Got juice? A test of health-related information processing.

Pan, Chia-Hsin

In recent years, sugary beverage has become one of the main sources of daily sugar intake. Over-consumption of sugar-sweetened soft drinks and juice may cause dental caries, decreased bone mineral density, obesity, diabetes, and cardiovascular diseases that are concerned by public health researchers (Bray, Nielsen, & Popkin, 2004; Brown, Dulloo, & Montani, 2008). Furthermore, the consumption habit of sugary drinks formed during young adulthood has a strong impact on beverage choices in later life (Ha, Bish, Holloman, & Gordon, 2009). That is the reason why many brandname sugary drinks target their advertising campaigns on young adults.

The controversial Bomy fruit and vegetable juice's advertising campaign is one of the cases in Taiwan. The leading brand-name juice company seems to promoted 5-a-day fruit and vegetables healthy diets, while they suggests that consuming a bottle of fruit and vegetable juice is a natural and ideal way for the need. Later a consumer protection group rebuked the campaign for misleading consumers to drink the unhealthy, juice-flavored sugary water and resulted in a cessation of the campaign. Especially both information campaigns employ the Internet and their official Web sites to reach young consumers.

The new media not only easily built and designed by both commercial and nonprofit organizations but also user-friendly visited by health information seekers. According to a survey by Taiwan Network Information Center, there are 15.8 million Internet users, 68.94% of the population, in Taiwan (Tccii, 2009). However, the question of credibility bothers these health seekers because some Web sites seem to be more concerned with selling products than with providing accurate information.

Some scholars suggest that the new media has blurred the line between advertising and informational content (Flanagin & Metzger, 2000; Metzger, Flanagin, Eyal, Lemus, & McCann, 2003). A commercial Web site can promote its products or services by enticing potential buyers with a huge volume of related objective information. Companies attempt to manage consumers' perceptions of their trustworthiness through high quality in their Web site design and emotional appeal to elicit feelings of identification with them (Chadwick, 2001).

However, message receivers tend to identify a higher source credibility of nonprofit organizations than corporations. Source credibility is defined and assigned by the perceptions of the receivers, which are not inherent in the objective characteristics of the speaker or organization (Berlo, Lemert, & Mertz, 1969; Hammond, 1987; Haley, 1996). Especially, Web sites are seen as more credible if they represent a nonprofit organization or have a domain name that ends with ".org" (Fogg, 2003). Still, the issue of source credibility is important for online health promotions.

Thus, the purpose of this study is to investigate how target audiences process Web-based health campaign messages, especially the effects of individual proclivity of thinking styles, campaign message sources, and strategies on attitude and behavior changes are examined.

Literature Review

Dual Process Theories of Persuasion

How do people process and make decisions about Web-based health messages provided by different sources? Basically, past research proposed that individuals tend to employ a two-route mechanism, either thoughtful or un-thoughtful processing in accordance with personal relevance or irrelevance, and levels of motivation or ability. One of the most studied dual-process theories is the elaboration likelihood model (ELM) that proposes: When motivation and ability are relatively low, people tend to rely on less-effort peripheral route to process such as source characteristics of the information. When motivation and ability are relatively high, people tend to use central route to thoughtfully scrutinize the arguments presented (Petty & Cacioppo, 1986; Petty & Wegener, 1999).

Sharing basic conceptions with ELM, the heuristic-systematic model (HSM) suggests that high-relevant recipients employ thoughtfully systematic processing strategy because of message cues, whereas low-relevant subjects employ heuristic processing strategy because of source cues (Chaiken, 1980). However, the two modes might co-occur and the co-occurrence of heuristic processing could bias systematic processing. Especially, when messages are ambiguous or incongruent to recipients' values, defensive processes will be activated (Chen & Chaiken, 1990; Libermann & Chaiken, 1992; Chaiken & Maheswaran, 1994). For example, a low-credible source provides relevant information but incongruent to a recipient's attitude or belief, thus the biased factor might be elaborately corrected when one is aware of the possible bias. Then a disliked source could be more persuasive than a liked source (Petty & Wegener, 1999).

