Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2022 Aug 1.
Published in final edited form as: Drug Alcohol Depend. 2021 May 28;225:108788. doi: 10.1016/j.drugalcdep.2021.108788

Cognitive and Affective Responses to Marijuana Prevention and Educational Messaging

Glenn Leshner 1, Elise M Stevens 2, Amy M Cohn 3, Seunghyun Kim 6, Narae Kim 1, Theodore L Wagener 4, Andrea C Villanti 5
PMCID: PMC8282745  NIHMSID: NIHMS1709423  PMID: 34119879

Abstract

Background:

Perceptions of risk of using marijuana have decreased significantly in the US over the last decade, while marijuana use has increased. In order to educate people on the risks associated with marijuana use, large-scale health messaging campaigns have been deployed to educate the public about the risks associated with marijuana use, particularly in states where medical or recreational marijuana is legal. Few studies have examined how messages about marijuana affect the audiences’ cognitive and emotional responsivity to these messages.

Methods:

To address this knowledge gap, this study used psychophysiological assessment (heart rate, skin conductance, facial action coding) and self-report measures to explore the impact of different marijuana risk messages on real-time cognitive and affective responses and self-reported message receptivity, likeability, and intentions to use marijuana in a sample of 50 young adult marijuana users and non-users. Each participant saw six messages. Three messages were used from each of two campaigns, representing one of three risks (cognitive ability, driving, health harms).

Results:

Psychophysiological responses showed that the driving-themed messages for both campaigns had the greatest cognitive resource allocation to encoding the message, the greatest arousal, and the most positive emotional response, regardless of user status. Self-reports showed a less consistent pattern.

Conclusions:

Overall, psychophysiological measures provided a more consistent picture of message processing and effects than self-report measures. Findings from this study provide immediately useful data for improving the development and effectiveness of marijuana health-risk prevention campaigns by elucidating cognitive and emotional processes that could be targeted in future programs.

Keywords: marijuana, psychophysiology, educational messages, cognitive and affective processing

1. Introduction

Marijuana is the most commonly used federally illicit drug. Past year and past 30-day marijuana use have both increased significantly among US adults aged 18 years and older, from 10.4% (past year) and 6% (past 30-day) in 2002 to 18.0% (past year) and 11.9% (past 30-day) in 2019 (SAMHSA, 2019). Notably, 23% of adults 18-25 years report past 30-day marijuana use (SAMHSA). Studies consistently cite a link between lower marijuana harm perceptions—a primary reason for the appeal of marijuana use—and greater likelihood of marijuana use (Azofeifa et al., 2016; Compton et al., 2016; Danseco et al., 1999; Lipari & Ahrnsbrak, 2017). This association may be due, in part, to a general lack of public knowledge of the health risks of marijuana use and dramatic changes in the cultural norms surrounding marijuana use. For example, according to a Pew Research Center poll, 61% of Americans believe marijuana use should be legalized, and nearly 190 million Americans now live in a state where some level of marijuana is legal (Medical Marijuana, 2016; Pew Research Center, 2018). The decrease in marijuana harm perceptions is particularly concerning given that its use is associated with a variety of health and psychosocial problems (Hasin, 2018), including high comorbidity with other drugs of abuse, mental health conditions, poor tobacco cessation outcomes, memory and cognitive impairment, motor vehicle accidents, and in some cases, increased cancer risk (Callaghan et al., 2013; Compton et al., 2016; Hasin, 2018; Huang et al., 2015; National Academies of Sciences, 2017).

Public health messaging campaigns (deployed via print, radio, television, web, or mobile platforms) are well suited for disseminating information about marijuana health risks to a large and diverse audience to affect behavior change. For tobacco use, media campaigns have been shown to be effective at decreasing cigarette smoking, increasing utilization of quitting services, and changing social norms related to tobacco use (Centers for Disease Control and Prevention, 2001, 2016), although ample research has shown “boomerang effects” (i.e., when messages designed to promote certain outcomes produce the opposite effect, such as defensive processing, increasing counter-recommended attitudes and/or behaviors, psychological reactance), for health message effectiveness in many health domains (e.g., Clayton et al, 2020; LaVoie et al., 2015; Leshner et al., 2009).

A recent systematic review and meta-analysis of the effectiveness of these types of public education campaigns targeting younger audiences showed only modest evidence of their effectiveness on changing drug use, including marijuana use (Allara et al., 2015; Ferri et al., 2013). Many of the studies included in this review, among others, have pointed to the problem of limited rigorous evaluation studies of marijuana messaging campaigns as a reason that might explain weaker than expected behavior change outcomes (Cermak and Banys, 2016; Gates, 2006). Several studies that examined the effectiveness of anti-marijuana messages by Ginsburg and his colleagues showed boomerang effects. For example, Czyzewska & Ginsburg (2007) found that individuals who watched anti-marijuana ads experienced reactance and rated marijuana significantly less negatively than individuals who watched the anti-tobacco ads and rated tobacco. Individuals who watched anti-marijuana ads also reported higher intentions of using marijuana afterward. Also, Ginsburg & Czyzewska (2005) found that college students made more negative comments in response to televised anti-marijuana ads than anti-tobacco ads and perceived the marijuana ads as exaggerated and unbelievable.

