g., age is associated with celebrity/celebrity culture hatred; see Shabahang et al. 2021; Shabahang et al., 2023), we examine the age and gender differences on perceived celebrity deception. Moreover, considering that in some cases individuals with high levels of engagement with celebrities can feel very close to celebrities and even see them as close friends or family members (see Brown et al., 2003), another goal of the research is to examine whether perceived deception from celebrities is similar to or different from perceived deception from family and others.
Hypothesis 2: Age and gender are associated with perceived celebrity deception.
Hypothesis 3: Individuals perceive varying levels of deception from family, others, and celebrities.
There is evidence that cognitive skills such as critical thinking and skepticism may predict people’s attitudes toward celebrities. Inordinate attachment to celebrities is associated with low cognitive skills (e.g., arithmetic, creativity, crystallized intelligence, critical thinking, spatial ability, and an attitudinal need to think; see McCutcheon et al., 2003). Previous research on perception of celebrity credibility and belief in a just world show that people who are critical of celebrity status often question whether celebrities deserve their places in society (Shabahang et al., 2021). Skeptical attitudes toward celebrities and celebrity culture hate may be protective for consumers. Employing critical thinking about celebrities may help people avoid blind trust and careless consumption (Kleinig, 2018). People who are skeptical toward celebrities will question whether celebrities deserve their status and privileges by substantially contributing to society, whether celebrity wellbeing results in negative consequences for communities, and whether celebrities help create or help solve community problems. Considering the impact of skepticism on negatively responding to suspicious targets (e.g., advertisement; Obermiller et al., 2005), it is logical to assume that skepticism will be associated with celebrity culture hate.
Hypothesis 4: Skeptical attitudes toward celebrities is associated with celebrity culture hate.
The perceived negative personality traits of celebrities may mediate the possible link between public’s skepticism toward celebrities and celebrity culture hate. Studies have investigated the impact of perceived deception on negative attitudes toward the deceiver and support for limiting the deceiver’s influence. For instance, Chaouachi and Ben Rached (2012) reported that perceived deception in advertisement engenders negative attitudes toward the brand and a decline in purchase intention among consumers. Lim et al. (2020) showed that perceived deception in online advertising of weight loss products appears to predispose consumers to support government supervision and regulative activities of such products. Darke and Ritchie (2007) argue that deceptive advertisements cultivate a cynical outlook leading to loathing of subsequent promotion from both the same source and similar sources. Moreover, deceptive advertisements make individuals believe that they have been fooled, which evokes self-protective goals to guard against later deceptions. Building on previous arguments explaining the link between perceived deception and negative attitudes toward deceptive targets, this investigation aims to shine new light on the association between skepticism toward celebrities and celebrity culture hate by probing the impact of perceived celebrity deception as an aggravating factor.
Hypothesis 5: Perceived celebrity deception mediates the association between skeptical attitudes toward celebrities and celebrity culture hate.
The attribution of dark triad characteristics (narcissism, psychopathy, and Machiavellianism) has been associated with dislike of popular media figures (Snyder et al., 2019). Young and Pinsky (2006) indicate that celebrities are significantly more narcissistic than the general population. Celebrity scandals, such as drug (Tiger, 2013) and philanthropy scandals (Jeffreys, 2011), also indicate signs of Machiavellian and psychopathic behaviors in celebrities. People who exhibit dark triad traits are considered to be deceptive, socially aversive, and have the potential for malevolent behaviors. It makes sense that one who perceives dark triad traits among celebrities might be skeptical of celebrities because they are viewed as potentially dangerous. Therefore, we predict that skepticism toward celebrities may affect celebrity culture hate through the mediator of perceived dark triad of celebrities.
Hypothesis 6: Perceived dark triad of celebrities mediates the association between skeptical attitude toward celebrities and celebrity culture hate.
