Impact of Anti-drug Use Advertisements in Hong Kong: A Randomized Control Experiment

Abstract

Background: This study examined how anti-drug use advertisements, along with perceptions of their realism and threat to personal freedom, impact perceptions toward drug use among Chinese university students in Hong Kong. Despite the widespread use of such ads globally, their potential for a “boomerang effect” has been understudied outside Western contexts. Methods: 483 students (aged 18–25) were randomly assigned to view either anti-drug or unrelated advertisements. Pre- and post-intervention surveys assessed their perceptions. Data were analyzed using two-tailed t-tests, a two-way mixed ANOVA, and ordinary least squares regression models. Results: No boomerang effect was found; exposure to anti-drug ads had limited influence on drug use perceptions. Unrealistic ads were associated with more favorable perceptions of drug use, while perceived threat to personal freedom did not significantly impact perceptions. Conclusion: Cultural context may influence the boomerang effect. Anti-drug campaigns should use coherent, realistic narratives to effectively promote negative drug views.

Keywords

anti-drug use advertisements perceived realism perceived threat on freedom psychological reactance Hong Kong

In 2022, over 292 million individuals worldwide, aged 15–64 years, were estimated to have used drugs (United Nations Office on Drugs and Crime, 2024). Notably, young people between 18 and 25 years old, including university students, exhibit the highest likelihood of drug use (Boden & Day, 2023; Shek et al., 2020; Wong et al., 2020), posing potential long-term negative effects on their physical and mental well-being (Kingston et al., 2017). Advertisements play a vital role in global anti-drug use policies by promoting anti-drug use perceptions (Palmgreen et al., 2001).

Hong Kong, one of the most densely populated cities in the world, actively employs anti-drug use advertising as a key component of its drug prevention strategy, utilizing a wide array of channels including news reports, banners, posters, flyers, and digital platforms like YouTube (Narcotics Division, 2019). For this paper, “drugs” refers to substances classified as “dangerous drugs” according to the Dangerous Drugs Ordinance (2019), such as heroin, methamphetamine, cannabidiol, and others listed in the ordinance.

Since 2010, the Hong Kong government has committed over three billion Hong Kong dollars to the Beat Drug Fund (Audit Commission, 2017). Despite these efforts, there was a 21.5% increase in newly reported young people using drugs (aged 21–35 years) from 511 in 2019 to 621 in 2022 (Narcotic Division, 2024). This trend raises questions about the effectiveness of the government’s substantial investment in wide-reaching anti-drug use advertisements.

Studies in American and European contexts have revealed that anti-drug use advertisements sometimes fail to achieve their intended objectives; in some cases, they even produce a boomerang effect, promoting perceptions contrary to their messaging (Clayton, 2022; Han & Kim, 2019; Slavin & Earleywine, 2019). Psychological reactance theory and perceived realism theory explain this effect. Psychological reactance theory posits that perceived threats to freedom of choice evoke counterarguments and a desire to reclaim autonomy, resulting in a boomerang effect (Brehm, 1966; Clayton, 2022). Perceived realism theory posits that unrealistic messages are more likely to be distrusted (Cho et al., 2013; Ewanation et al., 2017).

Despite extensive research in Western contexts, studies exploring this dynamic in non-Western settings are scarce. Hong Kong, with its unique blend of Western influence and Eastern culture, presents an intriguing case for such investigations. Understanding how individuals in this setting respond to anti-drug use messages can inform more effective drug prevention policies, with tailored interventions addressing the specific needs and cultural sensitivities of the population.

This study employed a randomized controlled trial (RCT) in Hong Kong, integrating psychological reactance theory and perceived realism theory to investigate the impact of anti-drug use advertisements on individuals’ perceptions of drug use. We addressed the following research question: How do anti-drug use advertisements affect individuals’ perceptions toward drugs? Additionally, we provide recommendations for designing more effective anti-drug use advertisements and discuss the limitations of our study.

Literature Review

Anti-Drug Use Advertisements and Boomerang Effect

Anti-drug use advertisements have been employed globally to promote anti-drug use messages, with the primary aim of reducing and preventing drug use. This goal aligns with Jowett and O’Donnell’s (2018) definition of anti-drug use advertisements, which are regarded as any messages via media—including newspapers, movies, and materials used in schools—that aim to prevent or reduce drug use behaviors, or the biological, psychological, and legal harm caused by these behaviors. Research on the effectiveness of anti-drug use advertisements and campaigns in Anglophone countries shows mixed results: some advertisements reduce audiences’ drug use (Evans et al., 2020; Palmgreen et al., 2001), whereas others have no impact (Buller et al., 2019; Erceg-Hurn, 2008; Hornik et al., 2008; Magura, 2012).

Several studies have presented alarming results, suggesting that some advertisements increase pro-drug use perceptions (Czyzewska & Ginsburg, 2007; Fishbein et al., 2002; Ginsburg & Czyzewska, 2005; Hornik et al., 2008; Jacobsohn, 2007; Slavin & Earleywine, 2019; Terry-McElrath et al., 2011; Zhao & Fink, 2021). This phenomenon is called a boomerang effect, which occurs when media content generates a perception or behavior opposite to what was originally intended (Cannon & Byrne, 2019). Two theories have been used to explain the boomerang effect: psychological reactance theory and perceived realism.

Psychological Reactance Theory

Psychological reactance theory suggests that when individuals feel that their personal freedom to choose is under threat, negative emotions and cognitions, such as anger and counterarguments, emerge. Individuals are then motivated to act to restore their freedom (Brehm, 1966; Clayton, 2022). The basic assumption is that individuals value personal freedom. When they perceive their personal freedom to be threatened, they take action to reassert that freedom. For example, when reading a message like “you must not take drugs,” audiences may take action to reassert their freedom by developing a pro-drug use perception or even by using illicit substances (Crano et al., 2017; Henriksen et al., 2006).