Most evaluations of marijuana public health messaging campaigns to date (e.g., Crano et al., 2017; Hornik et al., 2008; Nonnemaker et al., 2012; Zimmerman et al., 2014) have relied on retrospective self-report paper-pencil or focus group measures of message effectiveness, likeability, and appeal (c.f., Czyzewska and Ginsburg 2007, which measured attitudes via the IAT), which may be one reason underlying their weak impact on behavioral and attitudinal change outcomes. These measures do not capture the series of information-processing steps that occur automatically and outside of conscious awareness when individuals are confronted with a persuasive health message, nor do they measure, with a sufficient level of detail, the processes of health message encoding that are not visually perceptible and accessible through direct self-report (Cacioppo and Gardner, 1999; Larsen et al., 2008). These processes include changes in attention, arousal, and emotional response—implicit processes that have been shown to later result in attitudinal and change-related intentions (Dillard et al., 2007; Rice and Atkin, 2012).

Psychophysiological measurements of media and communication content provide observable and objective assessment of these processes that operate outside of conscious awareness and are well validated correlates and predictors of change-related intentions and risk perceptions (Cacioppo and Gardner, 1999; Clayton, Leshner, Bolls et al., 2017; Clayton, Leshner, Tomko et al., 2017; Larsen et al., 2008; Leshner et al., 2018; Leshner et al., 2011; Ravaja, 2004). Psychophysiological assessment techniques have several advantages over self-report methods: 1) cognitive and affective processes that unfold during message exposure and encoding can be objectively measured and assessed in “real time” (Larsen et al., 2008; Potter and Bolls, 2012; Wang and Lang, 2006); 2) biases or other demand characteristics are less likely to be used in judgement-making; and 3) assessments are not limited to what participants explicitly recall. The purpose of this study is to examine through psychophysiological measures, the cognitive and emotional processes engaged by anti-marijuana messages. This study also compares those processes to typical self-report measures, which to our knowledge, is the first study of its kind to take this approach in evaluating anti-marijuana campaign messages.

Changes in heart rate during message exposure, for example, can indicate the extent to which viewers allocate cognitive resources to encoding message content (Ravaja, 2004). Evidence from prior work shows that increased heart rate in response to viewing anti-tobacco public service announcements is correlated with increased smoking urges and reduced intentions to quit smoking among smokers (Clayton, Leshner, Tomko, et al., 2017), which led Clayton and colleagues to argue that this pattern of outcomes was an indication of motivational dissonance, resulting in avoidance of the messages.

Cognitive resource allocated to message encoding, as evidenced by decreasing heart rate, is thought to be a necessary condition for persuasion and is often validated with a self-report message recognition task taken after exposure (Lang, 2006). Further, sympathetic nervous system arousal, as measured by skin conductance, is greater when viewing high- vs low-stimulation images or movies (Hubert and de Jong-Meyer, 1990; Lang, 1990). As such, data revealed through psychophysiological assessment is qualitatively different than data collected through traditional retrospective reports and will add significant value to our understanding of the processes related to marijuana messaging receptivity and appeal.

Past research indicates that marijuana users and non-users respond differently to marijuana health messages. Research utilizing self-report measures has shown that non-users of marijuana tend to respond to public health messaging campaigns with greater perceived effectiveness and more positive attitudes toward the message (Alvaro et al., 2013; Stevens et al., 2019), and lower intentions to use marijuana in the future, even at a 1-year follow-up (Alvaro et al., 2013). While non-users may respond to these messages in more favorable way, it is important to know whether marijuana users may respond in the opposite way that the message was intended (Hornik et al., 2008). One study showed that increased exposure to an anti-marijuana messaging campaign caused audiences to increase their intentions to use marijuana (Hornik et al., 2008), which can be even more prominent among those who use the product presented in the message (Dillard and Shen, 2005; Stevens et al., 2019; Wolburg, 2006). Thus, marijuana use status is an important factor in determining responsiveness to public health messaging campaigns about marijuana.

We chose to include parallel themed messages from two different anti-marijuana campaigns with the goal of providing a conceptual replication. Consistency of responses across campaigns may indicate processes that are more generalizable than if messages were selected from only one campaign (Leshner, 2014; Reeves and Geiger, 1994). If responses diverge as a function of campaign, then those responses are likely to be campaign-specific. The strategy of using real-world, existing messages also affords greater ecological validity.

Research to date has not taken advantage of promising methods that capture cognitive and emotional factors the arise in response to marijuana prevention message exposure, despite their substantial potential to elucidate processes that could be targeted to enhance the effectiveness of these messages. To address this gap, this exploratory study employed a mixed methods approach of psychophysiological assessment (heart rate, skin conductance, facial action coding) and traditional self-report measures, to prospectively explore the impact of real-time cognitive and affective responses to marijuana public health messaging on message receptivity, likeability, and intentions to use marijuana in a sample of 50 young adults ages 18 to 25 who viewed 6 anti-marijuana messages across two real-world campaigns. We examine marijuana use status (current use vs non-current use) as a factor that could impact message responsiveness. We focused specifically on young adults because rates of marijuana use continue to increase in this age group, relative to youth or older adults (Azofeifa et al., 2016; Lipari et al., 2014). Our basic research question is: How do marijuana use status and message themes from two separate campaigns impact real-time cognitive and affective responses and self-reports (message receptivity, likeability, and intentions to use marijuana)?

2. Method

2.1. Design

The design of this study is a 2 (marijuana use status: current user/ non-user) × 3 (message theme: cognitive ability/driving/health harms) × 2 (campaign: Spread the Facts/Do the Math) mixed model repeated measures experiment. User status was a between-subjects factor. Message factors (message theme, campaign) were within-subjects factors. Participants viewed six messages based on combinations of each of three themes within two campaigns, which were presented in a random order. Exposure time was set at 10-seconds for each message (based on pilot testing) and was included as a within-subjects factor for analyses that included exposure time.

2.2. Independent Variables

2.2.1. Message theme.

Three messages were selected based on the availability of the message themes across the two anti-marijuana campaigns. Message themes targeted marijuana risks related to decreased cognitive ability, driving impairment, or general health harms related to use (see Appendix for messages).