The Seven-Stage Hate Model suggests that hate progresses from hateful thoughts towards the target to behaviors that harm the target (e.g., see; Schafer & Navarro, 2003). Celebrity culture hate has previously been shown to predict celebrity bashing (Shabahang et al., 2023). Skepticism toward celebrities might result in relatively low levels of hate, for example a hatred characterized by repulsion and disgust, without a behavioral component. However, when skeptical attitudes toward celebrities are combined with perceived celebrity deception, narcissism, Machiavellianism, and psychopathology, the hatred of celebrity culture may go beyond devaluing celebrities to include behaviors such as celebrity bashing. Accordingly, we evaluate a model predicting how celebrity deception and dark triad characteristics mediate the association between skeptical attitude toward celebrities and celebrity culture hate. Furthermore, considering diverse attitudes of individuals toward celebrities with different fields of expertise (e.g., celebrities in the fields of music, acting, and video-making are viewed as favorable to other fields; Zsila et al., 2018), participants' perspectives concerning the fields containing the most deceptive celebrities was investigated.
Hypothesis 7: Celebrities with various fields of expertise will elicit different levels of perceived deception and skepticism.
Method
Participants
A convenience sample of Iranian internet users (N = 1175; 904 females and 271 males; age 14–63; Mage = 29.20, SDage = 9.76) was recruited from two Iranian online shopping websites by posting an advertisement inviting cooperation in a celebrity-related study. The study was approved by the Institutional Review Board (IRB) at the University of Guilan. Ethical practices were observed in accordance with the World Medical Association Declaration of Helsinki. All respondents completed written informed consent before participating in the study.
Measures
The Perceived Celebrity Deception Inventory (PCDI) was developed for the present study to capture the perception that celebrities use manipulative and deceptive practices with the intent of misleading and/or deluding people into errantly believing a lie or inaccuracy. The seven items were based on published perceived deception measures (Chaouachi & Ben Rached, 2012; Riquelme et al., 2016; Román, 2010): 1) Celebrities tell us what they want, which may not reflect reality; 2) Reality is different from what celebrities say; 3) Celebrities mislead us by manipulating the truth; 4) Celebrities use clever manipulation about the truth to get their way; 5) Celebrities attempt to deceive society; 6) It isn’t a surprise that celebrities often lie; and 7) If you want the truth about a topic, you shouldn’t listen to a celebrity. The five-point response scale ranged from 1 (strongly agree) to 5 (strongly disagree).
In order to distinguish the difference between perception of celebrity deceptiveness from general deceptiveness other people, the items of the PCDI were reworded to examine perceived falsehood from family (7 items; e.g., “My family attempts to deceive me”) and others (7 items; e.g., “People around me attempt to deceive me”). The items had the same score range (7-35) and response options as the PCDI.
Three items were adopted from the Professional Skepticism Scale (Hurtt, 2010) to measure skeptical attitude toward celebrities (i.e., “I don't accept celebrity statements and explanations without further thoughts,” “I like to understand the reason for celebrity behaviors,” and “I don't tend to immediately accept what celebrities say”). Response options were on a 6-point scale (1 = strongly disagree, 6 = strongly agree).
Hatred toward celebrity culture was measured using the 17-item Celebrity Culture Triangular Hate Scale (Shabahang et al., 2023; e.g., “Celebrity culture is repugnant to me”) capturing three components of hatred of celebrity culture (i.e., negation of intimacy, passion, & commitment). The items were rated from 1 (not at all) to 9 (extremely).
Perceived dark triad personality characteristics of celebrities were assessed using three adapted items from the Short Dark Triad Scale (SD3; Jones & Paulhus, 2014). The original SD3 consists of 27 items divided into three domains (Machiavellianism, narcissism, and psychopathy). Items are presented in Likert-type format with anchors ranging from 1 (strongly disagree) to 5 (strongly agree). For the purpose of this study, three items from the SD3 were adapted to assess perception that celebrities are Machiavellian, narcissistic, and psychopathic, respectively: 1) “Celebrities talk and behave in a specific way to have important people on their side”; 2) “Celebrities insist on getting the respect they deserve”; and 3) “Celebrities are out of control.” The three adapted items were selected from the items that best described each aspect of dark triad in the SD3 (based on explained variances). When constructs are well-defined and narrow in scope, single-item measures are viable (Fuchs & Diamantopoulos, 2009). Extant evidence has confirmed the usefulness of single-item measures (e.g., Bergkvist & Rossiter, 2007; Gardner et al., 1998; Wanous et al., 1997).
The participants were also asked to respond to following questions: 1) “Which group of celebrities do you perceive as the most deceptive?” and 2) “Which group of celebrities are you most skeptical toward?” Response options were Actors, Musicians, Artists, Social media influencers, TV/Radio presenters, Athletics, Authors, Scientists, and Others.