Studies in health promotion, including anti-drug use advertisements, have found that the more audiences perceive their freedom as being threatened by the tone of an advertisement, the more likely it is that the advertisement will negatively impact their perception and behavior by driving freedom-restoration behavior, and the more likely a boomerang effect would occur (Armstrong et al., 2021; Crano et al., 2017; LaVoie et al., 2017; Massey et al., 2022; Richards et al., 2022; Slavin & Earleywine, 2019; Zhang, 2020). For example, in an experimental study, Slavin and Earleywine (2019) revealed that abstinence messages that instruct readers to completely stop using marijuana (imposing higher restrictions on freedom) resulted in higher levels of psychological reactance and a boomerang effect (increased craving for marijuana). When people are warned against health-threatening behaviors like drug use, they feel that their freedom is threatened and respond by generating more pro-drug perceptions or by taking even more drugs to restore their perceived loss of freedom (Hornik et al., 2008; Slavin, 2018; Zhao & Fink, 2021).

Interestingly, studies have also documented cultural variations in people’s reactions to freedom restrictions and psychological reactance, suggesting that Asian and Chinese audiences react more positively to freedom restrictions (Kim et al., 2017; Ng et al., 2021; Savani et al., 2008). One possible explanation for such findings is that collectivist cultures place less emphasis on individual freedom, which can lead to less negative consequences for individuals who have grown up in such environments compared to those who have grown up in individualistic cultures (Bang et al., 2021; Graupmann et al., 2012; Kim et al., 2017; Quick & Kim, 2009). For example, Kim et al. (2017), in their study on environmental advertising, discovered that, compared with South Korean participants, American participants experienced more psychological reactance towards assertive messages promoting recycling.

Perceived Realism

Another factor that affects the efficacy of anti-drug use advertisements is the audience’s perceived realism of the advertisement content. In communication studies, perceived realism refers to an audience’s perception of whether a narrative in the media accurately reflects real-life situations (Cao & Xu, 2021; Cho et al., 2013; Ewanation et al., 2017). Stories and characters portrayed in health promotion advertisements are often fictional (Busselle & Bilandzic, 2008; Van Leeuwen et al., 2013). As such, perceived realism involves determining whether a narrative can feasibly occur in the real world or accurately represent common real-world events. Communication studies have demonstrated that when audiences perceive media content as being more realistic, they are more inclined to accept and are positively influenced by the message conveyed (Cao & Xu, 2021; Czyzewska & Ginsburg, 2007; Fishbein et al., 2002; Terry-McElrath et al., 2011).

Research suggests that perceiving an anti-drug use message as unrealistic is associated with a higher likelihood of the boomerang effect (Czyzewska & Ginsburg, 2007; Terry-McElrath et al., 2011). When individuals perceive inconsistencies between the message content and their existing knowledge, they tend to react negatively to the message. For example, Terry-McElrath et al. (2011) found that exposure to anti-marijuana public service announcements increased young people’s intention to use illicit substances, possibly because the messages did not align with their prior experiences. Similarly, Cho et al. (2013), in a study involving 315 American college students, found that participants were more likely to distrust alcohol and illicit substance-related messages if they perceived the depicted events as unrealistic.

Current Study

Hong Kong has adopted an “enlightened prohibition” approach in its drug policy, focusing on law enforcement, control-based treatment, and prevention (Cheng & Lapto, 2023; Cheung & Cheung, 2018; Kong & Laidler, 2020). Anti-drug use campaigns have increased in Hong Kong since the 1970s (Cheung & Cheung, 2018). More recently, the Hong Kong government’s Narcotics Division has produced and distributed anti-drug use advertisements across various platforms, including news reports, banners, posters, flyers, and YouTube.

These advertisements aim to promote anti-drug use messages by fostering two subsets of perceptions supporting drug use prevention and rehabilitation. The first aims to discourage drug use by highlighting the negative consequences of drugs and advocating that individuals, particularly young people, avoid drug use. Examples of these messages include phrases such as “Simply say no, drugs are addictive. No thanks!,” “Cocaine and ice are addictive,” and “Ice can dissolve your brain.” The second subset encourages proactive attitudes towards seeking help. This approach aims to facilitate the early identification of the hidden population with SUD and promote prompt assistance among those struggling with drug use. Examples of these messages include prompts to call a helpline at “186 186” or send a message via WhatsApp or WeChat to “98 186 186” for assistance.

Two types of anti-drug use messages are used. The first type emphasizes the potential threat to personal freedom, utilizing statements like “Don’t lose your life to drugs,” “Quit Drugs Now!” and “You must know how to resist temptation.” The second type presents a more positive perspective, conveying messages such as “Comfortable, satisfied, free… There are many ways to enjoy life.” While message types may overlap, the perceived threat to freedom and perceived realism require further examination.

Additionally, it remains unclear whether psychological reactance also affects drug use perceptions in Hong Kong, where people are affected by both individualist and collectivist cultures (Tsoi & Liu, 2019) owing to its unique history as a former British colony and its status as an international metropolitan city influenced by Western liberal culture.

Thus, the current study aimed to investigate the effect of anti-drug use advertisements in Hong Kong by answering the following research question: How do anti-drug use advertisements affect individuals’ perceptions toward drugs?

Specifically, we tested several hypotheses. Despite the mixed results of previous studies, we hypothesized that anti-drug use advertisements function as educational tools that raise awareness regarding the negative consequences of drug use and alternative lifestyles. Despite its colonial history, Hong Kong continues to be influenced by traditional collectivistic Chinese values (Tsoi & Liu, 2019). Although this may seem contradictory to our other hypotheses, as previous studies conducted shows that psychological reactance is less common in Asian collectivistic contexts (Bang et al., 2021; Graupmann et al., 2012; Kim et al., 2017; Quick & Kim, 2009), we are optimistic about the main effect of anti-drug use advertisements while we also recognize that there may be some variability in the actual impact they have in different contexts or situations.

Exposure to anti-drug use advertisements is positively associated with a more unfavorable perception of drugs.

However, adopting the psychological reactance theory and a related framework on perceived realism, we hypothesized that both threats to freedom and perceived unrealism will negatively affect individuals’ acceptance of anti-drug use messages, and thus lead to more favorable perceptions of drugs.

A higher level of perceived threat to freedom from anti-drug use advertisements is associated with a more favorable perception of drugs.

A higher level of perceived unrealism from anti-drug use advertisements is associated with a more favorable perception of drugs.