Rather than testing individual situational variables, the cognitive-experiential self-theory (CEST; Epstein, 1994; Epstein & Pacini, 1999) looks at the personality of the self and considers the rational and irrational styles by which an individual processes new information. Those individual traits are more stable and worthy of longer-term examinations.

CEST proposes that: People adapt to their environments by ways of two information process systems: a preconscious experiential system and a primarily conscious rational system. The two systems operate in parallel and are interactive. The rational system operates through a person's understanding of logical rules of inference, a deliberative, analytical system that operates primarily in the medium of language and is relatively affect-free. In contrast, the experiential system encodes information in a concrete, holistic, primarily non-verbal form and operates according to heuristic principles, and despite the limitations of such operation, it has the advantage of being far more rapid and efficient than the rational system for coping with events in everyday life. The operation of the experiential system is assumed to be intimately associated with the experience of affect (Epstein & Pacini, 1999).

Berger, Johnson, & Lee (2003) conducted two experiments to examine the conjoint effects of participant rationality and base-rate information on apprehension about a threat. Results show that high-rationality individuals who first received context-expanding information about a relatively likely hazard (traffic deaths) manifested less apprehension in response to a subsequent news story about a less likely threat (anthrax death) than did highs who read only the anthrax death story. Among low-rationality individuals this relationship is reversed. Applying Rational-E-experiental Inventory (REI; Pacini & Epstein, 1999) to measure participants' responses to news stories about the outbreak of SARS, Zhou, Pan, & Zhoug (2008) found that high-rationality individuals would be affected by the story with contextual information provided to assuage their apprehension.

Applying this approach, the present study aims to understand whether people's thinking styles and levels of thoughtful processes bear any significance in Web-based health message processes. It is expected that individual rationality will mediate participants' processes of health-related information, interact with the message content variable, and thus conjointly affect their short-term changes of attitude toward healthy diets (H1) and behavioral intentions (H2).

Sleeper Effects

Kahneman & Tversky (1979) suggested that the way information is framed will affect its recipient's decision and that decision is based on benefit gain or loss. Such heuristic information processing is based on limited individual experience and circumstanced influence (Zillmann & Brosius, 2000). In practice, the applications of message framing emphasize not only the negative consequences of rejection of suggested behavior, which may be similar to fear appeals, but also the positive consequences of adoption of suggested behavior.

Rothman & Salovey (1997) suggested that the relative effectiveness of gain or loss framed appeal depends on whether the suggested behavior serves as an illness-detection or a health-affirming function. Generally, studies for illness-detection promotion found that negatively loss-framed messages were more persuasive (e.g. Schneider, Salovey, Apanovitch, Pizarro, McCarthy, Zullo, & Rothman, 2001, mammography). In contrast, studies for health-affirming promotions found that positively framed messages were more persuasive (e.g. McCaul, Johnson, & Rothman, 2002, flu shot). Since the promotion of healthy diets serves a health-affirming function, positively gain-framed messages should be more persuasive than loss-framed messages as revealed in previous studies.

However the intended effects might vary over time, then the unintended effects of health promotions should be examined (Cho & Salmon, 2007). Especially the present study is to test how participants process a positively gain-framed message and a negatively loss-framed counter-arguing message over time. In general, effects attenuate as time lapses. However, under some circumstances, effects may increase over time. Then unintended or undetected effects in the short term may be found over the long term. For example, the sleeper effect, first referred by Hovland, Lumsdaine, & Sheffield (1949), explained the phenomenon of immediate and delayed influence of a second-world-war propaganda film on American soldiers, who had greater attitude change rather than attitude decay nine weeks after viewing the film.

Later, Hovland & Weiss (1951) proposed the memory-for-source hypothsis that message recipients forget the noncredible source as time goes by, and thus the association between the source of the discounting cue and the message fade over time. Kelman & Hovland (1953) hypothesized that the source and message were reinstated should not observe the sleeper effect, whereas the source and message were not reinstated should observe the sleeper effect.

