Framing effects on disaster preparation: Is negative framing more effective?

The Australasian Journal of Disaster
and Trauma Studies
ISSN:  1174-4707
Volume : 2011-1

Framing effects on disaster preparation:
Is negative framing more effective?

John McClure School of Psychology, Victoria University of Wellington P. O. Box 600 Wellington New Zealand

Chris G Sibley Department of Psychology University of Auckland New Zealand

Abstract

The positive and negative framing of messages about a risk influences people’s intentions to adopt precautions. Framing research has confounded the framing of outcomes (experiencing harm and avoiding harm) with the framing of preventive actions (taking or not taking preventive action). A study manipulated these two factors with judgments about the importance of earthquake preparation and specific preventive actions with citizens (n = 200) in Wellington (New Zealand). Judgments of the general importance of preparedness were affected by outcome frames but not action frames; these judgments were higher with negatively framed outcomes than positive outcomes. Intentions to take specific actions were enhanced by the negative framing of both the outcome and the specific action. These findings clarify which framing messages are most likely to increase preventive actions in relation to risks.

Framing effects on disaster preparation:
Is negative framing more effective?

Communications designed to increase risk-reducing actions vary greatly in their effectiveness, so it is important to know which messages are more effective. A key aspect of these communications is the way that a message is framed (Block & Keller, 1995). Framing effects occur when two messages that are logically equivalent but phrased differently have divergent effects on people’s responses (Linville, Fischer, & Fischhoff, 1993). One major difference in messages is that they can be framed in positive or negative terms. Positive frames refer to the positive consequences from not performing a risky behavior (“if you do not smoke, you are less likely to get cancer”), whereas negative frames cite the negative consequences that result from performing a risky behavior (“if you smoke, you are more likely to get cancer”) (Block & Keller, 1995).

Risky choice framing and prospect theory

Early research on framing presented a choice between two options for reacting to a risk, such as a choice between two treatments for a disease that are framed as gains and losses (Tversky & Kahneman, 1981). The gain frame presents a safe option where a fixed number of people are saved and a risky option with a 1/3 chance that everyone is saved and a 2/3 chance that no-one is saved. The loss frame gives a safe option where a fixed number of people die and a risky option with a 1/3 chance that no one dies and a 2/3 chance that everyone dies. People prefer the fixed option with the ‘gain’ scenario and the risky option with the ‘loss’ scenario ( Tversky & Kahneman 1981). Put another way, if disease treatment is framed in terms of lives saved, people will be more likely to take the ‘sure bet’ and go with the option guaranteeing that a certain number of people are saved. If, however, disease treatment is framed in terms of lives lost, people will be more likely to risk it all, and go with the option offering a chance of saving everyone, and risk the chance that everyone dies.

Kahneman and Tversky (1979) linked the stronger effects of negative frames to prospect theory. This theory of decision making deals with people’s decisions about two prospects or options that involve gains and losses. The theory proposes that people display ‘loss aversion’, in that they weigh losses more heavily than gains. In other words, people act as if losses hurt more than equivalent gains satisfy (Hastie & Dawes, 2001). As a consequence, people’s decisions often reflect how a choice or decision is framed, and in particular, whether an option is framed as a gain or a loss. Options that are framed as losses have a stronger effect on people’s decisions. A different explanation of the stronger effects of negative frames was proposed in Pratto and John’s (1991) evolutionary theory, which claims that negative events such as averting danger had more salience for survival than positive events such as feeding. These two explanations are not incompatible. Druckman and McDermott (2008) suggested that framing effects reflect the different emotions triggered by positive and negative frames, rather than cognitive processes. Kuhberger (1998) noted that with some decisions, it is unclear which option is more risky. For example, having a mammogram entails the risk of detecting breast cancer, but not having a mammogram entails the risk of the cancer developing unchecked. Levin, Schneider, and Gaeth (1998) claimed that findings of framing effects where both outcomes involve equal risks or both are guaranteed offer challenges to prospect theory.

Three types of framing

Levin et al. (1998) attributed the mixed findings on framing to the fact that research has failed to distinguish three types of framing: risky choice framing, attribute framing, and goal framing. Risky choice framing varies the positive or negative frame and the risk in each option, whereas attribute framing manipulates a single characteristic of a target object using a positive or negative frame. For example, the frame may describe the quality of meat as 75% lean (positive) or as 25% fat (negative) (Hasseldine & Hite, 2003; Levin & Gaeth, 1988). Although the two descriptions are logically equivalent, positive framing produces more favorable evaluations, possibly because it evokes more positive associations (Levin et al., 1998). With goal framing, two messages recommend the same behavior to achieve a goal, but vary the valence of the outcome. Thus positive goal framing describes a desirable outcome (e.g., getting healthy) resulting from an action (regular exercise), whereas negative goal framing describes avoiding an undesirable outcome (sickness) by not performing the action (by failing to exercise regularly).