Procedure
The items of pilot PCDI were generated by reviewing previous theoretical and empirical literature on perceived deception (see Held & Germelmann, 2018 for a review) and hatred toward celebrities (Shabahang et al., 2021) and celebrity culture (Shabahang et al., 2023). Three researchers with extensive knowledge of celebrity psychology and experiences with validation of media psychology measures evaluated the items for their conceptual scope, clarity, simplicity, and relevance. After considering experts’ suggestions and an online pilot study (n = 30), minor revisions to the wording of the inventory were completed, resulting in the final version of the PCDI. The final online survey including the PCDI and additional measures was assembled and interested individuals were able to access the survey by clicking on the online ad which provided a link to the survey on Google® Forms. Factor (i.e., exploratory factor analysis—EFA & confirmatory factor analysis—CFA) and item (i.e., corrected item-total correlation, & Cronbach's alpha) analyses were conducted to examine the psychometric characteristics of the Perceived Celebrity Deception Inventory. Furthermore, Pearson correlation analysis, regression analysis, Wilcoxon signed-rank test, and structural equation modeling were employed for addressing the remaining research hypotheses regarding the differences in perceived deception from celebrities, family, and others, correlations between research variables, and the mediating role of perceived negative personality traits of celebrities. Data analysis was performed using SPSS statistical software (IBM SPSS Statistics 25.0; IBM Corp., 2017) and lavaan package (Rosseel, 2012) in R software (R Core Team, 2020).
Results
Psychometric Validation of the Perceived Celebrity Deception Inventory (PCDI; H1)
We first evaluated whether any item pairs were excessively correlated. The initial unidimensional EFA revealed that items 1 and 6 were suspected to be redundant with others. After eliminating the problematic two items, the Cronbach’s α with the remaining 5 items remained high (.87, 95% CI [.86, .88]). Next, we conducted EFA with the randomly split data (n = 587) to confirm the factor structure and implemented CFA with the remaining data (n = 588) to cross-validate the suggested factor structure. The results of Horn’s (1965) parallel analysis (see Figure 1) suggested up to one latent factor, and which was supported by EFA (see Table 1). The suggested one-factor model explained 55.86% of the total variance.
Figure 1
Scree Plot From the Parallel Analysis of the Perceived Celebrity Deception Inventory (N = 587)
Table 1
Exploratory Factor Analysis of Items in the Perceived Celebrity Deception Inventory
| Item | Perceived Celebrity Deception Inventory | h2 | u2 | M | SD |
|---|---|---|---|---|---|
| 2 | .69 | .48 | .52 | 3.72 | 1.00 |
| 3 | .80 | .64 | .36 | 3.44 | 1.06 |
| 4 | .78 | .61 | .39 | 3.86 | 1.05 |
| 5 | .80 | .64 | .36 | 3.18 | 1.11 |
| 7 | .65 | .42 | .58 | 3.50 | 1.16 |
Note. N = 587. h2 = communality; u2 = specific variance.
The results of the CFA (see Table 2) satisfied the recommended values for good fit (CFI = .994, TLI = .988, RMSEA = .056). In addition, we evaluated item characteristics using the full sample (N = 1175). The results (see Table 3) indicated that the internal consistency of the scale was acceptable: Ranges of corrected item-total correlations: [.62, .75], and Cronbach’s α if item deleted: [.83, .86].
Table 2
Goodness-of-Fit Indices for the One-Factor Model of the Perceived Celebrity Deception Inventory
| χ2 | df | p | CFI | TLI | RMSEA |
|---|---|---|---|---|---|
| 14.076 | 5 | < .05 | .994 | .988 | .056 |
Note. N = 588. The comparative fit index (CFI; cutoff ≥ .90; Bentler, 1990), the Tucker Lewis index (TLI; cutoff ≥ .90; Bentler & Bonnet, 1980), and the root mean square error of approximation (RMSEA; cutoff ≤ .08; Browne & Cudeck, 1993).