2.2.2. Campaigns.

Messages were selected from two existing, real world print-based anti-marijuana campaigns: Spread the Facts, developed by the National Institute on Drug Abuse for Teens (https://teens.drugabuse.gov/) and Do the Math, developed by the Liberty Alliance for Youth (http://libertyalliance4youth.com/). These campaigns were selected because, at the time of the study, these were among a very small number of campaigns that were being disseminated in the United States.

2.2.3. User status.

Participants were asked about their use of marijuana: never used, tried it once, at least once a year but not monthly, at least once a month but not weekly, at least once a week, but not daily, at least once a day but not daily, or at least once a day or most days of the month. Participants were categorized as either marijuana users or never-users. Users were defined as those who reported using marijuana in the past 30 days. Non-users were those who had never tried marijuana or who tried it once.

2.3. Dependent Variables

2.3.1. Psychophysiological measures

2.3.1.1. Heart rate.

Heart rate is an indicator of cognitive resources allocated to encoding the message and was measured from the photoplethysmogram (PPG) pulse. One PPG sensor was applied to the forefinger tip of the participant’s left hand with a Velcro strap and connected to a Shimmer 3 galvanic skin response (GSR) module. The raw data were sampled at 128 Hz. Heart rate was converted to beats per minute (BPM) values in iMotions (2018) biometric software. Heart rate change scores for each second of message exposure were computed by subtracting the last second of the baseline (measured prior the onset of each message) from scores for each second of message exposure. Heart rate acceleration can indicate multiple processes, such as activation of flight/fight responses (which should also correlate with an increase in arousal), retrieval from long-term memory (as would be expected during reading), and other internal cognitive processes, such as counterarguing. Heart rate deceleration typically suggests greater cognitive resources being allocated to encoding the message (Potter and Bolls, 2012).

2.3.1.2. Arousal.

GSR is an indicator of sympathetic nervous system arousal. Two GSR electrodes were placed on the distal phalanges of participants’ nondominant hand. These electrodes were connected to a Shimmer 3 GSR module. The raw GSR data were sampled at 128 Hz and expressed in micro-siemens (μS) in iMotions. The GSR data were converted to change scores by subtracting the last second of the baseline from scores for each second of message exposure. Increases in GSR suggest a greater arousal response to the message (Potter and Bolls, 2012).

2.3.1.3. Emotional valence.

Emotional valence was assessed by computer-based automatic facial expression analysis software installed in iMotions (FACET). Participants’ real-time facial expression data were collected via a separate video camera attached to the computer monitor during message exposure. The automatic facial coding software detects and analyzes any movement or change of participants’ facial landmarks (e.g., eyes, eyebrows, and mouth corners, etc.) corresponding to the Facial Action Coding System (FACS). FACET provides numeric values (evidence scores) of the presence of each of the valence dimensions (negative and positive). Negative valence reflects activation of the aversive (avoid) motivational system, while positive valence reflects activation of the appetitive (approach) motivational system (Lang, 2009; Lang, 1994; Lang & Bradley, 2008; Lang, Bradley, & Cuthbert, 1997). FACET evidence scores were used to analyze participants’ valence and were interpreted as the probability of categorizing the current facial expression as either negative or positive (iMotions, 2018), with higher values indicating a greater likelihood of the particular motivational response.

2.3.2. Self-report measures

2.3.2.1. Perceived effectiveness.

Perceived effectiveness was measured with an averaged composite index of six items. Participants recorded their response on a 1 (strongly disagree) to 5 (strongly agree) response scale to the following statements: (1) “This ad is worth remembering; (2) This ad grabbed my attention; (3) This ad is powerful; (4) This ad is informative; (5) This ad is meaningful; (6) This ad is convincing” (α= .97; Davis et al., 2013).

2.3.2.2. Intention to quit/avoid marijuana use.

Intention to quit or avoid using marijuana was measured in accordance with user status with an averaged composite index of four items. That is, users were asked about their intentions to quit and non-users were asked about their intentions to avoid. Participants recorded their response on a 0 (Not at all) to 10 (Extremely) response scale to the following four questions: (1) How important is it to you that you quit/avoid using marijuana in the next month? (2) How confident are you that you can quit/avoid using marijuana in the next month? (3) How ready are you to quit/avoid using marijuana in the next month? (4) How committed are you to quit/avoid using marijuana in the next month? (α=.93; indexes adapted from Boudreaux, et al., 2012).

2.3.2.3. Message acceptance

Message acceptance was measured with an averaged composite index of three items. Participants recorded their response on a 1 (strongly disagree) to 5 (strongly agree) response scale to the following statements: (1) I think this message makes a strong argument”, (2) “I will adopt this message’s recommendations”; (3) “This message convinced me of its recommendations” (α=.91).

2.3.2.4. Attitude toward the message.

Attitude toward the message was measured with an averaged composite index of three items. Participants recorded their response to the item stem “This ad was…” on a 1—7 scale to the following bipolar items: (1) bad/good, (2) negative/positive, (3) dislike/like (α=.93).

2.4. Demographics.

Age, race/ethnicity, and biological sex were collected with the baseline questionnaire prior to message exposure.