Research on goal framing

Levin et al.’s review (1998) found that, consistent with prospect theory, negative goal framing produced stronger effects. However, later research found inconsistent effects for goal framing (Levin, Gaeth, Evangelista, Albaum, & Schreiber, 2001; Levin, Gaeth, Schreiber and Lauriola, 2002). Researchers attributed this finding to moderating variables such as personal involvement and experience with the issue (Hasseldine & Hite, 2003; Krishnamurthy et al., 2001; Maheswaran & Meyers-Levy, 1990; Rothman & Salovey, 1997), depth of processing (Block & Keller, 1995; Rothman, Salovey, Antone, Keough, & Martin, 1993), or whether the the goal of the target behavior was prevention or detection (Rothman & Salovey, 1997; Rothman et al., 1993) (see also Maule & Villejoubert, 2007).

Several studies, however, have shown clear effects for goal framing. Krishnamurthy, Carter and Blair (2001) examined whether the type of frame (attribute or goal) and the valence of the frame (positive or negative), influenced people’s decisions to discuss a new health treatment with a doctor. The attribute frames read: “Treatment A provides (fails to provide) better results for 50% of the patients with your health condition”. The goal frames read: “By taking (not taking) treatment A, you get a 50% chance of getting better results”. Goal framing affected intentions to talk to the doctor more than attribute framing (see also Pinon & Gambara, 2005).

Several other studies have also shown stronger effects with negative goal framing. Robberson and Rogers (1988) examined how goal framing affected intentions to exercise. They examined whether the framing of an essay about the benefits of regular exercise for health and self-esteem would influence the intentions of unfit female students to begin a regular exercise programme. They found that a negative frame produced stronger intentions to exercise than a positive frame.

Banks, Salovey, Greener, Rothman et al. (1995) examined the effectiveness of negative and positive goal frames for persuading women who had had few mammograms to have a mammogram. The women viewed a short video about breast cancer and mammography, which was either positively framed (entitled The Benefits of Mammography), or negatively framed (The Risks of Neglecting Mammography). Women who had watched the negatively framed video were more likely to have obtained a mammogram 12 months later, a result consistent with prospect theory.

Ganzach and Karsahi (1995) examined the impact of goal framing on credit card usage with low use cardholders. In messages explaining the benefits of using the card, they framed the benefits in terms of gains (positive framing): “paying by Zioncard is not only more convenient, but also much more secure”; or losses (negative framing): “paying by cash is not only less convenient, but also much less secure.” A follow-up six months later showed that negatively framed messages led to participants using their cards more and making higher charges. The results again are consistent with prospect theory.

Is the framing of the action or the outcome more important?

A key issue in this research on goal framing is these studies did not distinguish between the framing of the outcome and the framing of the target behavior. In these studies, the target behavior always had the same valence (positive or negative) as the outcome of the behavior. In Robberson and Rogers’ (1998) exercise frames, for example, the positive frame read: “one of the greatest advantages of becoming involved in a regular exercise programme is the resulting increase in physical stamina and endurance”, and the negative frame read: “one of the greatest disadvantages of not being involved in a regular exercise programme is the resulting decrease in physical stamina and endurance”. With the positive frame, the valence of the target behavior (getting exercise) and the frame of the outcome (increase in physical stamina) are both positive, whereas with the negative frame, the valence of the target behavior (not getting exercise) and the frame of the outcome (decrease in physical stamina) are both negative.

Similarly, Banks et al. (1995) used videos where in one condition, both behavior valence (getting mammography) and the outcome (health benefits) were positive, whereas in the second condition, both were negative (neglecting mammography and health risks). Using a related design, Krishnamurthy et al. (2001) combined positive behavior frames (taking treatment) with a positively framed outcome (the chance of getting better results), and negative behavior frames (not taking treatment) with a negatively worded outcome (giving up a chance of getting better results). All these studies provide valuable results, but their designs do not cross the valence of the target behavior with the framing of the outcome, so it is uncertain which of the two components of the message shaped responses. It is possible that the behavior valence shaped responses more than the outcome frame.

Levin et al. (1998) noted that positive and negative outcomes can be described in different ways, in that good outcomes can be described in terms of obtaining a gain or avoiding a loss, whereas bad outcomes can be described in terms of forgoing a gain or experiencing a loss. Mandel (2001) similarly observed that research has confounded outcome descriptors (number saved or number who die) with the outcome (positive or negative). Lee and Aaker (2004) framed the outcomes of adaptive actions in terms of a promotion focus or a prevention focus, and showed that this framing makes a difference to responses. However, none of these studies addressed the confound between action valence and outcome frame. Krishnamurthy et al. (2001) noted that goal framing with different consequences involves an inherent framing confound because it varies the salience of desirable and undesirable outcomes (p. 390, footnote). They dealt with this issue by varying ‘same consequence’ goal framing but did not vary the action frames.