Methods and Data Collection

Experimental Procedure

Data for this project were obtained from an RCT conducted in 2021. The participants were recruited from a teaching-oriented university in Hong Kong that offers over 90 undergraduate and 30 postgraduate programs to more than 20,000 students. Based on the student population, a confidence level of 95%, and a margin of error of 5%, the target sample size was estimated at 377. Recruitment posters were placed on student noticeboards and sent via email. Undergraduate students aged 18 to 25, who could read and speak Cantonese and were permanent residents of Hong Kong, were eligible to participate. The ability to read and speak Cantonese was essential for this study as Cantonese is often presented in anti-drug use advertisements in Hong Kong. In addition, the permanent residency requirement requires at least seven years of residency in Hong Kong, thus suggesting a similar level of exposure to anti-drug use messages among participants. We recruited 513 participants who provided written informed consent to participate in the study. Participants received HK$150 (approximately USD$19) for their participation.

We conducted a simple randomization procedure and assigned 257 and 256 participants to the experimental and control groups, respectively. An independent researcher used Microsoft Excel to generate the random assignment sequence, ensuring the predetermined order was followed.

Both groups completed a survey that collected demographic and baseline information on their perception of drugs. The experimental group was presented with anti-drug use advertisements produced by the Narcotics Division of the Security Bureau of Hong Kong (including one leaflet, two posters, and two TV announcements released in 2018 and 2020; Narcotics Division, 2019). The Narcotics Division is one of the major providers of anti-drug advertisements in Hong Kong. These four experimental materials were widely used throughout Hong Kong to promote anti-drug messages. They were featured in schools, on the streets, in television commercials, and on online platforms such as YouTube, Instagram, and Facebook. We believe these materials can most effectively represent anti-drug advertisements in Hong Kong during the time of our study.

The control group received materials from hygiene guides unrelated to anti-drug use information. After the manipulation, perception of drugs was measured again. We also presented and measured respondents’ perceived threat to freedom, and perceived unrealism. Both groups were subjected to similar treatment durations. All the procedures were conducted via Zoom in light of COVID-19 policy implemented in Hong Kong. This research complies with the ethics regulations of [University name anonymous for review].

Data

After collecting the data, we identified and excluded invalid cases using various methods. First, despite the age requirement at the recruitment stage, we identified 22 participants who self-reported an age below 18 or above 25 years. Second, we identified five careless responses by cross-checking reversed and follow-up questions on demographic information. Specifically, if the respondent chose yes to the question “I don’t know any drug user” and any of the follow-up questions, such as “I know a drug user (coworker),” then we identified it as a careless response. Third, we excluded one observation that reported a family size of zero. All invalid responses were removed from the sample. In addition, considering the potential impact of being a drug user in this study, we excluded two observations self-identified as drug users. The final sample comprised 483 participants, with 237 in the experimental group and 246 in the control group.

We further conducted t-tests and chi-square tests to perform balance checks (Tables 1 and 2). The results suggested that randomization was successful, and the data were balanced in the final sample.

Table 1.

Descriptive Statistics for Respondents’ Demographic Information.

Variables	Control (n = 246)		Experiment (n = 237)		Total (n = 483)
Variables	N	%	N	%	N	%
Gender χ² = 0.0001 V = 0.0003
Male	80	32.52	77	32.49	157	32.51
Female	166	67.48	160	67.51	326	67.49
Grade χ² = 3.5533 V = 0.0858
1	32	13.01	41	17.30	73	15.11
2	49	19.92	42	17.72	91	18.84
3	94	38.21	78	32.91	172	35.61
4	69	28.05	75	31.65	144	29.81
5	2	0.81	1	0.42	3	0.62
Work experience dummy χ² = 0.2440 V = 0.0225
No	36	14.63	31	13.08	67	13.87
Yes	210	85.37	206	86.92	416	86.13
Religion χ² = 5.9548 V = 0.111
No religion	200	81.30	194	81.86	394	81.57
Buddhism	3	1.22	8	3.38	11	2.28
Taoism	1	0.41	1	0.42	2	0.41
Christian	39	15.85	29	12.24	68	14.08
Catholic	2	0.81	5	2.11	7	1.45
Islam	0	0.00	0	0.00	0	0.00
Hinduism	1	0.41	0	0.00	1	0.21
Household income χ² = 1.5232 V = 0.0562
<=5000	20	8.13	15	6.33	35	7.25
5,001–10,000	18	7.32	19	8.02	37	7.66
10,001–30,000	121	49.19	127	53.59	248	51.35
30,001–60,000	72	29.27	64	27.00	136	28.16
>=60001	15	6.10	12	5.06	27	5.59
TV time per day χ² = 7.2525 V = 0.1225
No TV time	38	15.45	38	16.03	76	15.73
<=1 hour	113	45.93	118	49.79	231	47.83
1–2 hours	60	24.39	50	21.10	111	22.77
2–3 hours	18	7.32	25	10.55	43	8.90
>=3 hours	17	6.91	6	2.53	23	4.76
Know drug user χ² = 0.1965 V = 0.0202
Don’t know	205	83.33	201	84.81	406	84.06
Know	41	16.67	36	15.19	77	15.94

Note. χ² = Chi-square test V = Cramer’s V.

Table 2.

Two-tailed t test on Demographics and Key Measures by Group Assignment.

Variables	Control (n = 246)		Experiment (n = 237)		Difference		t test (2-tailed)
Variables	Mean	SD	Mean	SD	Mean	SE	t	p
Age [18, 25]	21.341	1.682	21.177	1.800	0.164	0.158	1.037	0.301
Work experience [0, 9]	2.277	1.728	2.255	1.733	0.022	0.158	0.142	0.887
Family size [2, 8]	3.740	1.021	3.873	1.017	−0.134	0.093	−1.440	0.151
Siblings [0, 6]	1.520	1.163	1.624	1.073	−0.104	0.102	−1.022	0.307
Threat of freedom [1, 4.75]	2.522	0.785	2.543	0.767	−0.021	0.070	−0.295	0.768
Perceived unrealism [1, 3.5]	2.025	0.492	1.883	0.436	0.142	0.042	3.361	0.001
Pre-perception of drugs [2.33, 5]	4.319	0.442	4.338	0.453	−0.019	0.041	−0.478	0.633
Post-perception of drugs [2.75, 5]^a	4.271	0.454	4.352	0.483	−0.080	0.043	−1.883	0.060
Change in perception of drugs [-1.17, 0.92]	−0.047	0.278	0.013	0.251	−0.061	0.024	−2.522	0.012

^aOne-tailed t test: Post-Perception of Drugs (control < experiment) p = .030. SD = Standard deviation. SE = Standard error.

