2010 LIBRARY Midfgh; , (Mafia Vtulv University This is to certify that the thesis entitled THE EFFECTS OF MESSAGE FRAMING WITHIN THE STAGES OF CHANGE ON SMOKING CESSATION INTENTIONS AND BEHAVIORS presented by JENNIFER CORNACCHIONE has been accepted towards fulfillment of the requirements for the MA. degree in Communication ,flALi/MU LU“ Mt I‘vm/ Major Professor’s Signature 00/37 Mom Date MSU is an Affirmative Action/Equal Opportunity Employer - —.----a-.-u-—-—-—----—---—.----.-u-u—--o-n-u--o-o-u. en-u-u-o--.—.-.-.—.---—--—-nu.-~ -»&“‘ ,,__‘- PLACE IN RETURN BOX to remove this checkout from your record. To AVOID FINES return on or before date due. MAY BE RECALLED with earlier due date if requested. DATE DUE DATE DUE DATE DUE 5/08 K'lProIIAccltPrelICIRCI‘DatoDuahdd THE EFFECTS OF MESSAGE FRAMING WITHIN THE STAGES OF CHANGE ON SMOKING CESSATION INTENTIONS AND BEHAVIORS By Jennifer Comacchione A THESIS Submitted to Michigan State University in partial fulfillment of the requirements For the degree of MASTER OF ARTS Communication 2010 ABSTRACT THE EFFECTS OF MESSAGE FRAMING WITHIN THE STAGES OF CHANGE ON SMOKING CESSATION INTENTIONS AND BEHAVIORS By Jennifer Comacchione This study examines two commonly used and accepted theoretical models in health communication—the stages of change and message framing—to determine if gain- or loss- framed messages are more effective at getting people to intend to quit smoking depending on their current stage of change (precontemplation, contemplation, or preparation). One hundred forty eight current smokers were exposed to one of four gain or loss frame messages that emphasized the benefits of cessation or the costs of smoking. Message believability, message processing, and stage movement were measured to see if any differences existed as a function of the individual’s base stage of change and message fi'ame exposure. Overall, results indicated that all participants, regardless of stage and frame, engaged in more central than peripheral message processing. Additionally, gain- framed messages were most influential at getting individuals to progress from the contemplation to the preparation stage. Gain-framed messages were most influential for those with lower intentions to quit smoking, and loss-framed messages were more influential for those with higher intentions to quit. ACKNOWLEDGEMENTS First, I would like to thank my advisor, Sandi Smith, and my committee members, Maria Lapinski and Pamela Whitten, for their time, encouragement and guidance. Secondly, I would like to thank Carolyn LaPlante, Samantha Nazione, Chelsea F ristoe, and Lindsay Neuberger for their help and suggestions throughout the process. iii TABLE OF CONTENTS LIST OF TABLES ................................................................................ v LIST OF FIGURES .............................................................................. vi INTRODUCTION ................................................................................. 1 Smoking and Smoking Cessation ........................................................ 3 Stages of Change ................................. - ......................................... 4 Message Framing .......................................................................... 7 Different Frame for Different Stages .................................................... 9 Hypotheses and Research Questions ................................................... 11 METHOD .......................................................................................... 13 Overview and Participants ............................................................. 13 Procedure .................................................................................. 14 Measurement ............................................................................. 15 RESULTS ........................................................................................... l9 Hypothesis One ........................................................................... 19 Hypothesis Two ........................................................................... 19 Research Question 1 ....................................................................... 20 Research Question 2 ....................................................................... 21 Hypothesis Three ........................................... 21 Hypothesis Four ............................................................................ 22 DISCUSSION ........................................................ ‘ ............................... 23 Message processing measures ............................................................ 23 Belief and stage movement ................................................................ 25 Limitations and Directions for Future Research ....................................... 26 CONCLUSION ...................................................................................... 29 APPENDICES ....................................................................................... 37 Appendix A .................................................................................. 37 Appendix B .................................................................................. 43 REFERENCES ...................................................................................... 46 iv LIST OF TABLES Table 1 Table 2 Table 3 Table 4 Table 5 Table 6 Message condition distributed by base stage .............................. 30 Message frame distribution by base stage ................................... 31 Descriptive statistics for dependent variables ............................... 32 Stage Movement by Message Frame and Precontemplation Stage ......33 Stage Movement by Message Frame and Contemplation Stage ......... 34 Stage Movement by Message Frame and Preparation Stage ............. 35 LIST OF FIGURES Figure 1 Coding variables and examples .................................. 36 INTRODUCTION Smoking tobacco reduces the overall health of smokers because it damages almost every organ in the body (Centers for Disease Control & Prevention [CDC], 2004). Smoking is linked to 15 different cancers, such as lung, mouth, and stomach, and accounts for at least 30% of all cancer deaths (American Cancer Society [ACS], 2008). Smoking tobacco also causes coronary heart disease; chronic obstructive pulmonary diseases (COPD), such as emphysema and chronic bronchitis; and puts women at risk for low bone density (CDC, 2004). Furthermore, nicotine, a main ingredient in cigarettes, is so addictive that only four to seven percent of smokers who wish to quit smoking actually succeed the first time without any help (ACS, 2008). Despite the numerous health benefits of cessation, it is often difficult for people to quit smoking. The stages of change model (SCM) has been used successfully to study and facilitate self-initiated and professionally facilitated changes in addictive behaviors, such as smoking cessation (Prochaska, DiClemente, & Norcross, 1992). The model suggests that behavior change involves progression through five stages: precontemplation, contemplation, preparation, action, and maintenance (Prochaska et al., 1992). According to the SCM, behavioral and cognitive goals are specified at each stage, with the overall goal being to progress to the next stage of change (Prochaska et al., 1992). Oftentimes, action-oriented interventions are not successful at getting people to engage in or to avoid specific behaviors because the targeted audience is not in the preparation for action stage. Therefore, a method that works for someone who is ready to change will likely be ineffective for someone who is not aware that a problem even exists. One way to better instantiate the SCM is by using specific message flames in smoking cessation messages to advance individuals through the stages. In fact, Slater (1999) posits that the SCM can serve as a template to synthesize theories of persuasion. Message flames can emphasize the costs (loss-flamed) or benefits (gain-flamed) of engaging in a particular behavior (for a discussion, see Rothman, Bartels, Wlaschin, & Salovey, 2006). Therefore, message flaming should be integrated with the SCM. Salovey, Schneider, and Apanovitch (2002) suggest that gain-flamed messages are more effective for prevention behaviors, such as smoking cessation, and loss-flamed messages are more effective for detection behaviors. However, research also suggests that the effects of flamed messages depend on one’s intention to change (e.g., Moorman & van den Putte, 2008). Integrating message flaming into the SCM may help uncover when either gain- or loss-flamed messages are most effective in getting people to quit smoking at different stages in their intentions and actions to do so. It is predicted here that loss- flarned messages will be more effective for those in the later stages (i.e., preparation), and gain-flamed messages will be more effective for those in the earlier stages (i.e., precontemplation) of change with regard to smoking cessation with stage movement, believability, and message processing as the indicating dependent variables. To further elaborate on the ideas above, this paper will first discuss the issue of smoking and smoking cessation. Then, an overview of the SCM and message flaming literature will be presented. Finally, hypotheses and research questions will be advanced, and their results will be presented and implications will be discussed. Literature Review Smoking & Smoking Cessation From 2000-2004, approximately 443,000 people in the United States died prematurely flom cigarette smoking and exposure to secondhand smoke. Furthermore, during a four-year period (2000-2004), 5.1 million annual lives were lost for both men and women combined in the United States, excluding lives lost flom secondhand smoke (CDC, 2008). The tar in cigarettes is composed of over 4,000 chemicals, 60 of which are known to cause cancer (ACS, 2008). Additionally, smoking accounts for 90% of COPD- related deaths, the fourth leading