However, Gillig & Greenwald (1974) criticized the memory-for-source hypothesis and questioned the existence of the sleeper effect. Then, conditional explanations for the sleeper effect were supported by other studies. In order to observe the emergence of the sleeper effect, the discounting cue must inhibit the message's persuasiveness, such as a negative or untrustworthy source must dissociate from the message before the measure of delayed attitudes (Cook, Gruder, Hennigan, & Flay, 1979), and the trustworthiness of the source must be rated immediately after message exposure (Pratkanis, Greenwald, Leipe, & Baumgardner, 1988).

Recently, studies have found that greater amount of thought and elaboration of attitude-relevant message scrutiny will persist over time (Haugtvedt & Petty, 1992) and influence subsequent behavior (Cacioppo, Petty, Kao, & Rodriquez, 1986) than will less thoughtful information process. Priester, Wegener, Petty, & Fabrigar (1999) suggested that some factors tend to influence the amount of elaboration, such as individual levels of rationality.

That is, some conditions and mechanisms should be observed first as follows in order to predict the exhibition of sleeper effects: (1) The message must be persuasive on initial attitudes. (2) The discounting cue must be strong enough to inhibit the effect of the message on the attitudes. (3) The message and source must become dissociated before the delayed measure. (4) The delayed attitude of the message only group must be higher than the initial attitude of the discounting cue group. (5) The subjects pay attention to the message. (5) A discounting cue must be received after the message. (6) The subjects must rate the trustworthiness of the message source immediately after receiving the discounting cue. (Lariscy & Tinkham, 1999; Priester, Wegener, Petty, & Fabrigar, 1999)

Thus, the present study hypothesizes that the discounting cue of a nonprofit organization following a campaign message from a juice company will be more likely to influence health-related attitude and behavior changes over time among the high-rationality participants, however the effect will not be observed among low-rationality participants. Therefore, an interaction effect between rationality and discounting cue is expected. Changes of attitude toward healthy diets (H3) and behaviors (H4) will be observed over time.

Method

Participants

Participants were students taking introductory communication courses in a university in Taipei, Taiwan. They were recruited voluntarily and allowed to quit at anytime during the experiment. Students were told that they would receive extra class credits as encouragement. Participants who finished all experiment sessions included 69 undergraduate students, with 17 male and 52 female subjects (ages ranged from 18 to 33 years, M = 21.29, SD = 3.43). In addition, participants would be classified as high or low in rationality for processing information.

Procedure

The present experiment was a 2 x 2 (high/low rationality x with/without discounting cue) between-subject factorial design included two sessions. The first session of the experiment was conducted in two computer labs for two treatment conditions. Participants were randomly assigned to one of the two conditions: participants in both conditions would read a persuasive message inserted in a corporate Web page, but only in with-discounting-cue condition exposed to one more counter-arguing message inserted in a nonprofit organization's Web page following the persuasive message. Thus the persuasive-message-only condition could provide baseline data.

After signing an informed consent form, they were instructed to complete the first part of a two-part questionnaire and that there are no right or wrong answers and no talking could occur during the session. This part included 40 items of the 5-point REI scales, questions of health-related attitudes, behaviors, and some demographic questions (Time 0). The REI scales measure participants' rational and experiential thinking styles that consist of four subscales. Two subscales index an individual's rational thinking ability and experiential processing. The other two subscales index an individual's tendencies in and preferences for rational and irrational thinking. These sub-scales are combined to form two major subscales to index rationality and experientiality respectively (Epstein & Pacini, 1999).

Participants were then instructed to turn on the computer screen and to read a persuasive message on a corporate Web page. Immediately after exposure to the first Web page, only participants in the discounting-cue condition were asked to read the next message in a Web page operated by a nonprofit organization that dispute the first message. These treatment Web pages were broadcasted and controlled by moderators. Participants could not be exposed to other messages during the first session.

After 10-minute message exposure, participants were instructed to turn off the computer screen and to answered items in the second part of the questionnaire regarding their evaluations of the Web page's credibility, and questions of health-related attitudes and behavioral intentions were asked again (Time 1). Participants were instructed that they could not answer any questions about the experiment before leaving the experiment.

