Effects of Exposure to an Entertainment-Based Genetic Testing Narrative on Genetic Testing Knowledge, Attitudes and Counseling Discussion Intentions

ABSTRACT

This study explores the effects of exposure to a reality television narrative depicting genetic testing on attitudes and intentions, looking particularly at the effects of narratives containing elements of misinformation on genetics-related knowledge accuracy. In an experiment, participants completed a baseline survey, viewed a high versus low-accuracy narrative, then completed a follow-up survey. Exposure to a low-accuracy narrative was associated with lower knowledge accuracy. Indirect effects of identification and transportation on intentions to talk to a doctor about genetic testing also were detected via attitudes and reduced message counterarguing. Results illustrate the negative implications of inaccurate narratives on knowledge, which is concerning given the public’s low level of genetic literacy, as well as the critical role narrative engagement may play in shaping public attitudes and intentions regarding genetic testing.

It is well-established that germline pathogenic variants in BRCA1 (OMIM 113705) and BRCA2 (OMIM 600185) significantly increase cancer risk and warrant changes to cancer risk management (Daly et al., 2021). Given the number of BRCA mutation-positive individuals not yet identified in the population, hereditary breast and ovarian cancer (HBOC) has been deemed a Tier 1 genomic application, which means that the public at-large may benefit from the identification of at-risk individuals (Centers for Disease Control and Prevention, 2014). Mass communication regarding genetic risks may help increase awareness and shape public attitudes and intentions regarding genetic testing, as many people get information about genetics from the media (Geller, Bernhardt, & Holtzman, 2002; Nisker & Daar, 2006).

Genetic testing storylines have become more common in entertainment television (Hether, Huang, Beck, Murphy, & Valente, 2008) and an increasing number of celebrities have shared their experiences with genetic testing (Field, 2015, November 5). These entertainment narratives could be influential on public knowledge and attitudes regarding genetic testing, particularly since genetic literacy is generally low (Christensen, Jayaratne, Roberts, Kardia, & Petty, 2010). For example, exposure to the television drama 90210 featuring a BRCA mutation testing storyline was associated with increased viewer knowledge and greater intentions to talk to a doctor about BRCA testing (Rosenthal, Buffington, & Cole, 2018).

Narrative messages could make scientific topics like genetics more accessible to the general public (Henderson & Kitzinger, 1999) and the use of a narrative format could increase the likelihood that viewers will learn from and embed the information contained within (Slater & Rouner, 2002). However, knowledge acquisition resulting from exposure may depend on how well genetic information is integrated into the narrative (Quintero Johnson, Harrison, & Quick, 2013) and the accuracy of the information contained within. While content reporting on genetics is shown to be generally accurate, an overemphasis on the benefits and an under-reporting of the risks and limitations of genetic testing and personalized medicine is common (Bubela & Caulfield, 2004; Hicks-Courant et al., 2021).

Although public interest in genetic testing is high (Vermeulen, Henneman, van El, & Cornel, 2014), research on narrative engagement and the effects of exposure to media representations of genetics has lagged (Haran & Kitzinger, 2009). To better understand the effects of narrative misinformation (i.e., false or inaccurate information woven into a narrative; Southwell et al., 2019) on viewers, we assess the effect of exposure to a low-accuracy (versus high-accuracy) reality television narrative on knowledge, attitudes and intentions to talk to a healthcare provider about obtaining genetic counseling. The narrative is a storyline from the reality show “Keeping Up with the Kardashians” that illustrates aspects of the genetic counseling and testing process. Drawing on propositions from narrative theory, we also assess potential mechanisms of these effects including narrative engagement (i.e., transportation and identification) and resistance (Figure 1).

Figure 1. Study conceptual model and hypotheses of interest notes. Directionality of hypotheses is indicated in brackets.