cause of death in the US. (National Heart, Lung, & Blood Institute [NHLBI], 2009). Not only does smoking directly impact the smokers themselves, but it also has severe consequences for others. Secondhand smokers inhale the nicotine and toxic chemicals that smokers do (ACS, 2008), including more than 50 that are known to cause cancer (CDC, 2006). Exposure to secondhand smoke can cause the same negative effects that smoking does, such as heart disease, lung cancer, and decreased lung functioning (CDC, 2006). Not only is smoking detrimental to individuals, but it also accounts for over $150 billion each year in health care costs and lost productivity (American Lung Association [ALA], 2004). The benefits of cessation begin immediately (ALA, 2009). Twenty minutes afier quitting, one’s blood pressure drops. Eight hours afier cessation, carbon monoxide in the blood drops to a normal level, and oxygen levels also return to their normal state. Furthermore, the excess risk of heart disease caused by smoking is reduced by half. Long-term benefits, such as a reduced risk of stroke, also exist (ALA, 2009). Because of the detrimental consequences of smoking, and the immediate and enduring benefits associated with cessation, it is imperative for individuals to quit smoking to prolong the length and quality of their lives. Cessation is actually a process that begins when one considers the act of quitting. Stages of Change The SCM explains that behavior change involves multiple stages, each of which constitutes different behavioral and cognitive needs (Prochaska et al., 1992). The SCM is comprised of five stages. A person is in the precontemplation stage when there is no intention to change behavior in the next six months, and he/she is unaware of the problem. Oftentimes, people in this stage are forced into therapy, and attempts flequently fail because they are not yet ready to change. The second stage, contemplation, is characterized by individuals who are aware a problem exists, are seriously thinking about addressing it, but have not yet committed to action. Individuals in this stage weigh the pros and cons of the problem and solution. Preparation is the stage where individuals plan to take action in the next month, or have taken action and failed in the past year. For example, people who cutback on the number of cigarettes smoked each day are in this stage. Action, the fourth stage, occurs when individuals engage in, and are committed to, overt behavioral change. The problem behavior must be absent for one day to six months to classify one in this stage. Finally, the maintenance stage is characterized by relapse prevention. Behavior change is not linear, but instead a continual process. Individuals must remain flee of the addictive behavior and engage in a competing healthy behavior for at least six months to be in the maintenance stage. The SCM suggests that people do not move linearly through the stages; instead, progression often takes the form of a spiral, with individuals regressing and relapsing, having to continue through stage progression (Prochaska et al., 1992). The treatment or course of action an individual should take depends on his/her stage placement. Perz, DiClemente, and Carbonari (1996) found that smoking cessation success was a fimction of completing stage-specific tasks. They found that more cognitive and emotional activities, such as consciousness-raising, led to attitude change for those in the precontemplation and contemplation stages. Behavioral activities, such as committing to the behavior and creating new alternatives, should emerge for those in the later stages of change, resulting in increased smoking cessation success rates (Perz et al., 1996). Therefore, interventions relying on behavior change are not effective for those in earlier stages because awareness is a crucial component for long-term change. Furthermore, interventions for individuals ready to change their behavior are unsuccessful when the focus is on consciousness-raising and not on overt behavioral change (Prochaska et al., 1992). Different messages should thus be sent to individuals in different stages. Maibach and Cotton (1995) put forward that increasing knowledge and awareness is critical to move individuals flom precontemplation to contemplation. They further suggest that positive social and self-evaluative consequences are important to move individuals to the preparation stage flom contemplation. This helps individuals to see themselves as capable of behavior change. To move flom preparation to action entails encouraging people to identify and plan solutions to obstacles and messages to increase self-efficacy and reinforcement of appropriate behaviors. Finally, movement flom action to maintenance requires more reinforcement and self-efficacy, including the long-term positive consequences of behavior change (Maibach & Cotton, 1995). The effects of messages communicated to individuals vary depending on their current stage. Other studies have also emphasized the importance of tailored messages for individuals in different stages (e. g., Campbell et al., 1994). As the findings above indicate, it is important to tailor messages to obtain the desired outcomes. Slater (1999) posits that messages that increase attention to the issue are imperative to get individuals to move flom the precontemplation to the contemplation stage. Additionally, the arguments made in the message are not typically important for those low in involvement. Peripheral cues, such as message credibility, the number of arguments, and message quality, are most effective at increasing attention for this population. Slater suggests emphasizing the efficacy of responses to a threat and increasing the salience of the threat to be crucial message features in getting individuals to move to the preparation stage flom contemplation. Finally, Slater states that using quantitative evidence, reinforcing beliefs, and increasing attitude accessibility is useful in getting individuals to move flom the preparation to action stage. All of these message features are important driving forces in changing attitudes and behaviors at the different stages in the SCM. The SCM has been successfill in predicting and encouraging smoking cessation (DiClemente et al., 1991), weight loss (Sarkin, Johnson, Prochaska, & Prochaska, 2001), and condom use (Stark et al., 1998), among other topics. However, despite the abundance of literature that supports the transtheoretical model of tailored messages and audience segmentation, little research has examined whether health messages should emphasize the costs or the benefits of engaging in a particular health behavior for each stage of change. Message Framing As discussed previously, individuals in different stages of change need different intervention approaches, including tailored messages. Therefore, the effects of message flames may not be uniform across stages even though Salovey et a1. (2002) claim that gain-flamed messages should be superior for all prevention messages such as smoking cessation. Message flames can either emphasize the costs (loss-flamed) or benefits (gain-flamed) of engaging in a particular behavior. Each flame has been operationalized in two ways (Rothman & Salovey, 1997). Gain-flamed messages can describe the behavior outcome as attaining a desirable outcome or avoiding an undesirable outcome, and loss-flamed messages can describe the behavior outcome as failing to attain a desirable outcome or attaining an undesirable outcome (Salovey et al., 2002). O’Keefe and Jensen (2007) fin'ther articulated this by discussing kernel states. A kernel state is the basic, root state mentioned, and it can be stated in terms of something to be attained/avoided or as a desirable/undesirable outcome. For example, with the kernel state ‘cancer’, an example of a loss flame message is “If you continue smoking, you will be at an increased risk for developing cancer,” which is an undesirable outcome to be avoided, and an example of a gain flame message is “If you quit smoking, you will decrease your risk for developing cancer,” which is a desirable outcome to be attained. Kahneman and Tversky’s (1979) prospect theory has driven most message flaming research. Prospect theory suggests that when the potential outcomes are risky, people prefer loss-flamed messages, but when the potential outcomes are less risky people prefer gain-flamed messages (Salovey et al., 2002; Tversky & Kahneman, 1981). Therefore, Salovey and colleagues (Rothman etal., 2006; Salovey et al., 2002) suggest that gain-flamed messages influence prevention behaviors, such as smoking cessation and physical exercise, and loss-flamed messages influence detection behaviors, such as breast self-examinations and HIV testing. Many studies have supported this detection/prevention hypothesis. Loss-flamed messages have been found to be more effective for detection behaviors such as mammography (Banks et al., 1995; Schneider, Salovey, Apanovitch, et al., 2001), and gain-flamed messages have been found to be more effective for prevention behaviors such as sunscreen use (Rothman, Salovey, Antone, et al., 1993) and smoking cessation (Schneider, Salovey, Pallonen, et al., 2001). Furthermore, respondents’ own memorable messages about breast cancer were mostly gain-flamed and associated with prevention behaviors (LaPlante, Smith, Kotowski, & Nazione, 2009). However, results of three recent meta-analyses do not completely support the detection/prevention hypothesis (O’Keefe & Jensen, 2006; O’Keefe & Jensen, 2007; O’Keefe & Jensen, 2009). The authors found gain-framed messages to be more persuasive for prevention behaviors, but this difference was attributed to dental hygiene behaviors. Results of these meta-analyses indicate that the prevention/detection hypothesis cannot be supported. In fact, as a result of these meta-analytic findings, Latirner, Salovey, and Rothman (2007) stated that, “This next generation of flaming research is the result of investigators working to refine flaming