Approximately four weeks later, the second session of the experiment was conducted in participants' classrooms after their courses. Participants completed a questionnaire included the same questions in first session to evaluate credibility of the Web page they exposed in the first session by memory, and health-related attitudes and behaviors were also asked again in this session (Time 2). Participants were debriefed before leaving the experiment.

Treatments

The researcher created mock Chinese Web pages whose operators could be identified by the participants in the study. Considering the ecological validity of the stimulus materials, the Web page template was downloaded from Websites operated by a juice company and a nonprofit organization. The chosen company owns the Bomy fruit and vegetable juice (http://www.bolife.com.tw) in Taiwan. And the chosen nonprofit organization is Taiwan's number one nonprofit organization "Consumer foundation, Chinese Taipei (http://www.consumers.org.tw)." The Microsoft Frontpage was employed to retain Web pages' features such as links, headers, footers, and frames. However, the original messages was deleted and replaced by manipulated texts.

The manipulated text with the title "Juice naturally prevents diseases" was inserted in the Bomy Web page's online newsroom, which states, "nutritionists suggest that juice can reverse the processes of natural aging, cancer and other chronic diseases. Some vitamins and minerals in fruit and vegetable fibers are hard to be digested and absorbed by human bodies, however they can be juiced and released from cell membrane, then absorbed within 15 minutes. Nutritionists suggest that people should have 5-a-day vegetables and, especially for busy working people who eat fast foods frequently. Drinking fruit and vegetable juice is a natural and the best way for them to acquire 5-a-day vegetables and fruit."

Serving as a discounting cue, the consumer news story "Where is juice?" was inserted in the Consumer foundation Web page. It says, "There are numerous vegetable and fruit juice products on the market. On their containers and advertisements emphasize full of vitamin, cellulose, and other nutrients, even providing consumers with everyday need of nutrients from fruit and vegetables. It seems that consumers are no need to eat vegetables and fruit. Is it true? In fact, many nutrients are destroyed and cellulose released during vegetable and fruit juice production. Even worse, extra sugar, color, seasoning, and preservative are added for better flavor. There is not much nutrition left in juice."

Measures

The 20 five-point rationality items of the REI formed a reliable rationality index ([alpha] = .88, M = 3.32, SD = .6), which ranged from 1.8 to 4.85, with higher scores indicative of higher proclivity for rational processing. The 20 five-point experientiality items were combined to form the experientiality index ([alpha] = .76, M = 3.35, SD = .43), which ranged from 2.45 to 4.5. Two sub-scales were not significantly correlated (r = -.15), providing additional evidence for their orthogonality. K-means clustering analysis was used to recode the scores of rationality index into high (n = 30, M = 3.85) and low (n = 39, M = 2.91) rationality groups. Experientiality index was used as covariate in the analysis.

Across three sessions of the experiment, similar scales were used to test participants' health-related attitudes toward healthy diets, and intentions and consumption frequencies of vegetables, fruit, and juice.

Two items of the scales, "I carefully plan my meals to be sure they are nutritious" and "I carefully regulate the amount of sweets or calories I consume," were combined to form the index of attitude toward healthy diets (Time 0: a = .7, M = 3.37, SD = 1.15, ranged from 1 to 6; Time 1: [alpha] = .75, M = 3.71, SD = 1.34, ranged from 1 to 7; Time 2: [alpha] = .74, M = 3.54, SD = 1.26, ranged from 1.5 to 6.5).

Three questions were employed to solicit participants' weekly vegetables, fruit, and juice consumption intentions (Time 1) and frequencies (Time 0 and Time 2). Examples of the questions used were "In a typical week, how many days do you eat vegetables? (Time 0)" or "In the next week, how many days do you plan to eat fruit?" (Time 1), and "In last week, how many days do you drink juice? (Time 2) "

Eight 7-point adjective scales (anchored at 1=not at all, and 7=very) were used to measure participants' evaluation of Web pages' credibility, in which the credibility dimension of trustworthiness included five items: accurate, complete, unbiased, responsible, and trustworthy; the credibility dimension of expertness included three items: expert, important, and wise.