2.5. Procedures

The study took place in the media psychology laboratory at a large southwestern university, which is specifically designed for the assessment of cognitive and emotional processing of media and communication content using psychophysiological data collection. Fifty participants aged 18 to 25 were recruited through mass email and flyers around campus and included students, staff, and community members. After a brief screening over the phone, eligible participants were invited to complete a 1-hour experimental session in which they provided informed consent, completed the baseline questionnaire and then the psychophysiological measurements during exposure to each of six different anti-marijuana print messages, randomly presented, on a 22-inch computer monitor. Each message appeared for ten seconds in order to provide participants sufficient time to examine the message (exposure length was determined in a pilot test). Before each message appeared, participants viewed a black screen for 10 seconds to permit the collection of baseline levels of heart rate, arousal, and emotional valence coding. This design is recommended for communication message testing and consistent with our team’s prior work and balances efficiency for study resources (e.g., Leshner et al, 2011). All experimental stimuli presentation and psychophysiological data collection were controlled by the iMotions Biometric Research Platform 7.1 (2018).

Following exposure to each message, participants completed self-reported perceived message effectiveness, intentions to avoid/quit marijuana, message acceptance, and attitude toward the message. Participants were compensated $25 for their time. The study was approved by the university’s Institutiona1 Review Board (IRB #10693).

3. Results

3.1. Participants

Fifty people (25 marijuana users, 25 marijuana non-users) participated in the study. Twenty-eight were male, 21 were female, and one opted not to answer. The mean age was 21.00 (SD=2.00; range 18-25). Thirty-six reported they were White/Caucasian, five were Black/African American; one was Asian; six selected multiple race/ethnicities, and two did not provide an answer for race/ethnicity.

3.2. Heart rate

Participants’ heart rate change scores from baseline were analyzed with a 3 (message theme: cognitive ability, driving, health harms) × 2 (campaign: Spread the Facts, Do the Math) × user status (users, non-users) × time (10 second) repeated measures ANOVA. The theme × campaign × user status × time 4-way interaction was not significant (F(18, 522) = 1.29, p=.189). However, the theme × campaign × user status interactions were further probed. Figure 1 shows how heart rate varied as a function of user status and campaign for each of three themes.

Figures 1.

Figures 1

Open in a new tab

a-c. Beats per minute change from baseline as a function of user status and campaign for each of three themes.

Overall, there was a downward change in heart rate, although slight, for all message risks, user status, and campaign. However, the most consistent pattern emerged for users across campaigns for the driving theme. A close examination of the driving-themed messages showed that heart rate deceleration was greater for users than non-users, indicated by a time × user status quadratic near-significant trend (F(1, 46.49) = 3.35, p=.073, ηp2=.10). This pattern was observed across both Spread the Facts and Do the Math campaigns. In other words, users allocated more resources to encoding message content for the driving messages than non-users during the course of message exposure. No consistent patterns for heart rate were observed for the other message themes (cognitive ability or health harms). That is, the effects of theme on heart rate differed depending on unique combinations of user status and campaign.

3.3. Arousal.

GSR change from baseline levels was analyzed with a 3 (message theme: cognitive ability, driving, health harms) × 2 (campaign: Spread the Facts, Do the Math) × user status (users, non-users) × time (10 seconds) repeated measures ANOVA. The theme × campaign × user status × time 4-way interaction was not significant (F (18, 738) = 0.63, p=.88). Figs. 2a and 2b shows the pattern for the driving and health harms themed messages. A visual inspection of the GSR data for the driving messages (Fig. 2a) shows a tendency for users to experience higher arousal than non-users, although the difference was not statistically significant. There was a significant campaign × user status × time interaction on GSR for the health harms messages (F(9, 396) = 2.09, p=.029, ηp= .045). Users experienced greater GSR over time than non-users, regardless of campaign. The GSR pattern in response to viewing the messages with the theme about cognitive ability showed no consistent patterns.

Fig. 2.

Fig. 2

Open in a new tab

a & b. GSR response as a function of user status and campaign for the driving and health harms themed messages.

3.4. Emotional valence.

Emotional valence responses were recorded with FACET. For positive valence, users experienced greater positive emotional response to the driving themed messages than non-users (F(1, 65) = 4.49, p=.039, ηp = 0.087; Fig. 3). Positive valence patterns for cognitive ability and health harms showed no consistent patterns. Nor did negative valence show a consistent pattern of emotional response for user status, message themes, nor campaigns.

Fig. 3.

Fig. 3.

Open in a new tab

Positive valence evidence scores as a function of user status and campaign for the driving themed messages.

3.5. Perceived effectiveness.

Table 1 shows the mean responses for perceived effectiveness, intention to avoid/quit marijuana, attitude toward the message, and message acceptance. As expected, non-users responded more positively (i.e., higher perceived effectiveness F(1, 48) = 21.19, p<.001, ηp2 = .31; higher intention to avoid marijuana F(1, 48) = 49.48, p<.001, ηp2 = .51; more positive attitude toward the message F(1, 48) = 7.69, p=.008, ηp2 = .14; and higher message acceptance F(1, 48) = 24.33, p<.001, ηp2 = .34) than users. Otherwise, there was little consistency in responses, particularly for users, across message themes or campaigns. For example, users showed higher perceived effectiveness for the health harms messages in the Do the Math campaign than they did for messages in the Spread the Facts campaign. Further, self-report measures showed little differentiation across campaigns and message themes, as evidenced by users showing no differences between the two campaigns for the cognitive ability and driving messages. Users also showed no differences for intention to avoid/quit marijuana across either of the two campaigns or the three message themes. The same is true for attitude toward the message and message acceptance. Non-users showed no differences for the self-report measures across both the two campaigns and the three message themes.

Table 1.

Self-report means (SDs) as a function of user status, message theme, and campaign.