postulates. . .looking beyond categories such as prevention vs. detection” (p. 646). Thus, other variables can inform when each type of message is most effective for prevention behaviors. A possible direction that can reconcile these findings in message flaming research is to investigate the effectiveness of gain and loss flame messages in different stages of the SCM. As discussed earlier, different tactics have been found to be most effective at getting individuals to progress through the stages. Because the SCM emphasizes a tailored approach to behavior change, different flames may be more effective at raising issue awareness, enhancing response efficacy, or reinforcing smoking cessation attitudes depending on the individual’s current stage. Although little research in this area has been conducted, similar findings support the integration of flaming and SCM. Cho and Salmon (2006) studied the effects fear appeals had on skin cancer messages for each stage in the SCM. The fear appeal used in their study can be characterized as a loss- flamed message because it highlighted the severe consequences of skin cancer and the participants’ vulnerability to getting skin cancer. Results of this study suggest that participants in the precontemplation stage exhibited significantly less favorable attitudes and weaker intention toward adhering to the recommended behaviors compared to individuals in the other stages. The fear appeal also induced defensive avoidance and hopelessness for those in the precontemplation stage compared to the other stages, indicating that a negative (loss) flame may be ineffective for individuals in the precontemplation stage. Wong and McMurray (2002) tested the effects of gain- and loss-flamed messages for individuals in difl‘erent stages of change for smoking cessation. However, their study had some limitations, and also differs flom the current study. The researchers only had two groups, collapsing the stages into “no intention to quit smoking” (precontemplation stage) and “intention to quit smoking” (contemplation and preparation stages). Furthermore, their study was tested on a non-US. population with 70 participants. It is important to extend their findings by specifically examining each stage, not just overall intentions. The authors found that those with no intentions to quit smoking generated more message-specific thoughts for gain-flamed messages compared to loss-flamed messages. Wong and McMurray (2002) suggested that all smokers are aware of the negative consequences of smoking, so positively flamed information is seen as new information that requires scrutiny, especially for those with no intention to quit. One explanation for these findings rests on the assumption that messages containing different flames should be processed differently at different stages of change. Theories of message processing, such as the elaboration likelihood model (ELM), explain that there are two routes to persuasion: central and peripheral. Central processing is characterized by high involvement and extensive issue-relevant thinking, and peripheral processing is characterized by peripheral cues guiding behavior and attitudes, such as message credibility and length (Petty & Cacioppo, 1986). Both motivation and ability must be high for central processing to occur (for a review, see O’Keefe, 2002, ch. 6). Research suggests that highly involved individuals are more persuaded by negative flames, and positively flamed messages are more effective for low involvement individuals (Donovan & J alleh, 1999). Moorman and van den Putte (2008) found similar results for smoking cessation. They found that the effects of message flaming on smoking cessation depended on processing depth. The authors suggest that people with low intentions to quit smoking are not motivated to process the message (low involvement), so a positive flame is more effective, and people with high intentions to quit smoking are motivated (high involvement), so a negative flame is more effective (Moorman & van 10 den Putte, 2008). Other studies also support this outcome with findings suggesting that negative flames were more persuasive than positive flames for individuals who engaged in deep processing for health behaviors, and that positive flames were more persuasive when heuristic (peripheral) processing occurred (Block & Keller, 1995; Meyers-Levy & Maheswaran, 2004). Another study found that those who processed centrally responded more favorably to a loss-flamed message, and those who processed heuristically (peripherally) responded more favorably to a gain-flamed message (Steward, Schneider, Pizarro, & Salovey, 2003). Message believability is a component of the ELM that could potentially influence the effects of message flame on varying base stage placements for individuals. Although typically viewed as a peripheral process, Zuckerman and Chaiken (1998) suggest that when individuals centrally process messages, peripheral cues are also noticed, which may influence compliance to the message. Therefore, different flames may enhance message belief through the process of enhancing issue awareness or reinforcing anti-smoking attitudes. Because individuals in the precontemplation stage are often unaware of their problem, and likely have low issue involvement, it is probable that their level of motivation to process is low. Because of their low intentions, involvement, and motivation to process, individuals in the precontemplation stage should be more persuaded by a gain-flamed message, and engage in peripheral message processing, thus enhancing message believability. H1: Individuals in the precontemplation stage will engage in more peripheral message processing than central message processing for both gain- and loss- flamed messages. 11 H2: Individuals in the precontemplation stage will be more likely to a) report higher belief in the message, and b) move to the contemplation stage with a gain- flarned message than with a loss-flamed message. Contemplators may also have low motivation to process because they have yet to commit to any behavioral change, but it is also possible that they have high motivation because they are intending to change. Therefore, it is not possible to speculate about message belief. Additionally, contemplators’ level of intent to change and level of issue involvement cannot be speculated, thus the type of message flame that would be most effective is unknown and research questions are posed here. RQl: Do individuals in the contemplation stage engage in more central message processing or peripheral message processing for gain- and loss-flamed messages? RQ2: Will individuals in the contemplation stage be more likely to a) report belief in the message, and b) move to the preparation stage with a gain or loss flamed message? Individuals in the preparation stage likely have high issue involvement with smoking cessation, and thus are motivated to process the message. As a result, they should be more persuaded by a loss-flamed message, and engage in more central message processing because of their high intentions to change, high issue involvement, and motivation to process. Because of the predicted depth of processing, it is also expected that message believability will also be enhanced with a loss-flamed message. H3: Individuals in the preparation stage will engage in more central message processing than peripheral message processing for both gain- and loss-flamed messages. 12 H4: Individuals in the preparation stage will be more likely to a) report higher belief in the message, and b) move to the action stage with a loss-flamed message than with a gain-flamed message. Method Overview and Participants This study sought to persuade individuals to intend to quit smoking tobacco by determining when certain messages are most effective by integrating message flaming and the SCM. Specifically, this study examined individuals in the precontemplation, contemplation, and preparation stages of the SCM, along with both gain and loss flamed messages. A 3 (stage placement [precontemplation, contemplation, preparation]) x 4 (message flame [two gain flame and two loss flame instantiations]) between subjects factorial design was used, along with a pre- and post-test to determine message processing and stage movement. Two hundred participants enrolled in communication courses at a large Midwestern university completed the initial study survey, and received course credit for participation. Out of the 200, 37 did not return for time two of the study, 12 were removed flom the analyses due to non-smoker status, and three were excluded for irrelevant/explicit data. Thus, the total sample of participants who completed the two online survey questionnaires was 148 current smokers. Current smoking was defined as individuals who have smoked 100 cigarettes in their life, and smoke every day or some days (CDC, 1994; Healthy People 2010). Additionally, open-ended data were coded to determine whether or not there was a clear indication that the participant was a smoker. Open-ended data that explicitly referred to the participant’s smoking behavior was coded 13 as a clear indication that he/she was a smoker. Cohen’s Kappa reliability on smoking status was .94. To be included in the analysis, the participant had to be coded as positive on one of the three smoking measures. The 148 study participants were mostly female (61.5%) and Caucasian (84.5%). Other participants indicated their race/ethnicity as being Asian (5.4%), Afiican American (4.1%), Hispanic (3.4%), and Native American (0.7%). Two percent of the participants indicated their race/ethnicity as ‘other’. Over 30% of the participants were fleshmen (37.8%), followed by seniors (29.1%), sophomores (20.9%), and juniors (12.2%). Procedure Participants first indicated their smoking status by responding to two questions: “Have you smoked at 100 cigarettes during your lifetime?” and “Do you smoke at least one day a week on average?” Participants then indicated their current stage placement with a single-item measure that assessed their readiness to quit smoking. After completing this initial measure, all participants were randomly assigned to one of four gain- or loss-flamed smoking cessation messages that were similar in length. After reading the message, participants replied to questions indicating whether or not the message emphasized the benefits of cessation or the costs of smoking. Message believability was then measured. To measure level of message processing, participants completed a thought listing task (Cacioppo & Petty, 1981). Peripheral message processing was operationalized in terms of believability and the number of peripheral thoughts per participant. Individuals also indicated their intent to change after reading the message via a similar stages of change measure that will be described in detail below. 