Table 3
Item Characteristics of the Perceived Celebrity Deception Inventory
| Item | M | VAR | Scale mean if item deleted | Scale variance if item deleted | Corrected item-total correlations | Cronbach’s α if item deleted |
|---|---|---|---|---|---|---|
| 2 | 3.71 | 0.98 | 13.81 | 13.92 | .64 | .85 |
| 3 | 3.44 | 1.16 | 14.08 | 12.77 | .74 | .83 |
| 4 | 3.78 | 1.13 | 13.74 | 13.01 | .72 | .83 |
| 5 | 3.13 | 1.27 | 14.39 | 12.46 | .75 | .83 |
| 7 | 3.46 | 1.46 | 14.06 | 12.84 | .62 | .86 |
Note. N = 1175. Cronbach’s α = .87; 95% CI [.86, .88].
Gender and Age Variations in Perceived Celebrity Deception (H2) and Differences in Perceived Deception From Celebrities, Family, and Others (H3)
We initially conducted a multiple regression analysis to investigate whether perceived celebrity deception varies by age and gender. We created a composite score by averaging the perceived celebrity deception items. The minimum age was subtracted from each age to interpret the intercept with regard to a minimum age in our data (i.e., 14). The regression results (see Table 4) showed that perceived celebrity deception was not significantly related to gender nor age.
Table 4
The Effect of Age and Gender on Perceived Celebrity Deception
| Variable | B | SE | β |
|---|---|---|---|
| Intercept | 3.75*** | 0.08 | 0 |
| Female | -0.06 | 0.07 | -0.03 |
| Age | 0.00 | 0.00 | 0.01 |
Note. N = 1169 because 6 observations were deleted due to missing in the age variable. B = unstandardized coefficient; β = standardized coefficient.
*p < .05. **p < .01. ***p < .001.
We also compared differences between perceived deception of celebrities and those of family and others. The means (SDs) of perceived deception of celebrities, family, and others were 3.71 (0.99), 1.99 (1.05), and 3.19 (0.98), respectively. The results of Levene’s test (Levene, 1960) and the Shapiro-Wilk test (Shapiro & Wilk, 1965) supported neither the assumptions of homogeneity of variance (p < .05) nor normality (p < .05). Thus, we performed Wilcoxon signed-rank tests (Wilcoxon, 1945). In addition, we implemented the Bonferroni correction (Bonferroni, 1936), which offers adjusted p-values to avoid the elevated risk of Type I error. The results (see Table 5) indicated the three types of perceived deception were significantly different from each other (ps < .001). The largest difference occurred at the celebrity – family ( = 1.72), which were followed by others – family ( = 1.20), and celebrity – others ( = 0.52). When we consider that the range of the three types of perceived deception was [1, 5], it is notable that the mean perceived celebrity deception and mean perceived deception of other people were relatively high, while that of the family was relatively low.
Table 5
Mean Differences Between Perceived Deception of Celebrities, Family, and Others
| Comparison
|
Mean Difference | Adjusted p-value | |||
|---|---|---|---|---|---|
| Group1 | Group2 | ||||
| Celebrity | Family | 3.71 (0.99) | 1.99 (1.05) | 1.72 | < .001 |
| Celebrity | Others | 3.71 (0.99) | 3.19 (0.98) | 0.52 | < .001 |
| Others | Family | 3.19 (0.98) | 1.99 (1.05) | 1.20 | <.001 |
Note. N = 1175.
In addition, we investigated to what extent people perceived celebrities as Machiavellian, narcissistic, and psychopathic. Considering the range of each variable [1, 6], there was general consensus that celebrities are Machiavellian (M = 3.93, SD = 1.05), narcissistic (M = 4.11, SD = 1.03), and psychopathic (M = 3.30, SD = 1.22). Since both the Levene’s test and the Shapiro-Wilk test did not satisfy the assumptions (ps < .05), we conducted Wilcoxon signed-rank tests with the Bonferroni correction and compared the differences between perceived dark triad personality characteristics of celebrities. According to the results (see Table 6), there were significant differences between how people perceive the dark personality of celebrities (ps < .001). The largest difference occurred between narcissistic and psychopathic ( = 0.81), which was followed by Machiavellian and psychopathic ( = 0.63), and narcissistic and Machiavellian ( = 0.18). Overall, celebrities are perceived as Machiavellian, narcissistic, and psychopathic at the same time by respondents of this study, but at different levels.