Perceived Effectiveness (1 strongly disagree—5 strongly agree)
Non-user Non-user User User
Spread the Facts Do the Math Spread the Facts Do the Math
Cognitive Ability 2.89a
(1.00)		 2.40ab
(0.92)		 1.77bc
(0.84)		 1.61c
(0.66)
Driving 3.24a
(1.24)		 2.63ab
(1.07)		 2.12bc
(1.05)		 1.72c
(0.79)
Health Harms 2.63a
(1.07)		 2.67a
(1.18)		 1.61b
(0.66)		 1.91ab
(0.77)
Intention to Avoid Marijuana (0 won’t avoid/quit—10 will avoid/quit)
Non-user Non-user User User
Spread the Facts Do the Math Spread the Facts Do the Math
Cognitive Ability 7.84a
(2.95)		 8.01a
(2.95)		 2.54b
(2.39)		 2.48b
(2.32)
Driving 7.93a
(2.98)		 7.97a
(2.93)		 2.57b
(2.31)		 2.65b
(2.42)
Health Harms 7.76a
(2.96)		 7.75a
(2.93)		 2.54b
(2.42)		 2.71b
(2.46)
Attitude toward the Message (1 negative—7 positive)
Non-user Non-user User User
Spread the Facts Do the Math Spread the Facts Do the Math
Cognitive Ability 3.54a
(1.14)		 3.08ab
(1.16)		 2.80abc
(1.31)		 2.19c
(1.05)
Driving 3.58a
(1.42)		 3.17abc
(1.38)		 2.86abc
(1.33)		 2.44c
(1.25)
Health Harms 3.16abc
(1.32)		 3.59ac
(1.51)		 2.30bd
(1.34)		 2.72cd
(1.36)
Message Acceptance (1 strongly disagree—5 strongly agree)
Non-user Non-user User User
Spread the Facts Do the Math Spread the Facts Do the Math
Cognitive Ability 2.48a
(1.19)		 2.44a
(1.17)		 1.51b
(0.88)		 1.24b
(0.40)
Driving 3.05a
(1.43)		 2.51ab
(1.29)		 1.73bc
(1.01)		 1.40c
(0.57)
Health Harms 2.53a
(1.32)		 2.49a
(1.39)		 1.39b
(0.70)		 1.61b
(0.61)
Open in a new tab

Note. Means (SDs). Means within a row that do not share a subscript are significantly different at p < .05. Response scales for perceived effectiveness and message acceptance is 1-5; for attitude toward the message is 1-7; for intention to use marijuana is 0-10.

4. Discussion

This study is the first to prospectively compare real-time cognitive and affective responses to marijuana prevention messaging on perceived message effectiveness, message acceptance, attitudes toward the message, and intentions to use marijuana. Results showed that the driving messages in both campaigns evoked the greatest cognitive resource allocation to encoding message content (demonstrated by heart rate deceleration), the highest sympathetic nervous system activation (arousal), and the most positive emotional response (FACET) among users. The combination of these results indicates that users both cognitively and emotionally processed the driving theme messages better than the cognitive ability and health harms messages. This pattern of results is independent of the particular campaign the messages were from. It is possible that driving-themed messages are more salient to this age group because they drive daily, whereas concerns about cognitive decline may seem more distal. However, this possibility would likely not explain the differences shown between users and non-users on cognitive and emotional processing. The self-report results are less conclusive, particularly in the lack of discriminability across campaigns and message themes. While results did consistently show that non-marijuana users responded more positively than current marijuana users overall, the reliance solely on self-report measures in this study would yield the conclusion that neither campaign nor message theme mattered within user status.

The results from this study demonstrate the utility of psychophysiological measures in effectively designing and testing anti-marijuana messages. Specifically, we see that self-report alone does not give researchers a precise picture of the processes at work, nor were self-reports able to distinguish among campaigns and themes. The lack of significant differences across themes is particularly noteworthy, as message development often takes several iterative rounds of carefully reviewed concept and pilot testing before a campaign is deployed. These efforts, based on psychophysiological measures, appeared to show appreciable impact, suggesting that messages are not resonating in a way that is currently captured by traditional paper-pencil measures. In contrast, from the current study, we may conclude from the psychophysiological data that the messages about driving were most effective, particularly with respect to cognitive resource allocation and emotional response. We could make no such inference from the self-report data.

There are several limitations worth noting. First, orienting response, visual attention, and arousal are all sensitive to acute manipulations within messages, such as appetitive cues (e.g., cigarette smoking for smokers) and aversive cues (e.g., graphic images of the consequences of smoking; (Clayton, Leshner, Tomko et al., 2017; Sanders- Jackson et al., 2011). It is possible that different message strategies impact cognitive resource allocation, emotional response, appeal, and receptivity differently than they did in this study. Therefore, conclusions about the effectiveness of message themes and campaigns are limited to the particular themes and campaigns used in this study.

Second, the two campaigns varied in a number of important ways that may influence the cognitive and emotional responses to each. For example, the Spread the Facts messages used photographs of real images, black and green for the primary colors, and large typeface for the risks. The Do the Math messages used a muted blue for the primary color and pictographs to connote the primary risk, and small text that added information about the risk. Therefore, the two campaigns, one could argue, vary in message complexity. It appears the Spread the Facts message were more complex than the Do the Math messages in that the Spread the Facts messages contained more elements (e.g., text, pictures, colors, sponsor logos). In terms of content, the cognitive ability and health harms themed messages in the Spread the Fact campaigns appear to be targeted to individuals under 18-years of age, whereas the driving-themed Spread the Facts message and all of the Do the Math messages do not. However, these discrepancies across the two campaigns do not account for the results that the driving messages in both campaigns evoked the greatest cognitive resource allocation to encoding message content, the highest sympathetic nervous system activation, and the most positive emotional response (FACET) among users.