Narrative Communication

A narrative is a “cohesive and coherent story with an identifiable beginning, middle, and end that provides information about scene, characters, and conflict; raises unanswered questions or unresolved conflict; and provides resolution’’ (Hinyard & Kreuter, 2007, p. 778). Narratives foster involvement or engagement in a storyline or with a character (Moyer-Gusé, 2008, p. 409) and are shown to positively influence attitudes, intentions and behaviors (Braddock & Dillard, 2016). Specific to health contexts, exposure to narrative interventions is associated with behavior change (Perrier & Martin Ginis, 2018) and is often (but not always) shown to have beneficial effects on patient-decision-making (Shaffer et al., 2021). Furthermore, with some exceptions (Hinyard & Kreuter, 2007), narratives are shown to be more effective than non-narratives for educating viewers about health topics (Borrayo, Rosales, & Gonzalez, 2017; Murphy, Frank, Chatterjee, & Baezconde-Garbanati, 2013). In the context of health decision-making, studies show narrative exposure often enhances knowledge (Shaffer et al., 2021). Therefore, given the potential influence of health narratives on viewers, an investigation of how elements of misinformation embedded within a narrative could negatively influence public understanding of and attitudes toward genetics is of importance.

Effects of Misinformation

According to the capacity model (Fisch, 2000), when there is little distance between educational and narrative content, those exposed to a narrative are likely to retain story-consistent information. Existing or working knowledge on a topic is also likely to impact how new information (accurate or inaccurate) on a topic is processed (Biek, Wood, & Chaiken, 1996). In the case of genetics, where public knowledge is often low (Gupta et al., 2021), narrative misinformation could be problematic as it may lead viewers to embed inaccurate information, making it more easily accessible (Slater & Rouner, 2002) and less resistant to correction (Ecker, Lewandowsky, & Tang, 2010; Johnson & Seifert, 1994; Southwell & Thorson, 2015). Therefore, we propose that participants who view a low-accuracy narrative will exhibit lower knowledge of genetics post-exposure than those viewing a high-accuracy narrative (H1).

Narrative Engagement Mechanisms

According to the Entertainment Overcoming Resistance Model (EORM; Moyer-Gusé, 2008) and Extended-Elaboration Likelihood Model (E-ELM: Slater & Rouner, 2002), narratives can influence attitudes and behavioral intentions because they allow individuals to become absorbed into a story world via transportation and identification. Transportation allows individuals to experience a “vicarious cognitive and emotional response” to an unfolding story (Moyer-Gusé, 2008, p. 409) that may result in greater enjoyment (Green, Brock, & Kaufman, 2004), story-consistent beliefs (Green, 2004; Green & Brock, 2000), cognitive response (Banerjee & Greene, 2012) and persuasiveness (Braverman, 2008). Identification is the process of taking on or imagining oneself in the role of a character (Cohen, 2001). Although it can be hard to distinguish transportation from identification (De Graaf, Hoeken, Sanders, & Beentjes, 2012), transportation refers to storyline involvement whereas identification refers to character involvement (Moyer-Gusé, 2008). Identification acts as a mechanism of narrative persuasion that may help facilitate story-consistent attitudes (De Graaf et al., 2012) and health behaviors (Moyer-Gusé & Nabi, 2010). Considering that the mechanisms of narrative engagement are reliant on viewer enjoyment of the story and the quality of the entertainment narrative, rather than the accuracy of information embedded in the story, we do not expect transportation (H2a) or identification (H2b) to differ between those viewing a low versus high-accuracy narrative.

Despite their benefits to persuasion, transportation and identification may be attenuated by viewer counterarguing. Counterarguing involves the generation of rebuttals or counter examples that can lead to disengagement from the message (Slater & Rouner, 2002). Greater transportation should reduce motivation to counterargue (Green & Brock, 2000). Individuals who identify with characters may also be less capable of generating counterarguments (Igartua, 2010; Slater & Rouner, 2002). Indeed, transportation and identification often have a negative relationship with counterarguing (e.g., Moyer-Gusé, Jain, & Chung, 2012; Moyer-Gusé & Nabi, 2010). Reduced counterarguing, in turn, is associated with story-consistent attitudes and intentions (Green & Clark, 2013; Moyer-Gusé, 2008). In the case of the present study, narrative engagement, which reduces the means to counter argue, could increase the acceptance of informational inaccuracies embedded within the narrative and facilitate more positive attitudes regarding the message, as media may focus more on the benefits than the challenges of personalized medicine (Hicks-Courant et al., 2021). Thus, we propose that greater character identification (H3) and transportation (H4) will be associated with less counterarguing, which will be associated with lower knowledge (H5) and more positive attitudes toward genetic testing (H6) following narrative exposure.