Measures

Perception of Drugs

Perception of drugs was based on two measures. We first adopted the general drug use subscale from the Drug Attitude Scale (Goodstadt et al., 1978), which is a six-item Likert scale that measures attitudes toward drug use. The second subscale, Anti-Drug Message Agreement, adopted and developed from the Public Opinion Survey (Narcotics Division, 2019), was utilized to determine participants’ agreement levels with six core messages in anti-drug use advertisements. These messages include those that focus on the negative impacts of drug use (e.g., “Drugs such as ‘Ice’ and cannabis can be addictive and harmful to your body”) and those that encourage a proactive perception towards seeking help (e.g., “Do not battle against drugs alone, get help and quit drugs now”). The original survey was designed to measure whether the participants were aware of the six messages. We replaced the questions with whether they agree with these messages for the purpose of this study. We adopted this subscale because of its direct measures that connect to the anti-drug use advertisements adopted in the experiments. In addition, it reflects the goals outlined by Hong Kong’s Narcotics Division.

Participants were asked to rate each item on a five-point Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree) before and after the manipulation. Both subscales reached acceptable internal consistency for both times (Drug Attitude: α_pre = 0.729, α_post = 0.714; Anti-Drug Message Agreement: α_pre = 0.794; α_post = 0.822).

Combining the Drug Attitude and Anti-Drug Message Agreement scales, we conducted an exploratory factor analysis (EFA). The EFA results confirmed that the 12 items belonged to a single factor (eigenvalue = 3.860, proportion = 0.914). The new factor obtained a good construct validity (Kaiser–Meyer–Olkin = 0.876, χ² = 1700.49, p < .001). We then obtained the mean score for the twelve items as the value of the Perception of Drugs scale, which ranged from 2.333 to 5 in the sample. A higher mean score indicates a more unfavorable perception of drugs.

We then calculated the change in Perception of Drugs by subtracting the mean score collected before the manipulation from the score obtained after. A negative score indicated that the participant held a more favorable perception of drugs after manipulation; a positive score indicated a less favorable perception of drugs; and zero indicated no change. This measure ranged from −1.17 to 0.92 in the sample.

Perceived Threat to Freedom and Perceived Unrealism

Our main predictors included Perceived Threat to Freedom and Perceived Unrealism. Perceived Threat to Freedom was measured using four items rated on a five-point Likert scale from 1 (strongly disagree) to 5 (strongly agree), developed and tested by Dillard and Shen (2005). Participants were asked to rate all four items. The scale exhibited good consistency (α = 0.800). We then calculated the mean scores for the four items. A higher mean score indicates a greater perceived threat to freedom. The mean score ranged from 1 to 4.75 in the sample.

Perceived Unrealism was measured using the Perceived Realness Scale (Fishbein et al., 2002), which comprises four items rated on a four-point Likert scale ranging from 1 (definitely yes) to 4 (definitely no). Participants were asked to rate all four items. The scale reached acceptable consistency (α = 0.743). We then calculated the mean scores for the four items. A higher mean score indicates that the participants perceived the anti-substance messages as less realistic. The mean score ranged from 1 to 3.5 in the sample.

Analytical Strategies

We used a two-way mixed ANOVA and ordinary least squares (OLS) regression models with robust standard errors (Osborne, 2016) to answer our research question and test hypotheses. An alpha level of 0.05 was used for all tests, and effect sizes (R²) were reported. First, a two-way mixed ANOVA assessed the effect of group assignment and manipulation. Then, OLS regression models estimated the effect of main predictors on the change in drug perception. Marginal effects were calculated for improved illustration. We included all demographic information as controls in the analyses to ensure robustness. All analyses were performed in STATA 17.

Results

We conducted a two-way mixed ANOVA to examine the influence of group assignment and exposure to anti-drug use ads on participants’ perceptions of drugs (Table 3). The findings indicated a significant interaction between the group assignment (experiment vs. control groups) and the timeframe (pre- and post-manipulation) (F (1,481) = 6.36, p = .012). Post hoc analysis revealed that while perceptions of drugs remained largely unchanged in the experiment group from pre-to post-manipulation (diff = 0.013, SE = 0.017, t = 0.78, p = .438), the control group experienced a significant shift towards a more favorable perception of drugs post-manipulation compared to pre-manipulation (diff = −0.047, SE = 0.017, t = −2.81, p = .005). This interaction effect was also supported by the visualization (Figure 1).

Table 3.

Two-Way Mixed ANOVA Results on Perception of Drugs.

Source of variation	SS	df	MS	F	p-value
Group assignment (experiment or control)	0.601	1	0.601	1.56	0.212
Between-subject error	185.021	481	0.385
Timeframe (pre- or post-manipulation)	0.070	1	0.070	2.00	0.158
Interaction (group assignment × timeframe)	0.223	1	0.223	6.36	0.012
Within-subject error	16.869	481	0.035
Total	185.919	484	0.384	10.95	<0.001

Note. SS = Sum of squares, df = Degrees of freedom, MS = Mean sum of squares, F = F-ratio.

The analysis of simple main effects indicated that the group assignment did not significantly affect participants’ perception of drugs (p = .212). Similarly, the impact of anti-drug use ads on perceptions of drugs was not statistically significant (p = .158).

To test Hypotheses 2 and 3, we estimated a set of ordinary least squares regression models using the change in Perception of Drugs as the outcome variable (Table 4). The bivariate regression model results demonstrated that, holding all else constant, participants in the experimental group scored 0.069 higher in Perception of Drugs compared to those in the control group (Model 1).

Table 4.

Regression Models on Changes in Perception of Drugs (12 Items).