14 One week later, participants were invited to return to the study via the web to determine stage movement outcomes by selecting one of five phrases that best described them. Participants were also asked whether they had smoked within the last week. To allow the researcher to track the data between the two surveys while keeping the study confidential, participants were asked to report the name of their first best fiiend and their first pet. The web survey is located in Appendix A. As an incentive, two $50 Amazon.com gift cards were provided via a random drawing. Measurement of Variables Independent variables. Base stage. Participants were initially classified into a base stage of change with a single item SCM measure, which indicated their current readiness to quit smoking (Prochaska et al., 1992). Those indicating no intent to quit smoking tobacco were categorized as in the precontemplation stage; individuals intending to quit within the next six months were classified in the contemplation stage; and individuals intending to quit smoking within the next month were labeled in the preparation stage. Message frame. The four message conditions are presented in Appendix B. These flamed messages focused on the health effects of smoking and the benefits of cessation. Two different message instantiations of each message type (gain and loss flames) were used to prevent any confounds that may occur when only one gain and one loss flame message are used (Jackson & Jacobs, 1983). All participants in each of the three stages were randomly sent to one of the four message conditions. Participant distribution within each message condition is located in Table 1. Condition one (gain flame message one) was composed of 35 participants total. Condition two (gain flame 15 message two) was composed of 38 students total. Condition three (loss flame message one) was composed of 33 students total, and condition four (loss flame message two) was composed of 42 participants. Independent sample t-tests determined that there were no significant differences between the two gain-flame messages and the two loss-flame messages on all variables. For gain-flamed messages (conditions 1 and 2), independent sample t-tests revealed non— significant differences for central message processing, t (71) = -l.15, p = .25; peripheral message processing, t (71) = 1.10, p = .28; believability, t (71) = -0.21, p = .83; and stage movement to the Contemplation stage, t (70) = -0.08, p = .94; preparation stage, t (70) = - 0.02, p = .99; and action stage, t (70) = -0.02, p = .99. For individuals exposed to the loss-flamed messages (conditions 3 and 4), independent sample t-tests revealed non- significant differences for central message processing, t (73) = 0.62, p = .50; peripheral message processing, t (73) = -1.33, p '= .19; message believability, t (73) = 1.09, p = .28; and stage movement to the contemplation stage, t (69) = 0.20, p = .84; preparation stage, t (69) = 0.19, p = .85; and the action stage, t (69) = 0.19, p = .85. Therefore message conditions were collapsed by flame resulting in a total of 73 participants exposed to gain- flamed messages and 75 exposed to loss-flamed messages, and the distribution by base stage is located in Table 2. Data were therefore analyzed by message flame (gain vs. loss) rather than by each specific message. Dependent variables. Benefits vs. costs. A manipulation check was conducted to detemrine whether the messages were operating as planned. A paired samples t-test between those exposed to the loss-flame message conditions rated messages as significantly emphasizing the costs 16 (M = 6.03, SD = 0.96) more than the benefits (M = 4.60, SD = 1.62), p < .001. A paired samples t-test also revealed that those exposed gain-flame message conditions rated those messages as emphasizing the benefits (M = 6.23, SD = 0.63) significantly more than the costs (M = 4.84, SD = 1.15), p < .001. Therefore, the operationalization of gain and loss flame messages was successful. Message believability. Message believability was assessed using a four-item scale by Lapinski, Maloney, Braz, and Shulman (in preparation). Sample items included: “The message seems reasonable,” and “The information in the message is believable.” The message believability scale was found to be reliable, or = .82. Message processing. To assess message processing via a thought listing task (Cacioppo & Petty, 1981), participants were asked to list the thoughts they had when reading the message. Two researchers unitized the thoughts for 20% of the participants. The unit of analysis was a complete thought unit. Guetzkow’s U was .037 (Guetzkow, 1950) where a smaller figure indicates successful unitization. Coding reliability was then assessed using Cohen’s Kappa (Cohen, 1960). The units were coded by two researchers for 17% of the responses. First, the unit was coded as being relevant or irrelevant. If irrelevant, the unit was not coded. If relevant, the unit was then coded as a central or peripheral thought. A central thought was one that was specifically related to the content of the message, such as, “I was surprised to find out that after 2 hours of not smoking your blood pressure decreases,” and “My hair won’t smell if I quit.” A peripheral thought was one that referenced any features of the message itself outside of the content, including credibility and message quality. Examples of peripheral thoughts are, “[The message] didn’t have any evidence,” and “The article 17 needs more concrete arguments.” Overall, the dependent variables assessed here are: number of thoughts, central processing, and peripheral processing. Thought coding variables and examples can be found in Figure 1. Cohen’s Kappa ranged flom .82 to 1 on all thought variables. Thoughts were analyzed in SPSS as the sum of each type of unit per participant, such as the sum of peripheral thought units for each participant. Intent to change/stage movement. To measure intent to change, participants selected one statement that best described them according to the SCM (Prochaska et al., 1992). This was done after reading the message, and indicated any stage movement that may have occurred. This variable was measured differently than the initial base stage measure that was described as an independent variable. Instead of having three responses to choose flom, the participants had five options, with two new options. The two new options were, “I plan to quit smoking within the next week,” and “I plan to quit smoking today.” The statements were expanded and more specific to determine if movements that occurred were more micro in nature. This same measurement was also used when participants were invited back to complete a survey to measure stage movement one week later. Smoking status during past week. Smoking status was measured as a nominal level variable. Respondents answered yes or no to the question, “Have you smoked within the last week?” at time two of the study. Descriptive statistics for dependent variables are located in Table 3. 18 Results Precontemplation Hypothesis 1. Hypothesis one predicted that individuals in the precontemplation stage would engage in more peripheral rather than central message processing for both gain- and loss-flamed messages. A paired-samples t-test was run on the sum of peripheral and central thought units for those in the precontemplation stage, and the findings were non-significant, t (37) = 1.82, p = .08. In fact, the results trended in the opposite direction with those in the precontemplation stage tending to engage in more central processing (M = 2.05, SD = 1.63) than peripheral processing (M = 1.34, SD = 1.30) across both gain and loss-flame messages. Additional analyses were conducted to determine whether individuals engaged in more central or peripheral processing depending on the specific message flame (gain vs. loss) rather than the flames combined (gain and loss). A paired samples t-test between peripheral and central thoughts and gain-flamed messages indicated that precontemplation individuals engaged in significantly more central (M = 2.58, SD = 1.61) than peripheral (M = 1.47, SD = 1.31) processing, t (18) = -2.1 1, p = .05 when reading gain-flamed messages, counter to the prediction. No significant differences were present on the paired samples t-test between thought unit sums for central (M = 1.53, SD = 1.50) and peripheral (M = 1.26, SD = 1.28) processing and loss-flamed messages, t (18) = - 0.46, p = .65. Hypothesis 2. To determine believability of the message and stage movement after reading the message, hypothesis three predicted that those in the precontemplation stage would a) be more likely to report higher belief in the message, and b) move to the 19 contemplation stage with a gain-flamed message than a loss-flamed message. To examine message believability, an independent samples t-test was conducted between message flame and believability. The data were not significant, t (36) = 1.59, p = .12, and the trend was toward loss (M= 5.8, SD = 1.05) rather than gain (M= 5.31, SD = 0.89) flamed messages being more believable. To test stage movement at time two to the contemplation stage based on message flame, chi-square was