We further hypothesize that knowledge and attitudes will be directly associated. Theory suggests individual characteristic and difference variables may function as distal predictors of behavior via cognitive mediators, including attitudes (Yzer, 2012). Knowledge alone is unlikely to change a behavior (Fisher & Fisher, 1992). In the context of genetics, greater knowledge is correlated with more positive attitudes toward genetics (Etchegary et al., 2010); thus, we expect those who are more knowledgeable to hold more positive attitudes toward genetic testing (H7). However, the relationship between genetics knowledge and attitudes toward genetic testing is likely complex, as increased genetics knowledge and awareness could also be associated with more critical (and less positive) attitudes toward genetic technologies (Etchegary, 2014).

Effects on Behavioral Intentions

The EORM proposes that narrative engagement and decreased message resistance will facilitate story-consistent attitudes and behavior (Moyer-Gusé, 2008). Furthermore, the theory of planned behavior suggests that those with more positive attitudes will exhibit greater behavioral intentions (Ajzen, 1991). Meta-analyses support the attitude and intention relationship across health contexts (Armitage & Conner, 2001; Hagger, Chan, Protogerou, & Chatzisarantis, 2016), as well as in the context of genetics specifically (Shaw & Bassi, 2001; Sussner, Jandorf, Thompson, & Valdimarsdottir, 2010). Furthermore, across studies, attitudes are a consistent predictor of genetic testing decisions (Sweeny, Ghane, Legg, Huynh, & Andrews, 2014). Therefore, we propose that those with more positive attitudes toward genetic testing following exposure to the narrative will report greater intentions to talk to a provider about genetic counseling (H8). Genetic counseling involves a clinical risk assessment with a genetics professional, genetic testing (if appropriate), and a discussion of recommendations for managing genetic risk (National Cancer Institute, 2022). At the time of study launch (late 2017), testing for variants such as BRCA was done only clinically; thus, discussions with a care provider were the primary way to initiate counseling and/or testing.

Method

Women ages 18–49 participated in a randomized controlled experiment, which was approved by (*blinded)) and administered by an online survey research panel, coordinated by Qualtrics, Inc. Men were not recruited because the narrative featured women undergoing genetic testing; furthermore, this age range is representative of the characters and viewers of the show (Lerner, 2018, January 17). After providing consent online, participants completed a baseline survey (Time 1; T1). One week later (Time 2; T2), participants were randomly assigned to one of two narrative conditions (high or low accuracy). Three hundred eighty-eight participants in these two conditions completed surveys at both time points. Participants taking less than 5 minutes to complete the baseline survey or less than 10 minutes to complete the follow-up survey (n = 58), provided incomplete or questionable data (n = 6) or indicated they had not viewed the video (n = 12) were excluded. Thus, our effective sample size was 312.

Study Stimuli

Participants in the narrative conditions viewed a low or high-accuracy version of an edited episode of Keeping Up with the Kardashians from Season 11. The storyline begins with a family discussion of an opportunity to obtain genetic testing, including testing for a BRCA gene mutation, which increases risk for breast and ovarian cancer (Antoniou et al., 2003; Metcalfe et al., 2004). After the discussion, an executive vice president from Pathway Genomics is shown discussing testing with Kris Jenner, telling her that they will be looking for genetic mutations in her bloodstream that are associated with certain cancers and referencing her family history of cancer. Later, Kris is shown getting her blood drawn. The scene then cuts to Kourtney Kardashian entering the room to get her blood drawn for genetic testing and later to Kris having a discussion with Khloe Kardashian about genetic testing. The executive from Pathway Genomics later enters the room to talk to Khloe about testing, recommending she test because of her family history. Despite encouragement from family, Khloe is initially more reluctant to test, as she is concerned that the information could increase feelings of fear or impact her quality of life. She eventually decides to test after conversations with her family, who emphasize the benefits of testing to her health. Khloe and Kim Kardashian are later shown meeting with a medical engineer from Pathway Genomics. After the medical engineer explains that they will first take a medical history, Khloe is shown getting her blood drawn and being informed that she is being tested for a hereditary cancer disposition (specifically, BRCA 1 and 2 mutations). Finally, all four main characters (Kris Jenner, Kim, Kourtney and Khloe Kardashian) are shown traveling to San Diego to obtain results from Dr. Paul Nassif, who informs them that they do not have a BRCA gene mutation.