Variables	Full sample				Control	Experiment
Variables	Model 1	Model 2	Model 3	Model 4	Model 5	Model 6
Experiment versus control	0.069** (0.025)	0.069** (0.025)	0.062* (0.026)	0.063* (0.026)	-	-
Perceived threat of freedom	-	−0.025 (0.016)	-	−0.017 (0.017)	−0.020 (0.023)	−0.013 (0.025)
Perceived unrealism	-	-	−0.051* (0.025)	−0.042 (0.027)	−0.012 (0.039)	−0.089* (0.035)
Constant	−0.310 (0.172)	−0.263 (0.176)	−0.174 (0.187)	−0.165 (0.188)	−0.167 (0.258)	−0.077 (0.274)
N	483	483	483	483	246	237
R-squared	0.084	0.088	0.091	0.093	0.108	0.161

Note. Robust Standard Errors in parentheses. ***p < .001, **p < .01, *p < .05. Control variables are included in all models but omitted in results.

The models did not demonstrate significant effects of perceived threat to freedom on the change in Perception of Drugs (Models 2 and 4). Perceived unrealism demonstrated a significant association with a more favorable Perception of Drugs (Model 3). A one-unit increase in the perceived unrealism score was associated with a 0.051 point decrease in the change in Perception of Drugs. In other words, when a participant perceived an advertisement to be less realistic, the effect of the manipulation on their perception of drugs became substantially smaller. However, this effect was not significant when perceived threat to freedom was also included in the model (Model 4).

Given the significant interaction between group assignment and manipulation, we further estimated models for subgroups to examine the effects of the two main predictors (Table 4, Models 5 and 6). For the control group, neither perceived threat to freedom nor perceived unrealism had a statistically significant effect on the difference in perceptions toward drugs. In contrast, for the experimental group, perceived unrealism had a significant negative effect on the change in Perception of Drugs, suggesting that a higher score in perceived unrealism is associated with a more favorable perception of drugs after the manipulation. These results further corroborated the observations in the t-tests findings (Table 2), revealing significant differences in perceived unrealism and drug perceptions between the control and experimental groups.

Discussion

Despite the rise of online health campaigns, traditional media platforms such as TV programs and posters remain popular for disseminating anti-drug use messages. Academic research has explored various factors that contribute to the effectiveness of anti-drug use messages and perception of drugs (Armstrong et al., 2021; Massey et al., 2022; Slavin & Earleywine, 2019; Zhang, 2020). Psychological reactance theory and the perceived realism perspective offer two distinct frameworks for assessing the impact of anti-drug use advertisements (Brehm & Brehm, 1981; Li & Sundar, 2022). However, although research on the effectiveness of anti-drug use advertisements has been conducted in other regions, there is a dearth of similar studies in Hong Kong. This study contributes to the limited body of research in this context by bridging the insights from the two theoretical perspectives.

Our findings indicate that exposure to anti-drug use advertisements did not significantly affect participants’ perceptions of drugs. Interestingly, the perceived threat to freedom did not significantly influence perceptions of drugs, whereas higher perceived unrealism was associated with a smaller change in the perception of drugs between the pre- and post-manipulation periods.

Surprisingly, our findings do not support the effectiveness of anti-drug use advertisements in shaping individuals’ perceptions of drugs. However, we argue that the null finding is valuable as it denies the existence of a boomerang effect of anti-drug use advertainments in the Hong Kong context, highlighting the necessity of regional studies that situate in a non-western context. As aforementioned, studies in this field were predominantly conducted in American and European contexts (Czyzewska & Ginsburg, 2007; Fishbein et al., 2002; Ginsburg & Czyzewska, 2005; Hornik et al., 2008; Jacobsohn, 2007; Slavin & Earleywine, 2019; Terry-McElrath et al., 2011), which often emphasizes the potential counterintuitive effects of media use in terms of anti-drug education. Our results support a non-Western-centric perspective, as highlighted by prior studies conducted in Asian populations and cultures (Kim et al., 2017; Ng et al., 2021; Savani et al., 2008). Perhaps the deep-rooted Chinese collectivist culture, despite Hong Kong’s history of British administration and its status as a highly internationalized region, has contributed to limiting the negative effect of anti-drug use advertisements.

Similarly, our null finding regarding the relationship between Perceived Threat to Freedom and the Perception of Drugs may be explained by the collectivist culture, which leads to less concern for individual freedom in general (Kim et al., 2017; Ng et al., 2021; Xu, 2019). Previous studies have demonstrated that when certain behaviors are highly valued within a social context, individuals tend to react negatively when their freedom to engage in those behaviors is restricted (Boine et al., 2020; Buttrick, 2020). Therefore, our participants, who perceived drugs as harmful and attached less value to drug use, may have had a less negative perception of high freedom-threatening messages in anti-drug use campaigns, as compared with other contexts where the freedom of drug use is more valued. Alternatively, it may also be further explained by the predominantly negative perception of drugs among our participants from the outset, which may have limited the observable variation in our models. Further research is needed to determine whether this is a result of the low value attached to drug use, as we speculated, or of the generally lower value placed on individual freedom, as indicated in previous research (Kim et al., 2017; Ng et al., 2021; Xu, 2019). While this study is unable to further unpack the cultural elements from the advertisement, it calls for further considerations of a deep-rooted Chinese collectivist culture in anti-drug education.

The association between higher levels of perceived unrealism and more favorable perceptions of drugs aligns with previous research (Cho et al., 2013; Czyzewska & Ginsburg, 2007; Fishbein et al., 2002; Ginsburg & Czyzewska, 2005; Terry-McElrath et al., 2011), indicating that although the actors and narratives in health promotional advertisements are fictional, they can still impact real-world perceptions if they are perceived as realistic or likely to occur (Busselle & Bilandzic, 2008; Van Leeuwen et al., 2013). In particular, for anti-drug use advertisements to be effective, it is crucial for messages to be perceived as realistic. This aligns with recommendations that emphasize the importance of a coherent and logically realistic narrative, even within a fictional context (Busselle & Bilandzic, 2008).