used. The result was not significant, 2’2 (1, n = 38) = .13, p = .50. Thus, the data were not consistent with hypothesis 2. Overall stage movement at times one and two of the study are located in Table 4. Contemplation Research question 1. To determine the level of message processing that occurred for individuals in the contemplation stage, which was research question one, another paired-samples t-test was conducted between message flame and the sum of peripheral and central thought units. Results indicated that those in the contemplation stage engaged in significantly more central (M = 2.63, SD = 2.02) than peripheral (M = .98, SD = 1.23) message processing regardless of message flame, t (61) = 5.05, p < .001. Thus, the data reveal that individuals in the contemplation stage of their readiness to quit smoking engaged in more central processing than peripheral processing across message flame. Additional analyses were again conducted to determine if the level of message processing varied depending on the specific message flame. A paired samples t-test between thought unit sums and loss message flame indicated that individuals in the contemplation stage engaged in significantly more central (M = 2.52, SD = 2.19) rather than peripheral (M = 1.00, SD = 1.21) message processing for loss-flamed messages, t 20 (26) = -2.90, p = .008. Additionally, those in the contemplation stage engaged in significantly more central (M = 2.71, SD = 1.90) than peripheral (M = 0.97, SD = 1.27) processing for gain-flamed messages, t (34) = -4. 16, p < .001. Thus, individuals in the contemplation staged engaged in more central processing regardless of message flame. Research question 2. Research question two was proposed to examine a) message processing, and b) stage movement to the preparation stage for individuals in the contemplation stage. An independent sample t-test was conducted to determine message believability based on message flame. The results trended toward more individuals having higher belief with gain-flamed (M = 5.71, SD = 0.65) instead of loss-flamed (M = 5.36, SD = 0.92) messages. However, these results are not significant, t (60) = -1.73, p =. .09. Stage movement flom contemplation to the preparation stage based on message flame was analyzed with chi square. However, the results were non-significant, ,1; (1, n = 61) = .32, p = .38, indicating that message flame did not have an effect on stage movement. Thus, the data provide no definitive evidence for research question two. Overall stage movement at times one and two of the study are located in Table 5. Preparation Hypothesis 3. To test hypothesis three, which predicted that individuals in the preparation stage would engage in more central processing than peripheral for both gain— and loss-flamed messages, a paired-samples t-test was run between central and peripheral thought unit sums and message flame. The results yielded a significant difference, t (47) = 3.49, p = .001 with data indicating that participants in the preparation stage engaged in 21 more central message processing (M = 2.27, SD = 1.84) than peripheral message processing (M = .88, SD = 1.63). Therefore, the data were consistent with hypothesis 3. Additional analyses were conducted to determine whether individuals engaged in more central or peripheral processing depending on the specific message flame. Results of a paired samples t-test indicated that those in the preparation stage engaged in significantly more central (M = 2.48, SD = 1.90) message processing than peripheral (M = 0.90, SD = 1.61) for loss-flamed messages, t (28) = -3.03, p = .01. No significant differences emerged between central (M = 1.84, SD = 1.80) and peripheral (M = 0.84, SD = 1.71) processing and gain-flamed messages, t (18) = -1.56, p = .14. Hypothesis 4. Hypothesis four predicted that individuals in the preparation stage would a) be more likely to report higher belief in the message, and b) move to the action stage with a loss-flamed message than a gain-flamed message. An independent samples t-test was conducted to determine message believability. The data indicate individuals reporting higher belief in loss-flamed messages (M = 5.84, SD = 0.24) than gain-flamed messages (M = 5.78, SD = .73), but the relationship is non-significant, t (46) = 0.23, p = .66. To determine movement to the action stage for individuals in the preparation stage based on message flame, chi-square was used. The results were non-significant, [2 (1, n = 44) = 1.13, p = .24. Thus, message flame had no significant effect on movement to the action stage, and the data did not support hypothesis four. Overall stage movement at times one and two of the study are located in Table 4. Post hoc analysis. At time two of the study, participants were asked to report whether or not they had smoked within the last week. To test the association between 22 smoking within the last week and message flame, chi-square was used. The result was not significant, ,1; (1, n = 148) = 1.41, p = .16. Thus, message flame exposure had no influence on smoking behavior during the week between parts 1 and 2 of this study. Discussion The aim of this study was to combine two commonly used and accepted theoretical models in health communication to help researchers better understand how to get individuals to intend to quit smoking tobacco. Specifically, this study sought to determine whether individuals with varying levels of readiness to quit smoking would be more receptive to and influenced by gain or loss flamed messages. Overview and Explanation of Findings First, this study examined the level of message processing that individuals engaged in based on their current stage placement and message condition. Hypothesis one predicted that participants in the precontemplation stage would engage in more peripheral than central message processing regardless of message flame. The data were not consistent with the prediction. In fact, the trend of results was in the opposite direction, with precontemplation individuals engaging in more central message processing, especially for gain-flamed messages where the difference was significant. This significant finding that more central thoughts were provided for gain-flamed messages rather than for loss-flamed messages supports previous research that indicates that positively flamed messages are more effective for low involvement individuals (Donovan & J alleh, 1998), and supports findings by Wong and McMurray (2002) that those with no intention to quit smoking produced more message-specific thoughts for gain-flamed messages. The gain-flamed messages may have facilitated more motivation 23 to process new information that may not always be communicated to smokers, who are more accustomed to hearing about the negative effects of smoking (Wong & McMurray, 2002) Hypothesis three predicted that individuals in the preparation stage of change would engage in more central than peripheral message processing. The data were consistent with this prediction. Those who indicated their intention to quit within the next 30 days engaged in significantly more central message processing across message flames, especially loss-flamed messages. This finding supports previous research that indicates that those who have high intentions to quit smoking are motivated to process the message (Block & Keller, 1995; Meyers-Levy & Maheswaran, 2004; Moorman & van den Putte, 2008). Finally, research question one was posed to determine the level of message processing for individuals in the contemplation stage. The results indicate that those who intended to quit smoking within the next six months engaged in significantly more central than peripheral message processing for all message flame conditions. This finding also helps support previous research that has combined the contemplation and preparation stages of change in data analysis (e.g., Wong & McMurray, 2002). It is possible that these two stages are more similar than has originally been believed to be the case. Across all stages, central message processing occurred. For those with lower intention to quit smoking (precontemplation), more central processing occurred when exposed to gain-flamed messages. Additionally, those with higher intention to quit smoking (preparation) engaged in more central processing for loss-flamed messages. 24 Those in the contemplation stage engaged in central processing for both gain and loss flame messages. All of these findings were significant. Second, this study sought to examine message believability and stage movement as a function of an individual’s current stage of change and message flame exposure. Hypothesis two predicted that participants in the precontemplation stage would have higher belief in the message and move to the contemplation stage with a gain-flamed message. However, data were not consistent with this prediction. The data indicated that participants tended to have higher belief with a loss-flamed message, but this finding was non-significant. Additionally, no significant results were found for stage movement, although data indicated more stage movement with a loss-flamed message. This is interesting given the earlier findings that those in this stage of change engaged in more central processing with gain-flamed messages, and should be further explored. Researchers have suggested that gain-flamed messages are most effective for prevention behaviors, such as smoking cessation (Schneider, Salovey, Pallonen, et al., 2001). However, this finding provides no definite results for O’Keefe and Jensen’s (2006; 2007; 2009) meta-analyses. Hypothesis four predicted that participants in the preparation stage would have higher belief in the message and move toward the action stage with a 1033- rather than a gain-flamed message. The data did not support the hypothesis. Previous research indicates that negative flames are more effective for individuals with high issue involvement (Moorman & van den Putte, 2008), but these findings provide no definitive support for previous research. 