Table 6
Mean Differences Between Perceived Dark Triad Personalities of Celebrities
| Comparisons
|
Mean Difference | Adjusted p-value | |||
|---|---|---|---|---|---|
| Group1 | Group2 | ||||
| Narcissists | Machiavellian | 4.11 (1.03) | 3.93 (1.05) | 0.18 | < .001 |
| Machiavellian | Psychopath | 3.93 (1.05) | 3.30 (1.22) | 0.63 | < .001 |
| Narcissists | Psychopath | 4.11 (1.03) | 3.30 (1.22) | 0.81 | <.001 |
Note. N = 1175.
Correlates of Perceived Celebrity Deception (H4)
We investigated correlations between perceived celebrity deception and other measures of interest: skepticism toward celebrities, perceived dark triad personality of celebrities (Machiavellianism, narcissism, psychopathy, and total), and celebrity culture hate (negation of intimacy, passion, commitment, and total). The correlation matrix (see Table 7) indicated that perceived celebrity deception was positively correlated with skepticism toward celebrities (r = .10, p < .001), perceived celebrity Machiavellianism (r = .34, p < .001), perceived celebrity narcissism (r = .29, p < .001), perceived celebrity psychopathy (r = .31, p < .001), perceived dark personality of celebrity (r = .39, p < .001), negation of intimacy (r = .32, p < .001), passion (r = .20, p < .001), commitment (r = .24, p < .001), and celebrity culture hate (r = .28, p < .001).
Table 7
Descriptive Statistics, and Correlations of Measures
| Variable | Descriptive statistics
|
Correlations
|
|||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| M | SD | α | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | |
| 1. PCD | 3.71 | 0.99 | .87 | 1 | .10*** | .34*** | .29*** | .31*** | .39*** | .32*** | .20*** | .24*** | .28*** |
| 2. Skepticism toward celebrities | 4.90 | 0.93 | .51 | 1 | .20*** | .18*** | .07* | .18*** | .08** | .02 | .07* | .07* | |
| 3. Perceived celebrity Machiavellianism | 3.93 | 1.05 | — | 1 | .59*** | .42*** | .81*** | .33*** | .23*** | .35*** | .35*** | ||
| 4. Perceived celebrity narcissism | 4.11 | 1.03 | — | 1 | .46*** | .82*** | .31*** | .21*** | .33*** | .32*** | |||
| 5. Perceived celebrity psychopathy | 3.30 | 1.22 | — | 1 | .80*** | .43*** | .40*** | .43*** | .48*** | ||||
| 6. Perceived dark personality of celebrity | 3.78 | 0.89 | .73 | 1 | .44*** | .35*** | .46*** | .48*** | |||||
| 7. Negation of intimacy | 4.52 | 1.92 | .88 | 1 | .65*** | .64*** | .84*** | ||||||
| 8. Passion | 3.33 | 2.06 | .92 | 1 | .70*** | .88*** | |||||||
| 9. Commitment | 4.35 | 2.17 | .94 | 1 | .92*** | ||||||||
| 10. Celebrity culture hate | 4.10 | 1.82 | .95 | 1 | |||||||||
Note. N = 1175.
*p < .05. **p < .01. ***p < .001.
Mediating Role of Perceived Celebrity Deception and Perceived Dark Triad of Celebrities (H5 and H6)
Finally, we investigated the mediating role of perceived celebrity deception and perceived dark triad of celebrities in the relationship between skepticism toward celebrities and celebrity culture hate. As discussed earlier, perceived negative personality traits of celebrities (i.e., perceived celebrity deception and perceived dark triad of celebrities) seem to aggravate the skepticism toward celebrities that results in elevated celebrity culture hate. This assumption was evaluated by conducting structural equation modeling.
Descriptive statistics of each variable are in Table 7. As the first step, we checked the normality assumption of each indicator using the univariate and multivariate tests. The results did not support a normal distribution (ps < .05). To address this issue, we employed the maximum likelihood method (MLM), which provides robust standard errors and helps to resolve the multivariate normality violation issue. Gender and age variables significantly explained only celebrity culture hate, and thus they were considered as covariates for celebrity culture hate. Results of structural equation modeling (Figure 2) illustrated that skepticism toward celebrities influenced celebrity culture hate through the mediators of perceived celebrity deception and perceived dark triad of celebrities.