The episode was reviewed by a licensed genetic counselor to determine narrative inaccuracies, including statistics regarding cancer incidence articulated by Dr. Nassif, incorrect statements by family members about the types of cancers for which a BRCA mutation may increase risk (i.e., cervical cancer), lack of clarity about the type of testing performed, confusion of genes and genetic mutation (i.e., Dr. Nassif informing them they did not test positive for the BRCA gene, rather than mutation), and sensationalizing the potential impact of genetic information. There also were instances where genetic counseling and testing did not appear to align with practice recommendations at the time. The low-accuracy narrative contained the complete genetic testing storyline. In the high-accuracy narrative, major inaccuracies were removed. To ensure the two narratives were similar length (around 12 minutes), non-genetics footage from the episode (i.e., a family photo shoot and drive to San Diego) was included in the high-accuracy narrative.

Measures

Unless noted, all items were measured on a 1 (strongly disagree)-5 (strongly agree) scale.

Narrative Engagement

Identification with major characters in the narrative (Kris Jenner or Kim, Kourtney and Khloe Kardashian) was measured with six items (Cohen, 2001). Two items were dropped from the original scale (i.e., “While viewing the video, I wanted the [character] to experience a positive ending” and “When the outcome was positive, I felt joy”) given that the story outcome may be interpreted as positive to some and neutral or negative to others. Confirmatory factor analysis (CFA) indicated that a one-factor structure (for each character) provided an acceptable fit to the data (RMSEA = .09, CFI = .89, SRMR = .06, χ² (132) = 446.65 (p < .01), standardized item loadings from .53 to .89), per established recommendations (Brown & Cudeck, 1993; Hu & Bentler, 1999). Scale reliabilities were also high (Cronbach’s α’s = .86-.94). The mean of the items (collapsed across characters/conditions) was used in the analyses (M = 3.49, SD = .73).

Transportation was measured with ten items from Green and Brock (2000). An initial CFA showed a poor fit to the data. As in a previous study (Murphy et al., 2013), the reverse coded items loaded weakly on a single factor and were dropped from the model. With seven items, the one-factor structure of transportation was confirmed (RMSEA = .09, CFI = .96, SRMR = .03, χ² (14) = 45.45 (p < .01), standardized item loadings from .67 to .82). The mean of the items was used in the analyses (Cronbach’s α = .90, M = 3.29, SD = .90).

Narrative Resistance

Counterarguing was assessed with four items from Moyer-Gusé, Mahood, and Brookes (2011). CFA confirmed the one-factor structure of the items (RMSEA = .19, CFI = .93, SRMR = .04, χ² (2) = 23.77 (p < .01) with standardized item loadings from .67 to .91), although the fit was not outstanding. The mean of the items was used in the analyses (Cronbach’s α = .90, M = 2.55, SD = .91).

Knowledge

Knowledge accuracy was measured at T1 and T2 with seven true/false items (e.g., “All women have BRCA (breast cancer) genes.”). An explicit attempt was made to ensure the items used to measure knowledge matched the content of the narratives. No existing scale fully encompassed all of the information provided in the narratives, thus original and existing scale items were used to measure knowledge (from Fitzgerald‐Butt et al., 2016; Furr & Kelly, 1999; Vadaparampil et al., 2011). Participant responses to each item were scored as correct (1) or incorrect (0) and summed at T1 (range: 0–7, M = 4.41, SD = 1.50) and T2 (M = 4.41, SD = 1.33).