Scores on eight 7-point adjective items were averaged to form a reliable source credibility index [Time 1: [alpha] = .8, M = 4.74, SD = .75, ranged 3.6-6.5 (Bomy-only condition), [alpha] = .78, M = 4.12, SD = .93, ranged 2-6.13 (Bomy), and [alpha] = .87, M = 5.22, SD = .93, ranged 3.38-7 (Consumer foundation); Time 2: [alpha] = .86, M = 4.55, SD = .78, ranged 2.88-6 (Bomy-only condition), [alpha] = .72, M = 3.81, SD = .83, ranged 1.385.13 (Bomy), and [alpha] = .82, M = 4.22, SD = .9, ranged 1.88-5.63 (Consumer foundation)] with one item, biased, reverse scored, and the other seven items, accurate, complete, responsible, trustworthy, expert, important, and wise, positively scored.

Results

Manipulation Check

Regarding some required conditions posited by past studies (i.e. Kelman & Hovland, 1953; Cook et al., 1979; Pratkanis et al., 1988) in order to obtain the sleeper effect, the present study instructed participants to evaluate treatment Web page's credibility immediately after exposure to the message and checking the reinstatement four weeks later. Eight 7-point adjective items of source credibility were used to check the manipulation of the sleeper effect mechanisms.

It revealed that participants perceived the discounting cue from a nonprofit Consumer Foundation Web page (M = 5.22, SD = .93) at Time 1 to be more credible than the juice corporation Bomy Web page (M = 4.12, SD = .93). That is, the discounting cue might have enough strength to inhibit the influence of the campaign message.

However, after 4 weeks of memory decaying, participants at Time 2 evaluated the campaign message in Bomy Web pages (M = 4.55, SD = .78) to be more credible than Web pages in the other with-discounting-cue condition (M = 4.22, SD = .9 for Consumer Foundation, and M = 3.81, SD = .83 for Bomy). It revealed that these Web pages' source cues had been dissociated from the messages over time.

Random Assignment Check

Participants were randomly assigned to two conditions: a campaign message with or without discounting cue. To check whether the random assignment was successful, this study explored the differences among groups in terms of participants' initial measures health-related attitude, behaviors, rationality, and experientiality. A series of t tests were conducted and yield no statistically significant findings. Therefore, the random assignment appeared to be successful.

Hypotheses Tests

To test the hypotheses, a 3 x 2 x 2 (Time x condition x high/low rationality) repeated measure ANOVA was used to test for within-group differences across sessions in which measures of attitude toward healthy diet, consumption intentions and behaviors of vegetables, fruit, and juice were entered as dependent variables. Significant effects were then decomposed and examined by paired /-tests. Results of these analysis revealed a marginal significant three way effect on attitude toward healthy diet, F(2, 64) = 2.6, p = .082, n [p.sup.2] = .08.

Paired t-tests yielded significant attitude change toward healthy diets only in the condition with discounting cue for low-rationality participants between Time 0 and Time 1, /(20) = -2.63, p = .016, 2-tailed (Time 0: M = 3.26, SD = 1.17; Time 1: M = 4.02, SD = 1.5). An immediately positive attitude change toward healthy diets was observed after exposure to the discounting cue, however delayed attitude change toward healthy diet was not found for low-rationality participants (Time 2: M = 3.45, SD = 1.28), whose attitude toward healthy diet went back to the initial level after four weeks. Therefore only the short-term effect on change of attitude toward healthy diets proposed by H1 was supported by this analysis (see Figure 1).

On the other hand, no immediate or delayed attitude change toward healthy diets was found among high-rationality participants in both conditions. And no treatment effect on immediate or delayed behavior emerged for both high--and low-rationality participants. Therefore H2, H3, and H4 were not supported by the analysis. Experientiality proved to be a non-significant covariate (p > .70) when the repeated measure ANOVA was repeated as a repeated measure ANCOVA.

Discussion and Conclusion

The main purpose of this study was to investigate how individual proclivity of thinking styles and web-based message sources and strategies might affect target audiences' processes. The interaction effects between high/low rationality and campaign message with or without discounting cue on participants' attitude and behavior changes were thus examined.