Figure 2
Structural Equation Modeling
Note. Standardized coefficients (standard errors) are reported. The dashed line indicates a non-significant coefficient. Six observations were deleted due to missing age data (N = 1169).
Mediation analysis results (see Table 8) supported significant mediation effects of perceived celebrity deception (B = 0.01, p < .05) and perceived dark triad of celebrities (B = 0.19, p < .05).
Table 8
Mediation Analyses and Indirect Effect With 95% Confidence Interval
| Mediator | Path a B (SE) |
Path b B (SE) |
Indirect Effect
|
Direct Effect B (SE) |
Overall Effect B (SE) |
|
|---|---|---|---|---|---|---|
| B | 95% CI | |||||
| Perceived celebrity deception | 0.102*** (0.031) |
0.107* (0.050) |
0.011* | [0.001, 0.025] | -0.089 (0.048) | 0.116* (0.051) |
| Perceived dark triad of celebrities | 0.184*** (0.028) |
1.056*** (0.086) |
0.194* | [0.131, 0.262] | ||
Note. Path a: Skeptical toward celebrities → Mediator; Path b: Mediator→ Celebrity culture hate. Six observations were deleted due to missing the age variable (N = 1169). B = unstandardized coefficient.
*p < .05. **p < .01. ***p < .001.
Perceived Celebrity Deception and Skepticism Toward Celebrities by Field of Celebrity (H7)
In addition, we compared perceived celebrity deception and skepticism toward celebrities by field of celebrity (see Figure 3). Participants perceived the most deception from and were most skeptical toward 'social media influencers' (deception: 38.81%, skepticism: 36.34%) followed by ‘actors’ (deception: 29.28%, skepticism: 29.19%), and ‘TV/Radio presenters’ (deception: 16.25%, skepticism: 16.34%).
Figure 3
Bar Plots Presenting the Expertise of Celebrities Perceived as Most Deceptive (Left) and the Expertise of Celebrities Who Elicit the Greatest Skepticism (Right)
Discussion
The present study examined whether perceived celebrity deception and dark triad characteristics of celebrities mediated the relationship between skepticism and celebrity culture hate. After establishing the psychometric properties of the PCDI, we evaluated a model predicting how celebrity deception and dark triad characteristics mediated the link between skepticism toward celebrities and celebrity culture hate.
The PCDI showed good reliability and validity (H1). After eliminating redundant items, the resulting 5-item inventory demonstrated high internal consistency and validity, explaining most of the variability in scores using exploratory and confirmatory factor analyses. The PCDI will be useful in future investigations gauging the extent to which people perceive that celebrities are motivated to deceive the public. Future research could use the scale to examine whether perceived celebrity deception predicts consumer behavior and active aggression toward celebrity culture (celebrity bashing).
The results did not confirm the hypothesis that participant age and gender predict perceived celebrity deception (H2). Similarly, Chaouachi and Ben Rached (2012) did not find a significant impact of consumer gender and age on perceived deception in advertising.
Individuals perceived deception from celebrities as similar to perceived deception from others (H3). Perceived deception of celebrity and non-family (others) were relatively high while perceived deception from family was relatively low. Furthermore, celebrities were perceived as Machiavellian, narcissistic, and psychopathic at the same time, but at different levels. Above all, celebrities were more perceived as narcissistic by the participants.
In accordance with the Seven-Stage Hate Model (e.g., see; Schafer & Navarro, 2003), we hypothesized that skepticism toward celebrities would result in mild levels of celebrity hatred. However, we predicted that celebrity culture hatred would be magnified when skepticism is combined with perceived celebrity deception and dark triad characteristics (narcissism, psychopathy, and Machiavellianism). The result of the correlation analysis confirmed the weak direct association of the skepticism and celebrity culture hate (H4). However, this link was insignificant in the structural analysis. Furthermore, perceived celebrity deception (H5) and dark triad characteristics (H6) were found to mediate the relationship between celebrity skepticism and celebrity culture hate. These results suggest that disdain for celebrities is related to the extent to which celebrities are seen as deceptive, selfish, and manipulative.