Our findings regarding the significant shift in the control group towards less favorable perceptions of anti-drugs perception is unexpected. Our analysis shows that such effect is not a result of either perceived unrealism or perceived freedom threat to the material we presented to them. While our data cannot provide evidence to explain such phenomenon, one possible reason is that the material we had shown to the control group is taken from a hygiene manual used during the COVID-19 pandemic. Stress brought by it could have induced pro-drug perceptions (Patterson et al., 2021; Taylor et al., 2021). Whether such negative spillover effect existed and could indeed lead to more negative perception of anti-drug messages require further investigation (Angelucci & Di Maro, 2016). Another possible reason that the control group’s perception of drugs turned to be more positive, according to social learning theory, could be that the experiment itself could have provided our participants an opportunity and motivation to talk about drug use with their peers who might have more positive attitude toward drug use (Kruis et al., 2020). Whether such social learning effects indeed exist also require further investigations. However, it does indicate that drug perception is affected by multiple factors. It would be important for drug education efforts to consider external influences on drug perceptions, like the stress brought by COVID-19, and incorporating strategies to address them, rather than solely focusing on the content of the educational materials.

This study has several limitations that should be acknowledged. First, the participants were primarily Cantonese-speaking Chinese residents of Hong Kong in their early 20s who were enrolled in a degree program. Although multiple measures confirmed the success of randomization, replication in other cultures, regions, and populations with diverse backgrounds will strengthen the validity of our research. For example, we are unsure whether and to what extent our participants have overseas experiences, which may play an important role in shaping their attitudes toward drugs through multiple pathways (Cutler, 2014; Fan & Ashdown, 2014; Rosenthal et al., 2006). Therefore, it is necessary to conduct further research to explore the applicability of these findings to diverse populations with varying perceptions toward drug use. Second, this study employed relatively simple definitions of threats to freedom and perceived unrealism, which may not have captured the full complexity of these constructs (Dillard & Shen, 2005; Fishbein et al., 2002). More comprehensive scales can improve the accuracy of these measurements (Chan et al., 2018; Kim et al., 2020), allowing further unfurling of the mechanisms and interplay of these aspects. Finally, owing to the constraints of the experimental design, we could not identify the exact effect of each experiment, limiting the ability of this study to guide specific anti-drug use advertising designs.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the Faculty Development Scheme of the Hong Kong Research Grants Council; UGC/FDS16/H10/21.

ORCID iDs

Vincent Shing Cheng

Chenghui Zhang

Xinyi Situ

Yiu Chung Ma

Author Biographies

Vincent Shing Cheng is an Assistant Professor in the Department of Applied Social Sciences at the Hong Kong Polytechnic University. His research interests include substance use prevention and treatment, crime and deviant behaviors, and sociology of punishment.

Chenghui Zhang is an Assistant Professor in the Department of Sociology at the University of Nevada, Las Vegas. Her research interests include experimental designs, victim decision-making, bias crimes, and social inequalities.

Xinyi Situ is a Ph.D. Candidate in the Department of Criminology and Criminal Justice at the University of Maryland, College Park. His research interests include quantitative methods, crime and deviant behaviors, the spatial distribution of urban crimes, community safety, and policing strategies.

Yiu Chung Ma is an Assistant Professor in the School of Arts & Social Sciences at the Hong Kong Metropolitan University. His research interests include public policy analysis, managerial incentive, and bank governance.

References

Angelucci

Di Maro

(2016). Programme evaluation and spillover effects. Journal of Development Effectiveness, 8(1), 22–43. https://doi.org/10.1080/19439342.2015.1033441

Armstrong

Richards

A. S.

Boyd

K. J.

(2021). Red-hot reactance: Color cues moderate the freedom threatening characteristics of health PSAs. Health Communication, 36(6), 663–670. https://doi.org/10.1080/10410236.2019.1700885

Audit Commission . (2017). Efforts of the Narcotics division and Beat drugs Fund in combating drug abuse. Audit Commission.

Bang

Choi

Yoon

Baek

T. H.

Kim

(2021). Message assertiveness and price discount in prosocial advertising: Differences between Americans and Koreans. European Journal of Marketing, 55(6), 1780–1802. https://doi.org/10.1108/EJM-10-2019-0791

Boden

Day

(2023). Illicit drug use in university students in the UK and Ireland: A PRISMA-guided scoping review. Substance Abuse Treatment, Prevention, and Policy, 18(1), 18. https://doi.org/10.1186/s13011-023-00526-1

Boine

Siegel

Ross

Fleegler

E. W.

Alcorn

(2020). What is gun culture? Cultural variations and trends across the United States. Humanities and Social Sciences Communications, 7(1), 21. https://doi.org/10.1057/s41599-020-0520-6

Brehm

J. W.

(1966). A theory of psychological reactance. Academic Press.

Brehm

S. S.

Brehm

J. W.

(1981). Psychological reactance: A theory of freedom and control. Academic Press.

Buller

D. B.

Walkosz

B. J.

Gill Woodall

(2019). Use of Media and Social Media in the Prevention of Substance Use. In Sloboda

Petras

Robertson

Hingson

(Eds.), Prevention of Substance Use (pp. 319–334). Springer.

10.

Busselle

Bilandzic

(2008). Fictionality and perceived realism in experiencing stories: A model of narrative comprehension and engagement. Communication Theory, 18(2), 255–280. https://doi.org/10.1111/j.1468-2885.2008.00322.x

11.

Buttrick

(2020). Protective gun ownership as a coping mechanism. Perspectives on Psychological Science, 15(4), 835–855. https://doi.org/10.1177/1745691619898847

12.

Cannon

J. S.

Byrne

(2019). Boomerang effects. In Hobbs

Mihailidis

(Eds.), The international encyclopedia of media literacy. Wiley. https://doi.org/10.1002/9781118978238.ieml0016

13.

Cao

(2021). Seeing is believing: The impacts of visual exemplars on American young adults’ reactions to anti-prescription opioid campaigns. Journal of Drug Education, 50(3-4), 67–83. https://doi.org/10.1177/00472379211072856

14.

Chan

J. Y.

Harlow

A. J.

Kinsey

Gerstein

L. H.

Fung

A. L. C.

(2018). The examination of authoritarian parenting styles, specific forms of peer-victimization, and reactive aggression in Hong Kong Youth. School Psychology International, 39(4), 378–399. https://doi.org/10.1177/0143034318777781

15.

Cheng

V. S.

Lapto

F. K.

(2023). The Impact of Later Life Events on Cessation Motivation of Older Adults with Substance Use Disorder in Hong Kong. International Journal of Offender Therapy and Comparative Criminology, 67(13–14), 1362–1382. https://doi.org/10.1177/0306624X2311654

16.