When participants were asked to read and judge Web-based information about healthy-related issues, they would process the information according to their past experiences, memories, emotions, instincts (low-rationality), or with logical thinking (high-rationality) about it. Especially, factors of message attributes might affect participants during the experiment sessions and persistence over time. According to past research, the experiment condition design, such as discounting cues and dissociated source cues, might influence the occurrence of unintended long-term effects, i.e. the sleeper effect.

[FIGURE 1 OMITTED]

This study is characterized by many non-significant findings. Except for one marginal statistically significant finding, null hypothesis was supported in the case and qualified as a short-term effect. That raised some important issues with the design of this study, as well as implications for health-related information processing and influences.

One should first question if the design of the experiment was sound. In an effort to control extraneous variables, we designed the experiment such that two conditions used the identical campaign Web page. As the discounting cue condition added a nonprofit Web page was perhaps the weakest in terms of manipulation. Just providing more words with a negative frame might give off a sense of an unauthentic discounting cue.

However, the nonprofit Web page used in the discounting cue condition was a strong indicator of the gravity of the situation because of the prestigious reputation of the consumer protection organization. On the other hand, the importance of the health issue and the reputation of the brand-name juice might undermine the experiment design as well. Nonetheless, the sleeper effect is hard to find and less research to investigate. As Kumkle & Albarracin (2004) suggested that appropriate baseline measures should be obtained from independent groups of participants, especially a control group should be included in the experiment design in order to compare and observe the absolute sleeper effect. Thus further investigation of health promotions' long-term effects is needed.

On participants' processes of health-related information, Web pages' credibility might be an important attribute of the information especially when the campaign message's for-profit source is less credible than the following discounting cue's nonprofit source. In addition, such treatment design would elicit low-rationality participants' health consciousness and thus the discounting cue might successfully activate their attitude changes toward health diets (see Figure 1).

That is, low-rationality participants tend to heuristic process of the peripheral cues and to be more affected by the discounting cue condition with longer Web messages than affected by the no discounting cue condition. The other explanation for the effect is that the discounting cue condition, a positive message about fruit and vegetable juice added a negative description, can be qualified as a two-sided message, which might be more persuasive than the one-sided campaign message condition. In practice, a two-sided advertisement could be more credible and effective than a one-sided one (Mullen & Johnson, 1990). In short, the findings suggested that health promotions supported by corporate still had relative effectiveness on health promotions and comparative advantages as suggested by other research (Bloom, Hussein, & Szykman, 1997).

On the other hand, there is no significant difference among high-rationality participants exposed to either campaign messages with or without discounting cues. They might thoughtfully process the message or pay less attention to those treatment designs of source cues. A ceiling effect on attitude toward healthy diets was thus revealed, for instance, in response to the corporate persuasion message with a nonprofit's discounting cue across sessions (see Figure 1). In practical applications, these conscientious, thoughtful and value-protective campaign targets are hard to be persuaded (Slater & Rouner, 1996), therefore well-developed persuasion strategies are needed to attract and affect them.

The finding indicated that low-rational participants tended to be affected by the online health-related message's peripheral cues, thus individual differences and message strategies might be important factors in designing effective campaign: The target audiences should be well chosen and the appeal strategies should be soundly managed based on market analyses and the availability of the market data.

In conclusion, the finding of this study can be regarded as suggestive rather than definitive for future research. The generalization of this study is limited because of convenience sampling, treatments of the experiment design, and no control group. Future research could replicate this study in a more diverse population, such as rather younger or older samples than college students, and participants should be assigned to expose to a larger number of randomly selected health messages from different Web sites, such as news media, school's heath center, health departments, personal Web sites, or chat rooms. That is, Web pages combine the characters and advantages of both printed and electronic media. For testing the different effects of text, photograph, animation, video, and other special designs on Web pages, future research should conduct experiment online and have higher ecological validity.

References

Berger, C. R., Lee, E.-J., & Johnson, J. T. (2003). Gender, rationality, and base-rate explanations for increasing trends. Communication Research, 30, 198-223.

