Comparative impact of fear appeals and induced hypocrisy advertising in encouraging intent to quit smoking: Applying self-construal theory to consumers’ attitudes

Research on anti-smoking advertisements

Fear appeals are perhaps the single most representative form of appeal used in anti-smoking campaigns today. The findings on these appeals’ effects, however, remain varied and inconsistent. Many fear appeal theories point to positive linear results, but are unable to explain the boomerang or curvilinear results found in other studiesdies,(5) Witte(5) identified three models to account for fear appeal effects (Witte, 1992): (a) the drive model (6–8), (b) the parallel response model (9), and (c) expectancy value theories (10,11). By integrating these models into an advanced form of extended parallel process model, Witte identified several key concepts and constituent elements to fear appeals. Borrowing from the approach of O’Keefe (12), Witte argued that fear appeals can be defined in terms of their content, or by the reaction they engender from the audience. More specifically, Witte noted that fear appeals usually contain ‘gruesome content’ in the form of vivid language, personal language, or gory pictures. Alternatively, fear appeals have been defined in terms of the amount of fear aroused in and/or experienced by the audience. Fear appeals also offer feasible recommendations that are posited as effective in averting the threat. Thus, the three central constructs in fear appeals are fear, threat, and efficacy.

Also interesting are the findings from comparative analyses of theories applied toward anti-smoking advertising campaigns in the USA and South Korea, two countries with differing sociocultural characteristics. To examine the effects of com-municative anti-smoking campaigns, Hong and Lee (2012) conducted a content analysis on samples of 71 television advertisements aired in the USA and Korea over the 7 years between 2005 and 2012. They found that the advertisements had been based on core theories of health communications, namely the health belief model, theory of reasoned action, and social cognitive theory. Although the findings showed different characteristics in countries’ anti-smoking campaigns, Korean anti-smoking ads were found to rely mostly on social norm messages, followed by smoking attitudes. The message of modelling and self-efficacy was least used in Korea, whereas the US ads focused more on modelling and self-efficacy. Anti-smoking ads in both Korea and the USA most frequently adopted horror and humor as emotional effects rather than sadness, lack of appeal, and anger (13).

Cognitive dissonance theory and hypocrisy paradigm

Cognitive dissonance was defined by Festinger (1957) as an aversive state of psychological tension (dissonance), aroused when an individual holds two inconsistent cognitions. Festinger considered people to be motivated to reduce this uncomfortable state by changing one or both of the inconsistent cognitions (4,14). The classic theory of Festinger (4) of cognitive dissonance is in some respects a theory describing how hypocrisy leads to psychological tension. Using this idea, Dickerson et al. (15) and Fried and Aronson (16) induced recognition of hypocrisy in participants by confirming their support for specific issues and then reminding them that their own actions contradicted their past positions. They also confirmed that this induced recognition of hypocrisy gave rise to cognitive dissonance and made participants acutely uncomfortable.

Studies that have considered induced hypocrisy in the past have typically addressed public activities, e.g. the use of condoms (3,17), water conservation (15), recycling of household waste (16), respecting traffic laws (18), and sober driving (19).

Taken together, the above findings indicate that communication using induced hypocrisy may be applied to issues in which certain behaviours are judged to be socially desirable, and that induced hypocrisy may be a useful technique in changing attitudes and behavioural intentions by generating cognitive dissonance. This study attempts an empirical examination of the effects of anti-smoking advertisements using the induced hypocrisy method on smoker attitudes and behavioural intentions toward smoking cessation. It also seeks to compare these effects with those of fear appeals, which have frequently been used in anti-smoking messages in the past. Because advertising effects may differ by respondent characteristics, it applies self-construal theory to examine differences by personal characteristics.

Self-construal theory

The origins of self-construal theory lie in criticisms of research on the self in areas such as psychology and anthropology. Markus and Kitayama (20) criticised previous research in self-construal theory as focusing only on forms of the self as manifested in Western culture, while ignoring various other aspects of the self.

The researchers identified two categories of self in self-construal theory, which they identified as the independent and interdependent self. People who fall into the independent self category tend strongly to perceive themselves as independent from others and to prefer expressions of their own strengths as distinct from those of others. People who fall into the interdependent self category, in contrast, are greatly influenced by relationships with others and tend to avoid revealing themselves in collective life, focusing mainly on their role in community living (20).

Arnocky et al. (21) further expanded self-construal theory by considering the question of whether individual differences in self-construal predict differences in environmental concern, resource sharing, and pro-environmental behaviour. In their findings, they reported differences in perspectives on environmental issues. In particular, they found that independent self-construal uniquely predicted egoistic environmental concern and competitiveness in sharing resources, whereas interdependent self-construal predicted resource cooperation and metapersonal self-construal uniquely predicted biospheric environmental concern, ecological cooperation, and self-reports of environmental conservation behaviour.