Cheung

Y. W.

Cheung

W. T.

(2018). Psychoactive drug abuse in Hong Kong: Life satisfaction and drug use. Springer.

17.

Cho

Shen

Wilson

K. M.

(2013). What makes a message real? The effects of perceived realism of alcohol- and drug-related messages on personal probability estimation. Substance Use and Misuse, 48(4), 323–331. https://doi.org/10.3109/10826084.2012.762528

18.

Clayton

R. B.

(2022). On the psychophysiological and defensive nature of psychological reactance theory. Journal of Communication, 72(4), 461–475. https://doi.org/10.1093/joc/jqac016

19.

Crano

W. D.

Alvaro

E. M.

Tan

C. N.

Siegel

J. T.

(2017). Social mediation of persuasive media in adolescent substance prevention. Psychology of Addictive Behaviors, 31(4), 479–487. https://doi.org/10.1037/adb0000265

20.

Cutler

K. A.

(2014). Prescription stimulants are “A okay”: Applying neutralization theory to college students’ nonmedical prescription stimulant use. Journal of American College Health, 62(7), 478–486. https://doi.org/10.1080/07448481.2014.929578

21.

Czyzewska

Ginsburg

H. J.

(2007). Explicit and implicit effects of anti-marijuana and anti-tobacco TV advertisements. Addictive Behaviors, 32(1), 114–127. https://doi.org/10.1016/j.addbeh.2006.03.025

22.

Dillard

J. P.

Shen

(2005). On the nature of reactance and its role in persuasive health communication. Communication Monographs, 72(2), 144–168. https://doi.org/10.1080/03637750500111815

23.

Erceg-Hurn

D. M.

(2008). Drugs, money, and graphic ads: A critical review of the Montana meth project. Prevention Science, 9(4), 256–263. https://doi.org/10.1007/s11121-008-0098-5

24.

Evans

Andrade

Pratt

Mottern

Chavez

Calzetta-Raymond

(2020). Peer-to-peer social media as an effective prevention strategy: Quasi-experimental evaluation. JMIR mHealth and uHealth, 8(5), Article e16207. https://doi.org/10.2196/16207

25.

Ewanation

L. A.

Yamamoto

Monnink

Maeder

E. M.

(2017). Perceived realism and the CSI-effect. Cogent Social Sciences, 3(1), Article 1294446. https://doi.org/10.1080/23311886.2017.1294446

26.

Fan

Ashdown

B. K.

(2014). Chinese international students’ value acculturation while studying in the United States. International Journal of Psychology and Behavioral Sciences, 4(4), 157–164.

27.

Fishbein

Hall-Jamieson

Zimmer

von Haeften

Nabi

(2002). Avoiding the boomerang: Testing the relative effectiveness of antidrug public service announcements before a national campaign. American Journal of Public Health, 92(2), 238–245. https://doi.org/10.2105/ajph.92.2.238

28.

Ginsburg

H. J.

Czyzewska

(2005). National antimarijuana ads compared to antitobacco ads: Differences between immediate favorable and unfavorable post-viewing written comments. North American Journal of Psychology, 7(3), 367–378.

29.

Goodstadt

M. S.

Cook

Magid

Gruson

(1978). The drug attitudes scale (DAS): Its development and evaluation. International Journal of the Addictions, 13(8), 1307–1317. https://doi.org/10.3109/10826087809039344

30.

Graupmann

Jonas

Meier

Hawelka

Aichhorn

(2012). Reactance, the self, and its group: When threats to freedom come from the ingroup versus the outgroup. European Journal of Social Psychology, 42(2), 164–173. https://doi.org/10.1002/ejsp.857

31.

Han

K. J.

Kim

(2019). Toward more persuasive diabetes messages: Effects of personal value orientation and freedom threat on psychological reactance and behavioral intention. Journal of Health Communication, 24(2), 95–110. https://doi.org/10.1080/10810730.2019.1581304

32.

Henriksen

Dauphinee

A. L.

Wang

Fortmann

S. P.

(2006). Industry sponsored anti-smoking ads and adolescent reactance: Test of a boomerang effect. Tobacco Control, 15(1), 13–18. https://doi.org/10.1136/tc.2003.006361

33.

Hornik

Jacobsohn

Orwin

Piesse

Kalton

(2008). Effects of the national youth anti-drug media campaign on youths. American Journal of Public Health, 98(12), 2229–2236. https://doi.org/10.2105/AJPH.2007.125849

34.

Jacobsohn

L. S.

(2007). Explaining the boomerang effects of the national youth anti-drug media campaign. University of Pennsylvania.

35.

Jowett

G. S.

O’Donnell

(2018). Propaganda and persuasion. Sage.

36.

Kim

H. J.

Lee

Hong

(2020). Scale development and validation for psychological reactance to health promotion messages. Sustainability, 12(14), 5816. https://doi.org/10.3390/su12145816

37.

Kim

Baek

T. H.

Yoon

Choi

Y. K.

(2017). Assertive environmental advertising and reactance: Differences between South Koreans and Americans. Journal of Advertising, 46(4), 550–564. https://doi.org/10.1080/00913367.2017.1361878

38.

Kingston

R. E.

Marel

Mills

K. L.

(2017). A systematic review of the prevalence of comorbid mental health disorders in people presenting for substance use treatment in Australia. Drug and Alcohol Review, 36(4), 527–539. https://doi.org/10.1111/dar.12448

39.

Kong

T. S. K.

Laidler

K. J.

(2020). The Paradox for Chem-Fun and Gay Men: A Neoliberal Analysis of Drugs and HIV/AIDS Policies in Hong Kong. Journal of Psychoactive Drugs, 52(1), 77–85. https://doi.org/10.1080/02791072.2019.1568648

40.

Kruis

N. E.

Seo

Kim

(2020). Revisiting the empirical status of social learning theory on substance use: A systematic review and meta-analysis. Substance Use & Misuse, 55(4), 666–683. https://doi.org/10.1080/10826084.2019.1696821

41.

LaVoie

N. R.

Quick

B. L.

Riles

J. M.

Lambert

N. J.

(2017). Are graphic cigarette warning labels an effective message strategy? A test of psychological reactance theory and source appraisal. Communication Research, 44(3), 416–436. https://doi.org/10.1177/0093650215609669

42.