Third, the messages to which participants were exposed were focused on combusted marijuana use but did not address harms associated with the variety of products on the legal market. Understanding youth appeal and responsivity to messages that specifically target this nascent group of new or potential future users will be important for future research.

5. Conclusion

Findings from this study support the application of psychophysiological measures in formative work to improve the development and effectiveness of marijuana health-risk prevention campaigns. When developing large-scale public health campaigns, officials tend to rely on paper-pencil self-report assessments; however our study findings show that self-reports may not capture the dynamics of cognitive and emotional responses to these messages. Current public education campaigns appear to have limited effectiveness at changing misconceptions regarding the safety and consequences of marijuana use, as evidenced by the rapid increase in marijuana use among younger individuals in the US during the past decade (Compton et al., 2016). Our findings highlight that marijuana users cognitively and emotionally processed the driving theme messages better than the cognitive ability and health harms messages, across both campaigns. With greater liberalization of marijuana in the U.S., effective public health messaging will be needed to minimize the harms of marijuana use in young people. Increasing the effectiveness of anti-marijuana message development, perhaps through the application of new ways of measuring messaging responsiveness, could help determine what types of messages best engage cognitive and emotional systems, while also most effectively deterring future use.

Highlights:

  • psychophysiological assessment of the impact of marijuana public health messaging

  • driving-themed messages showed the greatest cognitive resource allocation to encoding the message, the greatest arousal, and the most positive emotional response

  • marijuana user status did not predict psychophysiological responses

  • self-reports showed a less consistent pattern.

Role of Funding Source:

GL was partially supported by the Edward L. & Thelma Endowed Chair in Journalism. AMC was partially supported by the Oklahoma Tobacco Settlement Endowment Trust and R21DA041548-01. EMS was supported by K99DA046563. ACV was supported by the Centers of Biomedical Research Excellence P20GM103644 award from the National Institute of General Medical Sciences and National Institute in Drug Abuse of the National Institutes of Health under Award Number R21DA051943.

Appendix—

Experimental Stimuli

Spread the Facts campaign Do the Math campaign
Cognitive Ability Cognitive Ability
graphic file with name nihms-1709423-t0004.jpg graphic file with name nihms-1709423-t0005.jpg
Driving Impairment Driving Impairment
graphic file with name nihms-1709423-t0006.jpg graphic file with name nihms-1709423-t0007.jpg
General Health Harms General Health Harms
graphic file with name nihms-1709423-t0008.jpg graphic file with name nihms-1709423-t0009.jpg
Open in a new tab

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Declaration of Competing Interest: No conflict declared.

All authors have read and agreed to the published version of the manuscript.