Skepticism of celebrities could be merely correlated with mild levels of celebrity hatred. However, this link seems to be indirect rather than direct. At this mild level of hatred, people may divert attention away from celebrity news and advertising. These skeptical attitudes toward celebrities and celebrity culture may be protective for consumers. Employing critical thinking about celebrities may help people avoid blind trust and careless consumption (Kleinig, 2018). Skepticism toward celebrities may result in negative evaluations of celebrity endorsements which may reduce intention to purchase products. In this sense, celebrity skepticism may be protective for consumers in avoiding unnecessary expenditures. That said, advertisers would be wise to consider the variable of celebrity skepticism when employing celebrities to increase positive affect toward brands and purchase intention.
Our results support previous research showing that people who attribute dark triad characteristics to celebrities tend to dislike celebrities more than others (Snyder et al., 2019). The characterization of celebrities as deceptive, selfish, and manipulative is not surprising insofar as many people are aware of celebrity scandals (Jeffreys, 2011; Tiger, 2013) and misleading celebrity endorsements (e.g., Balasubramanian et al., 2016; Harris, 2018). Celebrities are often viewed as narcissistic (Young & Pinsky, 2006) and undeserving of their stature in society (Shabahang et al., 2021).
The Seven-Stage Hate Model (Schafer & Navarro, 2003) proposes that low levels of hate are confined to hateful thoughts, while higher levels of hate are more likely to be manifested in hateful behaviors. While skepticism toward celebrities might result in relatively low levels of hate, when skeptical attitudes toward celebrities are combined with perceived celebrity deception, narcissism, Machiavellianism, and psychopathology, celebrity hatred may progress beyond diminutive thoughts about celebrities. Self-reported celebrity hatred predicts celebrity bashing (Shabahang et al., 2023) which is often manifested through posting aggressive comments online, many of which can result in considerable distress (Ouvrein et al., 2019).
In this study, social media influencers elicited the most skepticism and were judged to be the most deceptive (H7). A possible explanation for this result may be the marked contribution of influencers in false and conspiracy theory promotion (e.g., misinformation regarding COVID-19; see Harff et al., 2022). Furthermore, compared to traditional celebrities who are highly regulated, micro-celebrities (e.g., influencers) are more inclined to reveal intimate details of their thoughts and behaviors, representing themselves as more honest and less manipulative (Marwick, 2015). This frank and unregulated expression may increase suspiciousness toward influencers. Moreover, contrary to other types of celebrities, popularity of influencers is dependent upon generating eye-catching content and gaining paralinguistic digital affordances (e.g., “likes”; see Ruiz-Gomez, 2019), which may appear to elevate influencers’ probability of using manipulative fame-seeking behaviors. Aside from the public skepticism of social media influencers, investigations have documented societal contributions of this celebrity class in areas of public health communication (Pöyry et al., 2022) and promotion of health campaigns (Kostygina et al., 2020). Further research should investigate whether the influencer-public conflict, rooted in skeptical views toward influencers, has the potential to debilitate influencers’ constructive impacts on communities.
A number of noteworthy limitations need to be considered. Data were collected on a sample of internet users in Iran. Recently, Iranians have expressed higher levels of hatred toward native celebrities and their lifestyles in comparison to Americans (Shabahang et al., 2023). Considering the generally negative perspective of Iranians toward celebrities, we are unable to assess the degree to which our findings generalize to other populations and cultures. Most of the participants in the current investigation were women (904 women versus 271 men). Though age and gender were not associated with any of the constructs of the present study except celebrity culture hate (this effect was covariated in the SEM), further studies with an equal number of participants from each gender would be worthwhile. In current investigation, most of the participants were young adults (Mage = 29.20) with a small number of adolescents and older adults. Further research should compare a population that is more diverse in age to provide generation-related estimates. The design of this study was correlational. Therefore, we can only speculate about cause-and-effect relationships between variables. A three-item scale was used to capture perceived celebrity dark triad traits. Though single-item measures have been psychometrically supported in the literature (e.g., Bergkvist & Rossiter, 2007; Gardner et al., 1998; Wanous et al., 1997), further study with a multiple-item scale of perceived dark triad of celebrities is recommended. Finally, we did not measure behavioral aspects of hate such as celebrity bashing. However, celebrity culture hate has previously been shown to predict celebrity bashing (Shabahang et al., 2023). Nonetheless, future research should include measures of celebrity bashing to examine whether the perception of deception and dark triad attributes in celebrities predict celebrity bashing.