25 Finally, research question two was posed to determine whether message belief was higher or stage movement would occur more flequently with a gain or loss-flamed message for individuals in the contemplation stage. Although non-significant, higher belief was indicated with gain- rather than loss-flamed messages. Additionally, individuals reported moving to the action stage with a gain-flamed message significantly more often than with a loss-flamed message. These findings point to the possibility that gain-flamed messages are most effective for individuals planning to quit smoking within the next six months. Although central message processing occurred for both gain and loss-flamed messages, these findings support the idea that gain-flamed messages may be most effective in inducing attitude and behavior change related to smoking for those in the contemplation stage. Limitations and Directions for Future Work This study is not without limitations. First, stage movement was not uniformly measured. Three stages were present at the base measure, but five were possible post- message exposure. This was done to examine more finite changes in the participant’s intention to change. Measuring stage movement is a limitation across all SCM studies (W einstein, Rothman, & Sutton, 1998). Some studies use single-item measurements as in the current study, while others examine post-stage movement using a 40-item measure of the Processes of Change (DiClemente & Prochaska, 1985; Prochaska, Velicer, DiClemente, & Fava, 1988). Specifically, Weinstein and colleagues (1998) state that, “the specific strategies and beliefs that cause them to move flom one stage to the next are currently not well identified” (p. 293), and that “One would expect small, naturally occurring shifts along a continuum to be more common than large shifts, so movement to 26 nearby pseudostages would be more likely than movement to distant pseudostages” (p. 295). Thus, future research should aim to better understand the micro-stage movements that occur. Another limitation of this study was the use of only text-based persuasive materials, which limits the generalization of the messages to other contexts, such as visual and verbal. Future studies should test other message formats. Furthermore, additional variables could have been measured that may have highlighted or explained some of the findings. First, the role self-efficacy plays in this research context in individuals’ intentions to quit or to seek help to quit smoking could be studied. Perceived costs and benefits of quitting could be examined as well. Further research should also study the specific constructs of involvement and motivation at all levels of readiness to change to determine if they influence level of processing. - Other variables of interest might include nicotine dependence, previous number of quit attempts, and smoking history (Wong & Cappella, 2009). Although a specific population was sampled (smokers), this study only used college-aged participants. Despite the abundance of studies using this population, research has indicated that college-aged students have less-stable attitudes and are more easily influenced than adults, among other differences (Sears, 1986). Thus, different populations may result in different findings. This study should be replicated with different individuals, including older adults and varying ethnicities. Specific to the ELM, all participants may have had high involvement and motivation to process the message because the topic of smoking is issue relevant, even if there was no intention to quit smoking. Although this suggestion contradicts findings by 27 Moorman and van den Putte (2008), whose research suggests that those with no intention to quit are not motivated to process the message, this message topic may be perceived as relevant, thus enhancing issue involvement and motivation to process the message. Therefore, researchers should more carefully examine what, if any, influences message processing has on researching findings. Sample size is another limitation of this study. As seen in Tables 1 and 2, the distribution of participants per condition based on base stage is not uniform, with more individuals in the contemplation stage than precontemplation and preparation. Additionally, as seen in Table 2, very few analyses were conducted with more than 25 participants in each cell. This is especially important for those findings that were nearing significance. Small sample sizes make it less likely that the null hypothesis can be rejected because it increases the standard error (Cohen, 1988). Thus, a larger sample size would increase the validity of the results. By increasing the current sample size, researchers would begin to uncover the intricacies of when gain and loss flames are most effective. In addition to the suggestions already mentioned, a direction for future studies is to examine how risk perception influences belief, level of message processing, and stage movement based on message flames. Smokers may have different perceptions of the messages they read. For example, some people may feel vulnerable to the health risks discussed in the message while other may have engaged in reactance or minimized the risks. According to Weinstein (2001 ), “. . .smoker’s acknowledgement of smoking risks depends on the way in which risk judgments are assessed” (p. 95). Another possibility for future research is to examine the prevention/detection hypothesis within the stages of 28 change flamework. Finally, this study could be replicated on less addictive behaviors, such as HIV/STI testing, nutrition, condom negotiation, and health screening tests. Conclusion This study combined the SCM and message flaming to persuade people to intend to quit smoking tobacco. Overall, it was found that gain-flamed messages were most influential at getting individuals in the contemplation stage to progress toward the preparation stage. Central processing also occurred more for those in the precontemplation and contemplation stages who were exposed to gain-flamed messages. For those intending to quit within the next 30 days, loss-flamed messages were seen to be more effective. This study highlighted some key variables that should further be examined in SCM and message flaming research. 29 Table 1 Message Condition Distribution by Base Stage Message Precontemplation Contemplation Preparation Total Gain Frame 1 8 17 10 35 Gain Frame 2 11 18 9 38 Loss Frame 1 6 12 15 33 Loss Frame 2 13 15 14 42 Total 38 62 48 148 30 Table 2 Message Frame Distribution by Base Stage Message Precontemplation Contemplation Preparation Total Gain Frame 19 35 19 73 Loss Frame 19 27 29 75 Total 38 62 48 148 3 1 Table 3 Descriptive Statistics for Dependent Variables Precontemplation Contemplation Preparation Gain Loss Gain Loss Gain Loss Bfmdm‘ M so M so M so M so M so M so Range anables Benefits 6.21 0.50 5.15 1.52 6.28 0.51 4.00 1.68 6.12 0.91 4.78 1.49 1-7 Costs 4.37 1.30 6.1 1 1.02 4.87 1.03 5.90 0.97 5.28 1.08 6.09 1.03 1-7 Believability 5.30 0.89 5.80 1.05 5.71 0.65 5.36 0.92 5.78 0.73 5.84 1.29 1-7 Central 2.58 1.61 1.53 1.50 2.71 1.90 2.52 2.19 1.84 1.80 2.48 1.90 0-9 Peripheral 1.47 1.31 1.26 1.28 0.97 1.27 1.00 1.21 0.84 1.71 0.90 1.61 0-7 Smoking status 0.74 0.45 0.74 0.45 0.66 0.48 0.89 0.32 0.58 0.51 0.62 0.49 0-1 Note. Range is for entire possible range of the variable. Central and peripheral thought units were open-ended. 32 Table 4 Stage Movement by Message Frame and Precontemplation Base Stage Time One Time Two Gain Loss Gain Loss Movement 11 % n % n % n % 0 13 68.4 14 73.7 14 73.7 10 52.6 1 5 26.3 3 15.8 5 26.3 6 31.6 2 1 5.3 1 5.3 -- -- 1 5.3 3 -- -- 1 5.3 -- -- l 5.3 4 -- -- -- -- -- -- l 5.3 Total 19 100.0 19 100.0 19 100.0 19 100.0 Note. Possible range of movement was 0 to 5 33 Table 5 Stage Movement by Message Frame and Contemplation Base Stage Time One Time Two Gain Loss Gain Loss Movement n % n % n % n % -1 -- -- 1 3.7 -- -- l 3.7 0 18 52.9 19 70.4 15 42.9 16 59.3 1 10 29.4 5 18.5 14 40.0 7 25.9 2 4 1 1.8 2 7.4 5.7 3 1 1.1 3 2 5.9 -- -- 4 1 1.4 -- -- Total 34 100.0 27 100.0 35 100.0 27 100.0 Note. Possible range of movement was -1 to 4 34 Table 6 Stage Movement by Message Frame and Preparation Base Stage Time One Time Two Gain Loss Gain Loss Movement n % n % n % n % -2 -- -- 1 3.4 -- -- -- -- -1 -- -- 3 10.3 1 5.3 3 10.3 0 12 63.2 10 34.5 6 31.6 5 17.2 1 3 15.8 5 17.2 6 31.6 9 31.0 2 4 21.1 10 34.5 6 31.6 12 41.4 Total 19 100 29 100.0 19 100.0 29 100.0 Note. Possible range of movement was -2 to 3 35 Figure 1 Coding Variables &. Examples Coding variable Example Central positive “The message made me want to quit.” Central negative “I was also thinking that this information probably won't change my smoking habits.” Central neutral “1 already knew smoking was bad for me.” Peripheral positive “It [the message] was very straight forwar .” Peripheral negative “The article needs more concrete arguments.” Peripheral neutral “It [the message] was oriented towards the family” Smoking positive “A good smoke goes great with a good dark beer.” Smoking negative “I can’t afford financially to be smoking anymore.” Smoking neutral “I rarely smoke.” 