Berlo, D. K., Lemert, J. B., & Mertz, R. J. (1969). Dimensions for evaluating the acceptability of message sources. Public Opinion Quarterly, 33, 563-576.

Bloom, P. N., Hussein, P. Y., & Szykman, L. R. (1997). The benefits of corporate social marketing initiatives. In M. E. Goldberg, M. Fishbein, & S. E. Middlestadt (Eds.), Social marketing: Theoretical and practical perspectives (pp. 313-331). Mahwah, NJ: Lawrence Erlbaum Associates.

Bray, G. A., Nielsen, S. J., & Popkin, B. M: (2004). Consumption of high-fructose corn syrup in beverages may play a role in the epidemic of obesity. American Journal of Clinic Nutrition, 79, 537-543.

Brown, C. M., Dulloo, A. G., & Montani, J. P. (2008). Sugary drinks in the pathogenesis of obesity and cardiovascular diseases. International Journal of Obesity, 32, 28-34.

Cacioppo, J. T., Petty, R. E., Kao, C. F., & Rodriquez, R. (1986). Central and peripheral routes to persuasion: An individual difference perspective. Journal of Personality and Social Psychology, 51, 1032-1043.

Chadwick, S, A, (2001). Communicating trust in e-commerce interactions. Management Communication Quarterly, 14, 653-658.

Chaiken, S. (1980). Heuristic versus systematic information processing and the use of source versus message cues in persuasion. Journal of Personality and Social Psychology, 39, 752-766.

Chaiken, S., & Maheswaran, D. (1994). Heuristic processing can bias systematic processing: Effects of source credibility, argument ambiguity, and task importance on attitude judgment. Journal of Personality & Social Psychology, 66, 460-473.

Cho, H., & Salmon, C. T. (2007). Unintended effects of health communication campaigns. Journal of Communication, 57, 293-317.

Cook, T. D., Gruder, C. L., Hennigan, K. M., & Flay, B. R. (1979). History of the sleeper effect: Some logical pitfalls in accepting the null hypothesis. Psychological Bulletin, 86, 662-679.

Epstein, S. (1994). Integration of cognitive and the psychodynamic unconscious. American Psychologist, 49, 709-724.

Epstein, S., & Pacini, R. (1999). Some basic issues regarding dual-process theories from the perspective of cognitive-experiential self-theory. New York: Guilford.

Flanagin, A. J., & Metzger, M. J. (2000). Perceptions of Internet information credibility. Journalism & Mass Communication Quarterly, 77, 515-540.

Fogg, B. J. (2003). Persuasive technology: Using computers to change what we think and do. San Francisco, CA: Morgan Kaufmann.

Gellig, P. M., & Greenwald, A. G. (1974). Is it time to lay the sleeper effect to rest? Journal of Personality and Social Psychology, 29, 132-139.

Ha, E. -J., Bish, N. C., Holloman, C, & Gordon, K. L. (2009). Evaluation of effectiveness of class-based nutrition intervention on changes in soft drink and milk consumption among young adults. Nutrition Journal, 8, 1-6.

Haley, E. (1996). Exploring the construct of organization as source: Consumers' understanding of organizational sponsorship of advocacy advertising. Journal of Advertising, 25, 19-35.

Hammond, S. (1987). Health advertising: The credibility of organizational sources. In M. L. McLaughlin (Ed.), Communication yearbook 10 (pp. 613-628). Newbury Park, CA: Sage.

Haugtvedt, C., & Priester, J. R. (1997). Conceptual and methodological issues in advertising effectiveness: An attitude strength perspective. In W. D. Wells (Ed.), Measuring advertising effectiveness (pp. 79-91). Mahwah, NJ: Lawrence Erlbaum.

Hovland, C. I., Lumsdaine, A. A., & Sheffield, F. D. (1949). Experiments on mass communication. Princeton, NJ: Princeton University Press.

Hovland, C. I., & Weiss, W. (1951). The influence of source credibility on communication effectiveness. Public Opinion Quarterly, 15, 635-650.