Sundar

S. S.

(2022). Can interactive media attenuate psychological reactance to health messages? A study of the role played by user commenting and audience metrics in persuasion. Health Communication, 37(11), 1355–1367. https://doi.org/10.1080/10410236.2021.1888450

43.

Magura

(2012). Failure of intervention or failure of evaluation: A meta-evaluation of the national youth anti-drug media campaign evaluation. Substance Use and Misuse, 47(13–14), 1414–1420. https://doi.org/10.3109/10826084.2012.705706

44.

Massey

Z. B.

Duong

H. T.

Churchill

Popova

(2022). Examining reactions to smoking and COVID-19 risk messages: An experimental study with people who smoke. International Journal of Drug Policy, 102(9), Article 103607. https://doi.org/10.1016/j.drugpo.2022.103607

45.

Narcotics Division . (2019). Key findings of the public opinion survey on 2018 antidrug publicity measures. Narcotics division, security bureau, the government of the Hong Kong Special Administrative Region of the People’s Republic of China.

46.

Narcotics Division . (2024). Central registry of drug abuse (CRDA). Statistical tables customization. https://www.nd.gov.hk/en/crda.html

47.

A. H.

Kermani

M. S.

Lalonde

R. N.

(2021). Cultural differences in psychological reactance: Responding to social media censorship. Current Psychology, 40(6), 2804–2813. https://doi.org/10.1007/s12144-019-00213-0

48.

Osborne

W . (2016). Regression & Linear Modeling: Best Practices and Modern Methods. SAGE Publications.

49.

Palmgreen

Donohew

Lorch

E. P.

Hoyle

R. H.

Stephenson

M. T.

(2001). Television campaigns and adolescent marijuana use: Tests of sensation seeking targeting. American Journal of Public Health, 91(2), 292–296. https://doi.org/10.2105/ajph.91.2.292

50.

Patterson

Z. R.

Gabrys

R. L.

Prowse

R. K.

Abizaid

A. B.

Hellemans

K. G.

McQuaid

R. J.

(2021). The influence of COVID-19 on stress, substance use, and mental health among postsecondary students. Emerging Adulthood, 9(5), 516–530. https://doi.org/10.1177/21676968211014080

51.

Quick

B. L.

Kim

D. K.

(2009). Examining reactance and reactance restoration with South Korean adolescents: A test of psychological reactance within a collectivist culture. Communication Research, 36(6), 765–782. https://doi.org/10.1177/0093650290346797

52.

Richards

A. S.

Bessarabova

Banas

J. A.

Bernard

D. R.

(2022). Reducing psychological reactance to health promotion messages: Comparing preemptive and postscript mitigation strategies. Health Communication, 37(3), 366–374. https://doi.org/10.1080/10410236.2020.1839203

53.

Rosenthal

Russell

Thomson

(2006). The health and wellbeing of international students at an Australian university. Higher Education, 55(1), 51–67. https://doi.org/10.1007/s10734-006-9037-1

54.

Savani

Markus

H. R.

Conner

A. L.

(2008). Let your preference be your guide? Preferences and choices are more tightly linked for north Americans than for Indians. Journal of Personality and Social Psychology, 95(4), 861–876. https://doi.org/10.1037/a0011618

55.

Shek

D. T.

Zhu

Dou

Chai

(2020). Influence of family factors on substance use in early adolescents: A longitudinal study in Hong Kong. Journal of psychoactive drugs, 52(1), 66–76. https://doi.org/10.1080/02791072.2019.1707333

56.

Slavin

M. N.

(2018). Psychological reactance theory and marijuana craving: State University of New York at. Doctoral dissertation.

57.

Slavin

M. N.

Earleywine

(2019). Effects of messaging and psychological reactance on marijuana craving. Substance Use and Misuse, 54(14), 2359–2367. https://doi.org/10.1080/10826084.2019.1650771

58.

Taylor

Paluszek

M. M.

Rachor

G. S.

McKay

Asmundson

G. J.

(2021). Substance use and abuse, COVID-19-related distress, and disregard for social distancing: A network analysis. Addictive Behaviors, 114(1), Article 106754. https://doi.org/10.1016/j.addbeh.2020.106754

59.

Terry-McElrath

Y. M.

Emery

Szczypka

Johnston

L. D.

(2011). Potential exposure to anti-drug advertising and drug-related attitudes, beliefs, and behaviors among United States youth, 1995–2006. Addictive Behaviors, 36(1–2), 116–124. https://doi.org/10.1016/j.addbeh.2010.09.005

60.

Tsoi

L. Y. D.

Liu

F. M. C.

(2019). Translation, culture and politics: Implications of political slogans in Hong Kong. Translation Spaces, 8(2), 280–299. https://doi.org/10.1075/ts.18009.tso

61.

United Nations Office on Drugs and Crime . (2024). UNODC world drug report 2023. https://www.unodc.org/unodc/en/data-and-analysis/world-drug-report-2023.html

62.

Van Leeuwen

Renes

R. J.

Leeuwis

(2013). Televised entertainment-education to prevent adolescent alcohol use: Perceived realism, enjoyment, and impact. Health Education and Behavior, 40(2), 193–205. https://doi.org/10.1177/1090198112445906

63.

Wong

D. F. K.

Zhuang

X. Y.

T. K.

(2020). Effectiveness of a culturally-attuned integrated cognitive behavioural therapy for Chinese young people with drug abuse in Hong Kong. Children and Youth Services Review, 113(4), Article 104970. https://doi.org/10.1016/j.childyouth.2020.104970

64.

(2019). The impact of self-construal and message frame valence on reactance: A cross-cultural study in charity advertising. International Journal of Advertising, 38(3), 405–427. https://doi.org/10.3760/cma.j.issn.1673-0860.2019.06.002

65.

Zhang

(2020). Effects of freedom restoration, language variety, and issue type on psychological reactance. Health Communication, 35(11), 1316–1327. https://doi.org/10.1080/10410236.2019.1631565

66.

Zhao

Fink

E. L.

(2021). Proattitudinal versus counterattitudinal messages: Message discrepancy, reactance, and the boomerang effect. Communication Monographs, 88(3), 286–305. https://doi.org/10.1080/03637751.2020.1813317