Attitudes

Attitudes toward genetic testing (i.e., receiving a test to assess my cancer risk) were measured at T1 and T2 using six 10-point semantic differential scales (e.g., worthless/valuable, bad/good, useful/not useful; Ajzen, 1991). The items showed high reliability at T1 (Cronbach’s α = .97, M = 7.86, SD = 2.03) and loaded on a single factor (RMSEA = .06, CFI = .97, SRMR = .02, χ² (9) = 19.79 (p < .02) with standardized loadings from .89 to .94. Similar results were obtained for attitudes at T2 (RMSEA = .04, CFI = .99, SRMR = .01, χ² (9) = 12.62 (p > .05), with standardized loadings from .92 to .96, Cronbach’s α = .98, M = 7.98, SD = 2.17).

Genetic Counseling Intention

Intention was measured with a single item assessing the likelihood that one would “talk to a health care provider about my getting genetic counseling” (M = 3.12, SD = 1.06).

Individual Characteristics

Measures of age (continuous), race and ethnicity, and income (from 1 = ≤ $10,000 to 8 = ≥$200,000) were based on existing measures (Centers for Disease Control and Prevention. (2011), Centers for Disease Control and Prevention. (2013)). Previous testing experience (i.e., have you/partner or children ever had a genetic test) and personal/family cancer history were assessed, as well as perceived cancer vulnerability (“On a scale from 0–100, how likely are you to develop cancer/another cancer during your lifetime?”). The perceived risk item was rescaled to 0–10 for better alignment with other measures. Individuals also reported how often they watched reality television (hours and shows per week) and whether they regularly watched Keeping Up with the Kardashians (yes or no; Papacharissi & Mendelson, 2007).

Data Analysis

Descriptive statistics and Pearson correlations between model variables (values in the range of <.01 to .71) were first assessed. Multiple regression analyses were conducted in SPSS to examine study hypotheses and assess direct and indirect relationships between model variables. All analyses controlled for T1 age, income, and perceived cancer susceptibility. Viewers who perceive greater cancer susceptibility may identify more with characters or, alternatively, may resist the message due to their perceived invulnerability (Moyer-Gusé & Nabi, 2010), which is associated with behavioral intention (Bandura, 1986; Fishbein, 2009; Moyer-Gusé, 2008; So & Nabi, 2013) and interest in genetic testing (Bosompra et al., 2000; Lerman, Daly, Masny, & Balshem, 1994). Indirect effects of identification and transportation on attitudes and genetic counseling intentions via narrative resistance variables were assessed using the PROCESS macro for SPSS (Hayes, 2012).

Results

Participant Demographics

The mean age of participants was 36.55 years (SD = 7.86). The majority were non-Hispanic White (68.7%; n = 213). Most participants had incomes of $50,000 per year or more (56.7%; n = 165). Few (3.9%, n = 12) had a previous cancer diagnosis, but 31.9% (n = 99) had at least one parent, sibling or child diagnosed with cancer. Just over 11% (n = 36) had undergone genetic testing (any type) and 4.6% (n = 14) had a partner or child who had. Perceived cancer susceptibility was moderately low (M = 4.31, SD = 2.35). Independent samples t-tests and chi-square tests showed no differences between conditions on these key demographic variables (p < .05). Close to 68% (n = 211) regularly watched reality television and just over 16% (n = 52) regularly watched Keeping Up with the Kardashians. Forty participants (12.8%) had previously viewed the episode tested here. Because this study focused on a single storyline from a larger episode that appeared on television over a year prior to data collection, and due to the fact that no significant differences (p > .05) were detected between viewers and non-viewers of the episode on any main study variable (i.e., attitudes, knowledge, counseling intention and measures of narrative engagement), previous viewers were included in the analyses.

Main Study Hypotheses

We first examined knowledge and attitude differences by condition following exposure to the narrative (T2), controlling for baseline (T1). At T1, no condition differences were detected on knowledge (t (310) = −1.134, p > .05) or attitudes (t (309) = −.096, p > .05). However, at T2,

differences were detected on knowledge (b = −.26, SE b = .13, b* = −.11, p < .05); specifically, participants in the high-accuracy condition (M = 4.47, SD = 1.45) had more accurate knowledge than those in the low-accuracy condition (M = 4.34, SD = 1.18; H1 supported). A post-hoc t-test examining knowledge change scores (T2-T1) further showed differences between conditions (t (310) = −2.03, p < .05), whereby those in the low-accuracy condition experienced a slight drop in knowledge over time (M = −.17, SD = 1.36). No condition differences were detected on attitudes at T2 (p > .05), although T1 attitudes were associated with T2 attitudes (b = .690, SE b = .052, b* = .631, p < .01).