36 APPENDIX A MESSAGES ABOUT SMOKING WEB SURVEY Messages about Smoking Tobacco You are being asked to participate in a research project that investigates messages about smoking. You must be at least 18 years of age to participate in this study and be a current smoker. If you choose to participate in this study, you will read a message about smoking, and then complete a questionnaire regarding your thoughts and feelings about the message. One week later, you will log back on and complete a very short questionnaire. At the end of the study, two participating students will be randomly drawn to each receive one $50 gift card flom Amazon.com Participation in this study is voluntary, and will take 30 minutes or less and will occur at two times one week apart. Declining to participate in this study will not affect your status as a college student. You do not have to participate and you do not have to answer any questions that make you uncomfortable. You also have the right to withdraw flom the interview at anytime. No identifying information will be collected flom you during this survey. Your answers will be kept confidential and your privacy will be protected to the maximum extent allowable by law. You will be asked to report the name of your first pet and first best fliend as a way for the survey to remain anonymous and for the researchers to track data flom both surveys. There are no known benefits or risk for participants. If you have any questions about this study, such as scientific issues, how to do any part of it, or to report an injury, please contact one of the researchers. Jennifer Comacchione Comaccl @msu.edu 5 1 7-432-5 124 438 Communication Arts and Sciences Building East Lansing, MI 48824-1212 Dr. Sandi W. Smith smiths@msu.edu 517-353-3715 573A Communication Arts and Sciences Building East Lansing, MI 48824—1212 If you have any questions or concerns about your role and rights as a research participant, or would like to register a complaint about this study, you may contact, anonymously if you wish, the Michigan State University’s Human Research Protection Office at 517-355- 2180, Fax 517-432-4503, or e-mail irb@m_su.edu or regular mail at 202 Olds Hall, MSU, East Lansing, MI 48824. 37 Checking the box below indicates your voluntary agreement to participate in this study. Yes No Verification that participant is a smoker Have you smoked 100 cigarettes in your life? Yes No Do you smoke at least one day a week on average? _Yes __No Stage Placement Please select the one response that fits you best: . I intend to quit smoking tobacco within the next 30 days. I intend to quit smoking tobacco within the next six months. I have no intention to quit smoking tobacco in the next six months. Random assignment to messages. Please read the following message. When you are finished reading, you-will be asked to answer questions regarding this message. 2' 8 8 8 I- i“ 'é’ it a»; so ggl g, E“ i M M N 0 a a a m a z m a < The message emphasizes the costs of smoking. 1 3 5 6 7 The message shows good things that will happen if I 1 3 5 6 7 quit smoking. The message tells about good outcomes for quitting l 2 3 4 5 6 7 smoking. The message states bad outcomes of smoking. 1 2 3 4 5 6 7 The message suggests benefits for quitting smoking. 1 2 3 The message shows bad things that will happen if I 1 2 3 5 6 continue to smoke. I found the message to be believable. 1 2 3 4 5 6 7 This message presented credible information. 1 2 . 3 4 5 6 7 This message seems reasonable. 1 2 3 4 5 6 7 The information in the message is believable. 1 2 3 4 5 6 7 38 List the thoughts you had while reading the message. List the emotions you felt while reading the message. Emotion 39 Future Intent to Smoke Please select the phrase the fits you best: I plan to quit smoking today. I plan to quit smoking within the next week. I plan to quit smoking in the next month. I plan to quit smoking within the next six months I plan to continue smoking Demographic Information Sex: Male Female Year in College: Freshman Sophomore Junior Senior Graduate student Race/Ethnicity: Caucasian African American Hispanic Asian Native American Pacific Islander Other (list) What is the first name of your first best fiiend? What is the name of your first pet? One-weekfollow-up Survey You are being asked back to complete a short survey to complete this research- investigation, which investigates messages about smoking. Participation is voluntary, and 40 this survey should take 10 minutes or less. Declining to participate in this study will not affect your status as a college student. You do not have to participate and you do not have to answer any questions that make you uncomfortable. You also have the right to withdraw flom the interview at anytime. No identifying information will be collected flom you during this survey. Your answers will be kept confidential and your privacy will be protected to the maximum extent allowable by law. You will be asked to report the name of your first pet and first best fliend as a way for the survey to remain anonymous and for the researchers to track data flom both surveys. There are no known benefits or risk for participants. If you have any questions about this study, such as scientific issues, how to do any part of it, or to report an injury, please contact one of the researchers. Jennifer Comacchione Comaccl @msu.edu East Lansing, MI 48824-1212 Dr. Sandi W. Smith smiths@msu.edu 517-353-3715 573A Communication Arts and Sciences Building East Lansing, MI 48824-1212 If you have any questions or concerns about your role and rights as a research participant, or would like to register a complaint about this study, you may contact, anonymously if you wish, the Michigan State University’s Human Research Protection Office at 517-355- 2180, Fax 517-432-4503, or e-mail irb@m_su_.edu or regular mail at 202 Olds Hall, MSU, East Lansing, MI 48824. Have you smoked in the last Yes No week? Please select the phrase that fitiyou best. I plan to quit smoking today. I plan to quit smoking within the next week. I plan to quit smoking in the next month. I plan to quit smoking within the next six months I plan to continue smoking In the last week, have you talked to others about smoking? 41 Yes No If yes, who did you talk to, and what did you talk about? In the last week, have you had any thoughts about smoking? Yes No If yes, what were your thoughts? What is the first name of your first best fiiend? What is the name of your first pet? Thank you for participating in this research project. If you wish to receive the full study design and results when data have been collected and analyzed, contact Jennifer Comacchione, corn_accl muedu. To be entered in the random drawing for one of the $50 Amazon.com gift cards, send an email to msustudydrawing@gmail.com 42 APPENDIX B MESSAGE FRAME CONDITIONS First Set Gain: Condition 1 If you quit smoking, you are going to live longer. Quitting smoking will help to keep your organs healthy. If you quit smoking, benefits begin immediately. Twenty minutes after you quit, your blood pressure drops. Two weeks after quitting, your lung filnctioning increases. Your coughing and shortness of breath diminish. Long-term benefits include a reduced risk of heart disease and stroke. Quitting smoking lowers your risk for lung and other types of cancers, and your risk for developing cancer declines with the number of years you have abstained flom smoking. Quitting smoking can prevent bad breath, wrinkles, and premature aging. It can also decrease the stains on your teeth and gums. Quitting smoking will save you over $2000 each year if you smoke a pack of cigarettes a day. When you quit smoking you take control of your health. You save your money. You look healthy. You feel healthy. Quitting smoking is the best thing to do for your health. Word count: 160 Loss: Condition 3 If you continue to smoke, you are more likely to die sooner. Smoking cigarettes damages almost every organ in your body. It is linked to 15 different types of cancer, putting you at risk of getting mouth, stomach, lung, and liver cancers, among many others. Smoking also causes emphysema and chronic bronchitis, and increases your risk for heart disease 43 and stroke. Smoking cigarettes affects you immediately, such as you developing a chronic cough, an increase in heart rate and blood pressure, and airway irritability. Continuing to smoke will result in you having bad breath, wrinkles, and premature aging. Smoking also stains your teeth and gums. If you continue to smoke, you will spend over $2000 each year if you smoke a pack of cigarettes a day. If you continue smoking you are not taking control of your health. You waste your money. You look unhealthy. You feel unhealthy. Continuing to smoke is the worst thing to do for your health. Word count: 160 Second Set Gain: Condition 2 If you quit smoking, it can positively impact the quality of life for you and your friends and family. Quitting smoking will alleviate the harmful effects of second-hand smoking on those close to you. Quitting smoking may help you flom becoming ill, which will prevent your family and friends flom having to suffer. If you quit smoking, you will haVe more energy to do things with your friends and family. Quitting smoking before age 35 allows you to live as long and as healthy as someone who has never smoked. If you quit smoking, you will have more money for family events. If you determine your own reasons for quitting, you will have a better chance of success. Decide for sure that you want to quit. Think positively about how you will overcome obstacles and succeed. List all of the benefits for quitting smoking that you determined. Every night before going to bed, repeat one of those benefits 3 times. Word count: 160 Loss: Condition 4 If you continue smoking, it will hurt the quality of life for you, your fliends, and family. Smoking compromises the health of those close to you. Secondhand smoke is as harmful to them as smoking is to you. If you continue smoking, you could become ill, causing your family and friends to suffer. You will not have the energy to do things with your fliends and family. You will also have less money for family events. If you continue to smoke after age 35, you will not live as long or be as healthy as someone who has never smoked or quit. If you do not determine your reasons for quitting, you will have a greater chance of failure. Decide for sure that you want to quit. Try to avoid negative thoughts about how difficult it might be. List all of the costs of continuing to smoke. Every night before going to bed, repeat one of those costs 3 times. Word count: 160 45 REFERENCES American Cancer Society (2008). Cigarette smoking. Retrieved April 1, 2009, flom http://www.cancer.org/docroot/PED/content/PED_1 0_2X_Cigarette_Smoking.asp ?sitearea=PED American Lung Association (2004). Smoking cessation resources fact sheet. Retrieved April 5, 2009, flom http://www.lungusa.org/site/c.deUK900E/b.44456/k.7B2A/Smoking_Cessation _Resources_Fact_Sheet.htm American Lung Association (2009). Quit smoking: Benefits. Retrieved April 5, 2009, flom http://www.lungusa.org/site/c.deUK900E/b.33568/k.CEBD/Benefits.htm Banks, S.M., Salovey, P., Greener, S., Rothman, A.J., Moyer, A., Beauvais, J ., & Epel, E. (1995). The effects of message flaming on mammography utilization. Health Psychology, 14, 178-184. Block, L.G., & Keller, RA. (1995). When to accentuate the negative: The effects of perceived efficacy and message flaming on intentions to perform a health-related behavior. Journal of Marketing Research, 32, 192-203. Cacioppo, J .T., & Petty, RE. (1981). Social psychological procedures for cognitive response assessment: The thought-listing technique. In T.V. Merluzzi, C.R. Glass, & M. Genest (Eds), Cognitive Assessment, pp. 309-342. New York: Guilford Press. Campbell, M.K., DeVellis, B.M., Stretcher, V.J., Ammerman, A.S., Devellis, R.F., & Sandler, RS. (1994). Improving dietary behavior: The effectiveness of tailored messages in primary care settings. American Journal of Public Health, 84, 783- 787. CDC (1994). Cigarette smoking among adults—United States, 1992, and changes in the definition of current cigarette smoking. Morbidity and Mortality Weekly Report, 43, 342-346. CDC (2004). 