Kelman, H. C., & Hovland, C. I. (1953). "Reinstatement" of the communicator in delayed measurement of opinion change. Journal of Abnormal and Social Psychology, 48, 327-335.

Kahneman, D., & Tversky, A. (1979). Prospect theory: An analysis of decision under risk. Econometrica, 47, 263-291.

Kumkale, G. T., & Albarracin, D. (2004). The sleeper effect in persuasion: A meta-analytic review. Psychology Bulletin, 130, 143-172.

Lariscy, R. A. W., Tinkham, S. F. (1999). The sleeper effect and negative political advertising. Journal of Advertising, 4, 13-30.

Liberman, A., & Chaiken, S. (1992). Defensive processing of personally relevant health messages. Personality and Social Psychology Bulletin, 18, 669-679.

McCaul, K. D., Johnson, R. J., & Rothman, A. J. (2002). The effects of framing and action instruction on whether older adults obtain flu shots. Health Psychology, 21, 624-628.

Metzger, M. J., Flanagin, A. J., Eyal, K., Lemus, D. R., & McCann, R. M. (2003). Credibility for the 21st Century: Integrating perspectives on source, message, and media credibility in the contemporary media environment. In P. J. Kalbfleisch (Ed.), Communication yearbook 27 (pp. 293-335). Mahwah, NJ: Lawrence Erlbaum.

Mullen, B., & Johnson, G. (1990). The psychology of consumer behavior. Hillsdale, NJ: Lawrence Erlbaum Associates.

Pacini, R., & Epstein, S. (1999). The relation of rational and experiential information processing styles to personality, basic beliefs, and the ratio-bias phenomenon. Journal of Personality and Social Psychology, 76, 972-987.

Petty, R. E., & Cacioppo, J. T. (1986). Communication and persuasion: Central and peripheral routes to attitude change. New York: Springer-Verlag.

Petty, R. E., & Wegener, D. T. (1999). The elaboration likelihood model: Current status and controversies. In S. Chaiken & Y. Trope (Eds.), Dual-process theories in social psychology (pp. 41-72), New York: Guilford.

Pratkanis, A. R., Greenwald, A. G., Leippe, M. R., & Baumgardner, M. H. (1988). In search of reliable persuasion effects: III. The sleeper effect is dead. Long live the sleeper effect. Journal of Personality and Social Psychology, 54, 203-218.

Priester, J., Wegener, D., Petty, R., & Fabrigar L. (1999). Examining the psychological process underlying the sleeper effect: The elaboration likelihood model explanation. Media Psychology, 1, 27-48.

Rothman, A. J., & Salovey, P. (1997). Shaping perceptions to motivate healthy behavior: The role of message framing. Psychological Bulletin, 121, 3-19.

Schneider, T. R., Salovey, P., Apanovitch, A. M., Pizarro, J., McCarthy, D., Zullo, J., & Rothman, A. J. (2001). The effects of message framing and ethnic targeting on mammography use among low-income women. Health Psychology, 20, 256-266.

Slater, M. D., & Rouner, D. (1996). Value-affirmative and value-protective processing of alcohol education messages that include statistical evidence or anecdotes. Communication Research, 23, 210-235.

Tchii, E. (2009, Febuary 18). Taiwan's Internet Users Keep Rising: Survey. Taipei Times, p12.

Tversky, A., & Kahneman, D. (1981). The framing of decisions and the psychology of choice. Science, 211, 453-458.

Zhou, S, Pan, C., & Zhong, X. (2008). Effects of rationality and story attributes on perceptions of SARS. In J. J. Powers & X. Xiao (Eds.), The Social Construction of SARS (pp. 223-240) Amsterdam, Netherlands: John Benjamins.

Zillmann, D., & Brosius, H. B. (2000). Exemplification in communication: The influence of case reports on the perception of issues. Mahwah, NJ: Lawrence Erlbaum Associates.

Chia-Hsin Pan

Chinese Culture University

Correspondence to:

Dr. Chia-Hsin Pan

Department of Advertising

Chinese Culture University

55, Hwa-Kang Road, Yang-Ming-Shan, Taipei , Taiwan

Email: pjx@faculty.pccu.edu.tw