Narrative Engagement and Resistance

No significant condition differences were detected on identification, as expected (H2a supported). However, participants in the low-accuracy condition (M = 3.46, SD = .81) reported greater transportation into the narrative than those in the high-accuracy condition (M = 3.14, SD = .96, b = .27, SE b = .11, b* = .15, p < .05; H2b not supported).

Results examining narrative resistance (Table 1) showed age and identification to be negatively associated with counterarguing (p < .05; H3 supported), while transportation was unassociated with counterarguing (p > .05; H4 unsupported). Counterarguing, in turn, was negatively associated with attitudes (p < .05; H6 supported); whereas those who experienced greater transportation reported more positive attitudes toward genetic testing (p < .05). T2 knowledge was unassociated with attitudes (p > .05; H7 unsupported). We also did not detect an indirect effect of knowledge on intention via attitudes (est. = .013, 95% CI [−.027, .067]).

Table 1. Results of multiple regression analyses for models predicting message counterarguing, knowledge, genetic testing attitudes and genetic counseling discussion intentions

	Model 1		Model 2		Model 3		Model 4
	MessageCounterarguing^a		Genetics Knowledge^b		Genetic Testing Attitudes^c		Counseling Discussion Intentions^d
	b	SE b	B	SE b	b	SE b	b	SE b
Condition	.065	.107	−.224	.128	.012	.193	−.064	.102
Age	−.018	.007*	.009	.008	−.014	.013	−.008	.007
Income	−.012	.029	.006	.035	.014	.053	−.010	.028
Perceived Vulnerability	.029	.023	.025	.027	.007	.041	.015	.022
Identification	−.274	.107*	.017	.130	.036	.195	.004	.103
Transportation	−.138	.087	−.059	.105	.378	.157*	.410	.084**
Counterarguing			−.083	.073	−.421	.117**	−.064	.062
Knowledge (T1)			.369	.045**
Knowledge (T2)					.077	.082	−.076	.044
Attitudes (T1)					.584	.053**
Attitudes (T2)							.177	.027**
R^2	.12		.22			.48		.38

Notes: ^a Test of Hypothesis 3 (H3) and H4; ^b Test of H5; ^c Test of H6 and H7; ^d Test of H8* p < .05, ** p < .01

Contrary to our predictions, counterarguing was unassociated with genetics knowledge (p > .05, H5 unsupported). Furthermore, tests of indirect effects provided no evidence that counterarguing mediated the effects of identification (est. = .033, 95% CI [−.027, .095] or transportation (est. = .026, 95%f CI [−.017, .081] on participant knowledge.

Finally, as predicted, participants with more positive attitudes toward genetic testing following exposure to the narrative had greater intentions to talk to a doctor about genetic counseling (Table 1, p < . 01; H8 supported). Results also showed a direct and positive relationship between transportation and intentions (p < .05). Tests of indirect effects further showed effects of transportation (est. = .011, 95% CI [.002, .029]) and identification (est. = .020, 95% CI [.004, .047]) on intentions via counterarguing and attitudes.

Discussion

We examined the effects of exposure to high and low-accuracy genetic testing narratives featuring well-known celebrities. Our results offer some support for study predictions (Table 2) and provide practical and theoretical insights regarding the effects of entertainment narratives on genetics-related knowledge, attitudes, and counseling intentions. The most striking finding was that exposure to a low-accuracy narrative was associated with lower knowledge accuracy. Furthermore, those viewing the low-accuracy narrative exhibited a significant decrease in knowledge after viewing.