This study attempts an empirical examination of the effects of anti-smoking advertisements that use the induced hypocrisy method on smokers’ attitudes and behavioural intentions toward smoking cessation. It also seeks to compare these effects with those of fear appeals, which have frequently been used in anti-smoking messages in the past. Because advertising effects may differ by respondent characteristics, it applies self-construal theory to examine differences by personal characteristics. The following research questions were thus formulated:

Research design and experimental stimuli

To examine differences in responses to induced hypocrisy and fear appeal advertisements, this study adopted a between-subjects factorial design using two experimental groups. Mock anti-smoking newspaper advertisements were developed as experimental stimuli. Stimuli were respectively designed with induced hypocrisy and fear appeal content, although otherwise being kept as similar as possible in terms of size, layout, copy, and photographs. The mock advertisements were then examined by two professors of advertising with several years of experience producing advertisements at a professional agency, after which necessary adjustments were made.

The study’s induced hypocrisy advertisement used an image of a crying child who has been exposed to cigarette smoke. The smoke in the advertisement was digitally altered into the shape of a clear plastic bag to create the visual effect of the child being suffocated by it. The main copy on the advertisement read, ‘You said you value your promises! Aren’t you ashamed of yourself?’

The fear appeal advertisement used in the study shows the face of a smoking adult, but with the cigarette embedded in the subject’s head rather than between the lips or fingers. The image of emerging smoke is depicted in a horrific manner. The headline reads, ‘When you smoke, cigarettes burn your brain’.

Data collection and research procedure

This study adopted a web-based experimental method for sample extraction. The survey was conducted in August 2015 by a specialised company with a panel of 680,000 and ample experience with online questionnaires. Emails were sent to a randomly chosen group of the company’s 680,000 panelists who had participated in at least one survey in the 12 months before the date of sending. To ensure consistency between experimental groups, gender and age ratios were selected to conform to their distribution in the 2010 population census as announced by Statistics Korea.

In view of the average recent rate of response for the company, emails were sent to a total of 1400 participants who indicated that they were currently smokers. Measures were taken to ensure that no participant would participate in multiple surveys. Emails for the sample in the induced hypocrisy advertisement condition were sent to a total of 700 people, of whom 387 opened the email. One-hundred did not participate in the survey, leaving a group of 287 who participated. A total of 207 responses were discarded according to quota or because of responses judged to be insincere, such as those with a response time of 30% or less compared to the average or identical responses to multiple items. The final sample group thus consisted of 80 individuals.

For the fear appeal advertisement sample, emails were sent to another 700 individuals, of whom 401 opened the email. A total of 102 did not participate, leaving a group of 299 who participated. Out of those responses, 219 were discarded according to quota or because of responses judged to be insincere, such as those with a response time of 30% or less of the average or identical responses to multiple items, leaving a final sample group of 80 individuals. Each experimental group consisted of 40 males (50%) and 40 females (50%), with 15 respondents aged 20–29 (18.8%), 17 aged 30–39 (21.3%), 18 aged 40–49 (22.5%), and 30 aged 50 or over (37.5%).

The experimental survey consisted of the following steps. For the group exposed to the induced hypocrisy condition, participants first received an explanation of the study’s purpose, after which their self-construal tendencies and attitudes toward smoking cessation were measured. After presentation of the induced hypocrisy stimulus, their attitude toward smoking cessation and intention to quit were measured. The same standards were applied in the experimental group exposed to a fear appeal advertisement.

Measures

Self-construal tendency. The self-construal tendency was measured on a nine-point Likert scale using seven items adapted from the measure in Hardin et al. (22). Statements to measure self-construal included ‘Speaking up during a class (or a meeting) is not a problem for me’, ‘I am comfortable with being singled out for praise or rewards’, ‘I do my own thing, regardless of what others think’, ‘I feel comfortable using someone’s first name soon after meeting’, ‘I enjoy being unique and different from others in many respects’, ‘Personal identity, independent of others, is very important to me’, and ‘I act the same way no matter who I am with’.

Attitude toward smoking cessation. The attitude toward smoking cessation was measured on a nine-point Likert scale using six items also adapted from those used in Holbrook (23) and Simons et al. (24). Statements included, ‘I have positive feelings about quitting smoking’, ‘I think quitting smoking is useful’, ‘I think quitting smoking is beneficial’, ‘I think quitting smoking has considerable value’, ‘I think quitting smoking would be satisfactory for me’, ‘I think other people would think positively about me quitting smoking’, and ‘Quitting smoking is something of high quality’.

Intention to quit. The intention to quit was measured on a nine-point Likert scale using three items adapted from the measures in Bearden and Mason (25). Statements included, ‘I may quit smoking (possible)’, ‘I would advise that others quit smoking (likely)’, and ‘I think it would be wise for someone close to me to quit smoking (probable)’.

Although the aforementioned research design, experimental stimuli, and data collection methods were used for empirical testing of the research questions, a number of limitations suggest that caution is warranted in interpreting or generalizing the findings. First, because the participants in the study were all Korean, it may be possible to draw more meaningful theoretical and practical implications by applying the same approach to countries with different sociocultural characteristics. Moreover, only printed advertisements were used in both the fear appeal and induced hypocrisy conditions. Because effects may differ according to advertisement form, future researchers may reach more effective conclusions by examining a broader range of media, including broadcast, online, and mobile advertisements.