References

  1. Allara E, Ferri M, Bo A, Gasparrini A, Faggiano F 2015. Are mass-media campaigns effective in preventing drug use? A Cochrane systematic review and meta-analysis. BMJ Open 5(9), e007449. 10.1136/bmjopen-2014-007449 [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Alvaro EM, Crano WD, Siegel JT, Hohman Z, Johnson I, Nakawaki B 2013. Adolescents’ attitudes toward antimarijuana ads, usage intentions, and actual marijuana usage. Psychol Addict Behav, 27(4), 1027–1035. 10.1037/a0031960 [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Azofeifa A, Mattson ME, Schauer G, McAfee T, Grant A, Lyerla R 2016. National Estimates of Marijuana Use and Related Indicators — National Survey on Drug Use and Health, United States, 2002–2014. MMWR Surveill Summ, 65 (No. SS-11), 1–25. DOI: 10.15585/mmwr.ss6511a1externalicon. [DOI] [PubMed] [Google Scholar]
  4. Boudreaux ED, Sullivan A, Abar B et al. 2012. Motivation rulers for smoking cessation: a prospective observational examination of construct and predictive validity. Addict Sci Clin Pract 7(8). 10.1186/1940-0640-7-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Cacioppo JT, Gardner WL 1999. Emotion. Annu Rev Psychol, 50(1), 191–214. 10.1146.annurev/psych.50.1.191. [DOI] [PubMed] [Google Scholar]
  6. Callaghan RC, Allebeck P, Sidorchuk AJ 2013. Marijuana use and risk of lung cancer: a 40-year cohort study. Cancer Causes Control, 24(10), 1811–1820. DOI: 10.1007/s10552-013-0259-0. [DOI] [PubMed] [Google Scholar]
  7. Centers for Disease Control and Prevention. 2001. Message strategy research to support development of the youth media campaign (YMC): revealing target audience receptiveness to potential YMC message concepts. Retrieved from https://www.cdc.gov/youthcampaign/research/PDF/MessageStrategy.pdf
  8. Centers for Disease Control and Prevention. 2016. Impact of first federally funded anti-smoking ad campaign remains strong after three years. Retrieved from https://www.cdc.gov/media/releases/2016/p0324-anti-smoking.html
  9. Cermak TL, Banys P 2016. Prevention of youthful marijuana use. J Psychoactive Drugs, 48(1), 21–23, DOI: 10.1080/02791072.2015.1117689 [DOI] [PubMed] [Google Scholar]
  10. Clayton RB, Leshner G, Bolls PD, Thorson E 2017. Discard the smoking cues—Keep the disgust: An investigation of tobacco smokers’ motivated processing of anti-tobacco commercials. Health Commun, 32(11), 1319–1330. DOI: 10.1080/10410236.2016.1220042 [DOI] [PubMed] [Google Scholar]
  11. Clayton RB, Leshner G, Sanders-Jackson A, Hendrickse J 2020. When counterarguing becomes the primary task: Examination of dogmatic anti-vaping messages on psychological reactance, available cognitive resources, and memory. J Commun, 70(4), 522–547. 10.1093/joc/jqaa010. [DOI] [Google Scholar]
  12. Clayton RB, Leshner G, Tomko RL, Trull TJ, Piasecki TM 2017. Countering craving with disgust images: Examining nicotine withdrawn smokers’ motivated message processing of anti-tobacco public service announcements. J Health Commun, 22(3), 254–261. 10.1080/10810730.2016.1268222 [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Compton WM, Han B, Jones CM, Blanco C, Hughes A 2016. Marijuana use and use disorders in adults in the USA, 2002–14: analysis of annual cross-sectional surveys. Lancet Psychiatry, 3(10), 954–964. doi: 10.1016/S2215-0366(16)30208-5. [DOI] [PubMed] [Google Scholar]
  14. Crano WD, Alvaro EM, Tan CN, & Siegel JT 2017. Social mediation of persuasive media in adolescent substance prevention. Psychol Addict Behav, 31(4), 479–487. 10.1037/adb0000265 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Danseco ER, Kingery PM, Coggeshall MB 1999. Perceived risk of harm from marijuana use among youth in the USA. Sch Psycho Int, 20(1), 39–56. 10.1177/0143034399201004 [DOI] [Google Scholar]
  16. Davis KC, Nonnemaker J, Duke J, & Farrelly MC (2013). Perceived effectiveness of cessation advertisements: The importance of audience reactions and practical implications for media campaign planning. Health Commun, 28(5), 461–472. 10.1080/10410236.2012.696535 [DOI] [PubMed] [Google Scholar]
  17. Dillard JP, Shen L 2005. On the nature of reactance and its role in persuasive health communication. Commun Monogr, 72(2), 144–168. 10.1080/03637750500111815 [DOI] [Google Scholar]
  18. Dillard JP, Weber KM, Vail RG 2007. The relationship between the perceived and actual effectiveness of persuasive messages: A meta-analysis with implications for formative campaign research. J Commun, 57(4), 613–631. 10.1111/j.1460-2466.2007.00360.x [DOI] [Google Scholar]
  19. Ferri M, Allara E, Bo A, Gasparrini A, Faggiano F 2013. Media campaigns for the prevention of illicit drug use in young people. Cochrane Database of Syst Rev, 6. doi: 10.1002/14651858.CD009287. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Gates S 2006. Interventions delivered to young people in non-school settings for the prevention of drug use. Cochrane Database of Syst Rev, 1. doi: 10.1002/14651858.CD005030. [DOI] [PubMed] [Google Scholar]
  21. Hasin DS 2018. US epidemiology of cannabis use andassociated problems. Neuropsychopharmacology, 43(1), 195–212. doi: 10.1038/npp.2017.198 [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Hornik R, Jacobsohn L, Orwin R, Piesse A, Kalton G 2008. Effects of the national youth anti-drug media campaign on youths. Am J Public Health, 98(12), 2229–2236. 10.2105/AJPH.2007.125849 [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Huang YHJ, Zhang ZF, Tashkin DP,Feng B, Straif K,Hashibe M 2015. An epidemiologic review of marijuana and cancer: An update. Cancer Epidemiol Biomarkers & Prev, 24(1), 15–31. doi: 10.1158/1055-9965.EPI-14-1026. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Hubert W, de Jong-Meyer R 1990. Psychophysiological response patterns to positive and negative film stimuli. Bio Psychol, 31(1), 73–93. DOI: 10.1016/0301-0511(90)90079-c [DOI] [PubMed] [Google Scholar]
  25. iMotions Biometric Research Platform v. 7.1 (2018). iMotions A/S. Copenhagen, Denmark. [Google Scholar]
  26. Lang A 1990. Involuntary attention and physiological arousal evoked by structural features and emotional content in TV commercials. Communic Res, 17(3), 275–299. 10.1177/009365090017003001 [DOI] [Google Scholar]
  27. Lang A 1994. What can the heart tell us about thinking? In Lang A (Ed.) Measuring Psychological Responses to Media Messages (Vol. II, pp. 99–111). Hillsdale, NJ: Erlbaum. [Google Scholar]