Table 2. Summary of study findings

Dependent Variable	Associated Prediction	Results
Knowledge	Participants who view a low- accuracy narrative (versus a high-accuracy narrative) will exhibit lower genetics knowledge post-exposure. (H1)	Supported*
	Lower counterarguing will be associated with lower knowledge. (H5)	Not Supported
Identification	Identification will not differ between those viewing a low versus high accuracy narrative. (H2a)	Supported*
Transportation	Transportation will not differ between those viewing a low versus high accuracy narrative. (H2b)	Not Supported
Counterarguing	Greater character identification will be associated with less counterarguing. (H3)	Supported*
	Greater transportation will be associated with less counterarguing. (H4)	Not Supported
Attitudes	Lower counterarguing will be associated with more positive attitudes toward genetic testing. (H6)	Supported**
	Greater knowledge will be associated with more positives attitudes toward genetic testing. (H7)	Not Supported
Genetic Counseling Discussion Intention	Positive attitudes toward genetic testing will be associated with greater intentions to talk to a doctor about genetic counseling. (H8)	Supported**

Notes: * p < .05 ** p < .01

Because people’s knowledge of genetics often derives from the internet and the media (Dougherty, Lontok, Donigan, & McInerney, 2014; Roberts, Archer, DeWitt, & Middleton, 2019), the negative effect of exposure to inaccuracies found here is concerning; particularly since participants had moderate levels of genetic knowledge to begin with. Individuals who view entertainment narratives about genetics may presume the information presented is realistic and accurate (Papacharissi & Mendelson, 2007) and passively accept false information (Dahlstrom, 2021). The narrative format may have contributed to perceptions of source credibility; for example, participants who view scientists presenting scientific information in a first-person narrative style (versus a control) rate them as more authentic (Saffran et al., 2020).

The narrative we examined made specific mention of BRCA mutations associated with increased risk of breast and ovarian cancer, which may develop at a younger age than people without the mutations (National Cancer Institute, 2021). It is recommended that individuals with a BRCA mutation begin breast magnetic resonance imaging (MRI) to screen for cancer at age 25 (Elezaby et al., 2019) and to consider risk-reduction salpingo-oophorectomy (RRSO) between ages 35–40 (Daly et al., 2021), which is in the age range of many viewers of Keeping Up with the Kardashians (Lerner, 2018, January 17). Given its’ status as a flagship show of the E! network (Perez, 2017), and the popularity of the Kardashians (e.g., surveys show Kim Kardashian is the third most famous contemporary U.S. television personality (YouGovAmerica, 2021)), it is especially concerning that women at-risk of HBOC might view the show and endorse inaccuracies contained within. Because audiences can have trouble identifying narrative inaccuracies, it may be difficult to correct them with facts (Dahlstrom, 2021).

Our findings suggest that media producers should consult with experts to ensure the accuracy of genetics content, which the public may know less about (Christensen et al., 2010; Gupta et al., 2021), particularly given the increase in storylines featuring genetic testing (e.g., 90210 and The Bold Type). Still, it is not a given that narrative accuracy will increase genetic knowledge or literacy. For example, although 75% of Americans were aware of Angelia Jolie’s surgery following her editorial, only 10% could accurately interpret her cancer risk (Borzekowski, Guan, Smith, Erby, & Roter, 2014). Research on the effects of exposure to genetic testing narratives, accurate or inaccurate, is therefore warranted. While news media aren’t over-sensationalizing the topic of genetics, they may focus more on the benefits of testing than limitations or challenges (Hicks-Courant et al., 2021). Thus, research examining genetic message framing, with the accuracy of the information held constant, seems particularly warranted.

One hopeful finding is that narrative inaccuracies in the storyline appeared to have little effect on public attitudes toward genetic testing, as no differences were detected between conditions following narrative exposure. While the lower accuracy narrative did sensationalize the potential impacts of testing to a larger degree than the higher accuracy narrative, this difference was minimal and likely undetectable to the general public. Both messages also framed testing in a positive and worthwhile light. Public attitudes toward genetic testing were quite positive at baseline (T1) and remained so following narrative exposure (T2). Furthermore, T1 attitudes, along with transportation and message counterarguing, were most influential on attitudes following exposure to the narrative.

Interestingly, knowledge was unassociated with genetic testing attitudes at T2 or intentions to discuss genetic counseling with a healthcare provider. Although it failed to reach statistical significance, parameter estimates suggest greater knowledge may be negatively associated with intentions to speak to a provider about genetic counseling. This negative relationship could reflect people’s awareness of their own risk or how relatively rare these mutations are in the general population (e.g., only about 5–10% of breast cancer and 10–15% of ovarian cancers are hereditary; Centers for Disease Control and Prevention, 2022). Either way, the results validate the need for research to understand the complex relationship between knowledge, attitudes and behaviors (Etchegary, 2014).