2004 Surgeon General ’s report: The health consequences of smoking. Retrieved April, 1, 2009, flom http://www.cdc.gov/tobacco/data_statistics/sgr/sgr_2004/highlights/9.htm. CDC (2006). Fact sheet: Secondhand smoke. Retrieved April 6, 2009, flom http://www.cdc.gov/tobacco/data_statistics/fact_sheets/secondhand_smoke/secon dhandsrnokehtm 46 ‘ CDC (2008). Smoking-attributable mortality, years of potential life lost, and productivity losses: United States, 2000-2004. Morbidity & Mortality Weekly Report, 5 7, 1226-1 228. Cho, H., & Salmon, CT. (2006). Fear appeals for individuals in different stages of change: Intended and unintended effects and implications on public health campaigns. Health Communication, 20, 91-99. Cohen, J. A. (1960). A coefficient of agreement for nominal scales. Education and Psychological Measurement, 20, 37-46. Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2"d ed.). Hillsdale, NJ: Lawrence Erlbaum Associates, Inc. DiClemente, C.O., & Prochaska, J .O. (1985). Processes and stages of change: Coping and competence in smoking behavior change. In S. Shiffman & T.A. Wills (Eds.), Coping and substance abuse (pp. 319-343). New York, NY: Academic Press. DiClemente, C.O., Prochaska, J .O., Fairhurst, S.K., Velicer, W.F., Velasquez, M.M., & Rossi, J .S. (1991). The process of smoking cessation: An analysis of precontemplation, contemplation, and preparation stages of change. Journal of Consulting & Clinical Psychology, 59, 295-304. Donovan, R.J., & J alleh, G. (1999). Positively versus negatively flamed product attributes: The influence of involvement. Psychology & Marketing, 16, 613-630. Guetzkow, H. (1950). Unitizing and categorizing problems in coding qualitative data. Journal of Clinical Psychology, 6, 47-58. Healthy People 2010. Operational definition: Tobacco use. Retrieved September 9, 2009, flom @z/lftpcdcgov/pub/Health Statistics/NCHSIDatasets/DATAZOJ 0/Focusarea27/ 02701 aflf Jackson, S., & Jacobs, S. (1983). Generalizing about messages: Suggestions for design and analysis of experiments. Human Communication Research, 9, 169-191. Kahneman, D. & Tversky, A. (1979). Prospect Theory. An analysis of decision under risk. Econometrica, 47, 263-291. Lapinski, M.K., Maloney, E.K., Braz, M.E., & Shulman, H.C. (in preparation). Modifying perceptions of descriptive norms and behaviors through messaging: A field experiment. LaPlante, C., Smith, S.W., Kotowski, M., & Nazione, S. (2009). The effects of flaming of memorable breast cancer messages on engagement in detection or prevention 47 A behaviors. Paper presented at the Annual Conference of the National Communication Association, Chicago, 2009. Latirner, A.E., Salovey, P., & Rothman, A.J. (2007). The effectiveness of gain-flamed messages for encouraging disease prevention behavior: Is all hope lost? Journal of Health Communication, 12, 645-649. Maibach, E.W., & Cotton, D. (1995). Moving people to behavior change: A staged social cognitive approach to message design. In E. Maibach & R.L. Parrot (Eds.), Designing health messages: Approaches fiom communication theory and public health practice (pp. 41-61). Thousand Oaks, CA: Sage. Meyers-Levy, J ., & Maheswaran, D. (2004). Exploring message framing outcomes: When systematic, heuristic, or both types of processing occur. Journal of Consumer Psychology, 1 4, 159-167. Moorman, M., & van den Putte, B. (2008). The influence of message flaming, intention to quit smoking, and nicotine dependence on the persuasiveness of smoking cessation messages. Addictive Behaviors, 33, 1267-1275. NHLBI (2009). It has a name: COPD. Retrieved April 6, 2009, flom http://www.nhlbi.nih.gov/health/public/lungcopdl O’Keefe, DJ. (2002). Elaboration likelihood model. In DJ. O’Keefe (Ed.), Persuasion theory and research, 2’” ed., (pp. 137-167). Thousand Oaks, CA: Sage. O’Keefe, D. J ., & Jensen, J. D. (2006). The advantages of compliance or the disadvantages of noncompliance? A meta-analytic review of the relative persuasive effectiveness of gain-flamed and loss-flamed messages. Communication Yearbook, 30, 1-43. O'Keefe, D. J. & Jensen, J. D. (2007). The relative persuasiveness of gain-flamed and loss-flamed messages for encouraging disease prevention behaviors: A meta- analytic review. Journal of Health Communication, 12, 623-644. O'Keefe, D. J. & Jensen, J. D. (2009). The relative persuasiveness of gain-flamed and loss-flamed messages for encouraging disease detection behaviors: A meta- analytic review. Journal of Communication, 5 9, 296-316. Perz, C.A., DiClemente, C.C., & Carbonari, J .P. (1996). Doing the right thing at the right time? The interaction of stages and processes of change in successful smoking cessation. Health Psychology, 15, 462-468. Petty, R.E., & Cacioppo, J .T. (1986). Communication and persuasion: Central and peripheral routes to attitude change. New York: Springer-Verlag. 48 Prochaska, J .O., Velicer, W.F., DiClemente, C.C., Fava, J. (1988). Measuring processes of change: Applications to the cessation of smoking. Journal of Consulting and Clinical Psychology, 56, 520-528. Prochaska, J .O., DiClemente, C.C., & Norcross, J .C. (1992). In search of how people change: Applications to addictive behaviors. American Psychologist, 4 7, 1102- 1114. Rothrnan, A. J., Bartels, R. D., Wlaschin, J ., & Salovey, P. (2006). The strategic use of gain- and loss-flamed messages to promote healthy behavior: How theory can inform practice. Journal of Communication, 56, $202-$221. Rothman, A.J., & Salovey, P. (1997). Shaping perceptions to motivate healthy behavior: The role of message flaming. Psychological Bulletin, 121 , 3-19. Rothman, A.J., Salovey, P., Antone, C., Keough, K., & Martin, CK. (1993). The influence of message flaming on intentions to perform health behaviors. Journal of Experimental Social Psychology, 29, 408-433. Salovey, P., Schneider, T.R., & Apanovitch, A.M. (2002). Message flaming in the prevention and early detection of illness. In J .P. Dillard & M. Pfau (Eds.), The Persuasion Handbook, (pp. 391—406). Thousand Oaks, CA: Sage. Sarkin, J .A., Johnson, S.S., Prochaska, J .O., & Prochaska, J .M. (2001). Applying the _ transtheoretical model to regular moderate exercise in an overweight population: Validation of a stages of change measure. Preventive Medicine, 33, 462-469. Schneider, T.R., Salovey, P., Apanovitch, A.M., Pizarro, J ., McCarthy, D., Zullo, J ., & Rothman, A.J. (2001). The effects of message flaming and ethnic targeting on mammography use among low-income women. Health Psychology, 20, 256-266. Schneider, T.R., Salovey, P., Pallonen, U., Mundorf, N., Smith, N.F., & Steward, W. (2001). Visual and auditory message flaming effects on tobacco smoking. Journal of Applied Social Psychology, 31, 667-682. Sears, DO. (1986). College sophomores in the laboratory: Influences of narrow data base on social psychology’s view of human nature. Journal of Personality and Social Psychology, 51, 515-530. Slater, MD. (1999). Integrating application of media effects, persuasion, and behavior change theories to communication campaigns: A stages-of-change flamework. Health Communication, 11, 335-354. Stark, M.J., Tesselaar, H.M., O’Connell, A.A., Person, B., Galavotti, C., Cohen, A., & Walls, C. (1998). Psychological factors associated with the stages of change for condom use among women at risk for HIV and STDs: Implications for 49 intervention development. Journal of Consulting and Clinical Psychology, 66, 967-978. Steward, W.T., Schneider, T.R., Pizarro, J., & Salovey, P. (2003). Need for cognition moderates responses to flamed smoking-cessation messages. Journal of Applied Social Psychology, 33 , 2439-2464. Tversky, A., & Kahneman, D. (1981). The flaming of decisions and the psychology of choice. Science, 211, 453-458. Weinstein, ND. (2001). Smokers’ recognition of their vulnerability to harm. In P. Slovic (Ed.) Smoking: Risk, perception, and policy, pp. 81-96. Thousand Oaks, CA: Sage. Weinstein, N.D., Rothman, A.J., & Sutton, SR. (1998). Stage theories of health behavior: Conceptual and methodological issues. Health Psychology, 1 7, 290-299. Wong, N.C.H., & Cappella, J .N. (2009). Antismoking threat and efficacy appeals: Effects on smoking cessation intentions for smokers with low and high readiness to quit. Journal of Applied Communication Research, 3 7, 1-20. Wong, C.O., & McMurray, NE (2002). Framing communication: Communicating the antismoking message effectively to all smokersJournal of Community Psychology, 30, 433-447. Zuckerman, A., & Chaiken, S. (1998). A heuristic-systematic processing analysis of the effectiveness of product warning labels. Psychology & Marketing, 15, 621-642. 50 STATE V TY LBRARIES 7908 ...|..|.m| Illlllllllllll 3 03063 1293