  28. Lang A 2006. Using the limited capacity model of motivated mediated message processing to design effective cancer communication messages. J Commun, 56(s1), S57–S80. 10.1111/j.1460-2466.2006.00283.x [DOI] [Google Scholar]
  29. Lang A 2009. The limited capacity model of motivated mediated message processing. In Nabi R & Oliver M (Eds.), The SAGE Handbook of Media Processes and Effects (pp. 193–204). Thousand Oaks, CA: Sage. [Google Scholar]
  30. Lang PJ, Bradley MM 2008. Appetitive and defensive motivation in the substrate of emotion. In Elliot AJ (Ed.) Handbook of Approach and avoidance motivation (pp. 51–66). New York: Psychology Press,. [Google Scholar]
  31. Lang PJ, Bradley MM, Cuthbert BN 1997. Motivated attention: Affect, activation, and action. In Lang PJ, Simons RF, & Balaban MT (Eds.) Attention and Orienting: Sensory and Motivational Processes (pp. 97–135). Hove, UK: Psychology Press. [Google Scholar]
  32. Larsen JT, Berntson GG, Poehlmann KM, Ito TA, Cacioppo JT 2008. The psychophysiology of emotion. Handbook of Emotions, 3, 180–195. [Google Scholar]
  33. LaVoie NR, Quick BL, Riles JM, Lambert NJ 2015. Are graphic cigarette warning labels an effective message strategy? A test of psychological reactance theory and source appraisal. Communic Res, 44, 416–436. doi: 10.1177/0093650215609669 [DOI] [Google Scholar]
  34. Leshner G 2014. The basics of experimental research in media studies. In Darling-Wolf F (Ed.) Blackwell’s International Companion to Media Studies: Research Methods in Media Studies (pp. 236–254). Wiley-Blackwell. [Google Scholar]
  35. Leshner G, Bolls P, Gardner E, Moore J, Kreuter M 2018. Breast cancer survivor testimonies: Effects of narrative and emotional valence on affect and cognition. Cogent Soc Sci, 4(1), 1426281. DOI: 10.1080/23311886.2018.1426281 [DOI] [Google Scholar]
  36. Leshner G, Bolls P, Thomas E 2009. Scare ‘em or disgust ‘em: The effects of graphic health promotion messages. Health Commun, 24, 447–458. doi: 10.1080/10410230903023493 [DOI] [PubMed] [Google Scholar]
  37. Leshner G, Bolls P, Wise K 2011. Motivated processing of fear appeal and disgust images in televised anti-tobacco ads. J Media Psychol, 23(2), 77–89. 10.1027/1864-1105/a000037 [DOI] [Google Scholar]
  38. Lipari R, Ahrnsbrak R 2017. Risk and protective factors and estimates of substance use initiation: Results from the 2016 National Survey of Drug Use and Health. Retrieved from https://www.samhsa.gov/data/sites/default/files/NSDUH-DR-FFR3-2016/NSDUH-DR-FFR3-2016.htm [PubMed]
  39. Lipari RN, Hedden SL, Hughes A 2014. Substance use and mental health estimates from the 2013 National Survey on Drug Use and Health: Overview of findings. In: The CBHSQ Report. Rockville (MD): Substance Abuse and Mental Health Services Administration (US). Available from: https://www.ncbi.nlm.nih.gov/books/NBK385055/ [PubMed] [Google Scholar]
  40. Medical Marijuana, Inc. 2016. Nearly 60 percent of US population now lives in states with marijuana legalization. Retrieved from https://news.medicalmarijuanainc.com/nearly-60-percent-u-s-population-now-lives-states-marijuana-legalization/
  41. National Academies of Sciences, Engineering, and Medicine. 2017. The health effects of cannabis and cannabinoids: The current state of evidence and recommendations for research. Washington (DC):National Academies Press. Available from: https://www.ncbi.nlm.nih.gov/books/NBK423845/ doi: 10.17226/24625 [DOI] [PubMed] [Google Scholar]
  42. Nonnemaker JM, Silber-Ashley O, Farrelly MC, Dench D 2012. Parent-child communication and marijuana initiation: Evidence using discrete-time survival analysis. Addict Behav, 37(12), 1342–1348. 10.1016/j.addbeh.2012.07.006 [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Pew Research Center. 2018. About six-in-ten Americans support marijuana legalization. Retrieved from http://www.pewresearch.org/fact-tank/2018/01/05/americans-support-marijuana-legalization/
  44. Potter RF, Bolls P 2012. Psychophysiological Measurement and Meaning: Cognitive and Emotional Processing of Media. New York: Routledge. [Google Scholar]
  45. Ravaja N 2004. Contributions of psychophysiology to media research: Review and recommendations. Media Psychol, 6(2), 193–235. 10.1207/s1532785xmep0602_4 [DOI] [Google Scholar]
  46. Reeves B, Geiger S 1994. Designing experiments that assess psychological responses. In Lang A (Ed.), Measuring Psychological Responses to Media (pp. 165–180). Hillsdale, NJ: LEA. [Google Scholar]
  47. Rice RE, Atkin CK 2012. Public Communication Campaigns. Thousand Oaks, CA: Sage. [Google Scholar]
  48. SAMHSA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2002-2019. Retrieved from: https://www.samhsa.gov/data/sites/default/files/reports/rpt29394/NSDUHDetailedTabs2019/NSDUHDetTabsSect7pe2019.htm [Google Scholar]
  49. Sanders- Jackson AN, Cappella JN,Linebarger DL, Piotrowski JT, O'Keeffe M, Strasser AA 2011. Visual attention to antismoking PSAs: Smoking cues versus other attention- grabbing features. Hum Commun Res, 37(2), 275–292. 10.1111/j.1468-2958.2011.01410.x [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Stevens EM, Cohn AM, Villanti AC, Leshner G, Wedel A, Wagener TL 2019. Perceived effectiveness of anti-marijuana messages in adult users and nonusers: an examination of responses to messages about marijuana’s effects on cognitive performance, driving, and health. J Stud Alcohol Drugs, 80(4), 415–422. DOI: 10.15288/jsad.2019.80.415 [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Stevens EM, Kim S, Leshner G, Leavens EL, Vidrine JI, Frank-Pearce SG, … & Wagener TL 2019. The attitudinal and motivational effects of anti-waterpipe messages. Am J Health Behav, 43(3), 478–489. DOI: 10.5993/AJHB.43.3.3 [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Wang Z, Lang A 2006. Ad placement matters: A psychophysiological examination of program context effects on advertising processing. Psychophysiology, 43, S104. [Google Scholar]
  53. Wolburg JM 2006. College students’ responses to antismoking messages: Denial, defiance, and other boomerang effects. J Consum Aff, 40(2), 294–323. 10.1111/j.1745-6606.2006.00059.x [DOI] [Google Scholar]
  54. Zimmerman R, Cupp PK, Abadi M, Donohew RL, Gray C, Gordon L, Grossl AB 2014. The effects of framing and fear on ratings and impact of antimarijuana PSAs. Subst Use Misuse, 49(7), 824–835. 10.3109/10826084.2014.880721 [DOI] [PubMed] [Google Scholar]

RESOURCES