While the messages did not have differential effects on attitudes, our results do show that those with more positive attitudes toward genetic testing had greater intentions to talk to a doctor about counseling. Thus, attitudes are an important factor for message designers to consider. If the goal of narrative messaging is to increase genetic counseling intentions, messages may need to reinforce or seek to increase positive attitudes about genetic testing (i.e., promoting testing and counseling as worthwhile, useful or good). Message designers seeking to influence attitudes using narratives should consider the way genetic information is framed (e.g., gain versus loss) as that could impact beliefs (Condit & Shen, 2011). The use of simple metaphors is also shown to support learning of genetic topics (Kaphingst et al., 2009), which may impact attitudes.

In addition to attitudes, our data suggest that narrative engagement, via reduced counterarguing and attitudes, is influential on intentions to discuss genetic counseling with a care provider. Indirect effects of character identification and transportation on intentions to talk to a doctor about genetic counseling were detected via counterarguing and attitudes, although tests of their direct relationships suggest that identification and transportation functioned differently in this context. The effects of identification on intention appear to have been a function of reduced counterarguing, whereas the effect of transportation on attitudes and intentions was direct. Of the variables tested, transportation was the most influential (even more influential than attitudes) on intentions to discuss genetic counseling with a health care provider.

Studies often show differential effects for identification and transportation (see review by Tal-Or & Cohen, 2016); thus, our findings fit within the larger literature. While the lack of relationship between transportation and counterarguing was unexpected based on theory, some have suggested that increased transportation into the narrative may make it more difficult for viewers to detect things that seem wrong about the narrative and counter argue (Green & Brock, 2000; Tal-Or & Cohen, 2016). While we cannot be certain, increased transportation due to the entertaining format of the narrative may have limited participants’ ability to counterargue, as well as their ability to detect misinformation. Interestingly, transportation scores were higher in the low-accuracy condition than in the high-accuracy condition. Although both narratives featured Khloe Kardashian, the low-accuracy narrative featured more of her story, including her indecision regarding genetic testing and emotional scenes with family members; thus, differences in transportation may have reflected increased engagement.

Either way, our findings illustrate the critical role narrative engagement plays in shaping attitudes and intentions in regard to genetic counseling and testing, as well as the importance of message accuracy. Still, given that model variables explained only 38% of the variance in intentions to discuss genetic counseling with a health care provider, there are clearly other factors influencing these intentions that should be identified. Genetic testing narratives are intrinsically persuasive (Dahlstrom, 2014) and can help put a human face to the science (Henderson & Kitzinger, 1999); they are also likely to increase in prevalence over time as more people seek out testing. Our study highlights the importance of research regarding their effects and investigations into how narratives might be used as an intervention strategy to educate individuals and families about genetic testing.

Limitations

Because we collected data at only two time points, we were not able to determine whether exposure effects were long-lasting or impacted actual testing behaviors. We also did not include a non-narrative control in our analyses or compare multiple narratives due to our focus on the effects of misinformation. Furthermore, the items measuring perceived cancer susceptibility were not specific. Finally, because the narrative featured well-known characters, it is possible that Existing beliefs about the characters, perceived character credibility or likability may have influenced narrative engagement (identification and transportation) and subsequent attitudes and intentions; thus, more research is needed. Character liking, for example, could reduce reactance and decrease resistance to persuasion, which should increase the likelihood of story-consistent attitudes and behavioral intentions (Moyer-Gusé, 2008; Slater & Rouner, 2002).

Conclusions

Results suggest engaging narratives are more likely to influence attitudes and genetic counseling intentions, while showing that identification and transportation may have differential effects. Most importantly, our data illustrate the negative implications of misinformation embedded within a narrative on public knowledge accuracy.

Disclosure Statement

No potential conflict of interest was reported by the author(s).

Correction Statement

This article has been republished with minor changes. These changes do not impact the academic content of the article.

Additional information

Funding

The author(s) reported there is no funding associated with the work featured in this article.