In contrast, McClure, White and Sibley (2009) dealt with this issue by controlling for potential confounds between action valence and outcome frame, a distinction that was discussed in Levin et al.’s (1998) review but that has not been directly researched. McClure et al. (2009) examined framing in the hazards domain, and focused on preparation for earthquakes. The target actions dealt with earthquake preparation, and the desired outcome was to avoid harm in an earthquake. The experimental manipulation adapted the design used by Robberson and Rogers (1988) with health messages. However, instead of having two versions (positive and negative frames), it had four versions, where the messages crossed the valence of the action of being prepared (i.e. being “well prepared”, or “poorly prepared”) with the framing of the outcome (positive: “surviving unharmed” vs. negative: “experiencing harm”). This design enabled the study to clarify which of these two manipulations is crucial to framing effects. To avoid a confound between action/inaction and desired versus undesired behavior, the positive and negative framing of the target action referred to different levels of preparation (“well prepared” and “poorly prepared”), rather than an action and non-action (“being prepared” or “not being prepared”).

Preparation was defined in terms of general intentions to prepare (e.g., “How important do you think it is to be well prepared for a major earthquake?”) and specific actions that people could perform to prepare for an earthquake (e.g., “obtain a supply of tinned food that could be used in an emergency”). Earthquake preparation was described as the participant taking action, rather than people in general, to ensure that participants would judge the risk to themselves, because people tend to think that others are more at risk than they are across a number of domains, including earthquakes (McKenna & Myers, 1997; Spittal, McClure, Siegert, & Walkey, 2005).

Based on the proposition in prospect theory that losses loom larger than gains, McClure et al. (2009) predicted that negatively framed outcomes would lead to stronger positive effects on intentions to prepare than positively framed outcomes. Because research suggests that negative information in general, not just information about outcomes, is weighted more strongly than positive information (Levin et al., 1998; Meyerowitz & Chaiken, 1987), they predicted that negatively framed actions (poor preparation) would have stronger effects on intentions to prepare than positively framed actions (being well prepared) (Paton, Kelly, Burgelt, & Doherty, 2006). As predicted, negative outcome framing had more effect on general attitudes towards earthquake preparation than a positive outcome frame. However, with specific actions, outcome framing interacted with the framing of the action, in that negative outcome framing increased judgments of the importance of specific preparations more when combined with positive actions (being well prepared). This finding shows that behavioral intentions are influenced not only by the framing of the outcome but also by the framing of the relevant actions.

The present study

A key goal of McClure et al.’s study was to separate the framing of the actions and the framing of the outcome, as previous research has not distinguished these two components of a message. Although this design separated the valence of the action from the framing of the outcome, the wording of the outcome component of the scenarios may have been confusing, because the scenarios referred not only to the different likelihood of the outcome, but also to a greater or lesser chance of the given outcome. In the two conditions where the action and outcome have the opposite valence (e.g., positive action with negative outcome), high likelihood is combined with a low chance of the outcome. For example, the item pairing the negative action with the positive outcome was expressed: “If you are poorly prepared for a major earthquake, you are likely to have a lesser chance of surviving the event unharmed.” In this phrasing, higher likelihood (You are likely) is combined with lesser chance of the positive outcome (surviving unharmed).

In the present study, we eliminated the possible confusion entailed by combining conflicting chance and probability directions in the same statement by simplifying the wording of the outcomes to refer only to likelihoods. We also clarified the wording so that with positive outcomes, rather than referring to the positive condition as ‘survive unharmed’, which is a negation of the word ‘harmed’, we used a completely positive phrase: ‘survive in good shape’. So, for example, the phrasing for ‘Negative behavior with positive outcome’ read: ‘If you are poorly prepared for a major earthquake, you are less likely to survive the event in good shape’. We predicted that messages with negatively framed outcomes and negatively framed actions would lead to higher intentions to prepare than messages with positively framed outcomes and positively framed actions, and than messages with either positive actions and negative outcomes or negative actions and positive outcomes.

Several other minor modifications were made to the questionnaire from McClure et al.’s (2009) study. We used identical anchor points in the general preparedness scale and obtained demographic data on the participants’ length of residence in Wellington. In addition, because the framing manipulation is subtle and could easily be missed by participants wanting to proceed with filling in the rating scale, we added an example of an actual earthquake producing negative outcomes with the same phrasing as the key manipulation sentence.

Method

Participants

Participants were 200 citizens (129 female, 71 male). Time lived in Wellington was: Less than a year, 17%; 1-5 years, 21%; over 5 years, 54%; not given, 8%. Most were recruited in Wellington’s ( New Zealand) central business district during weekday lunch hours. The response rate was 82%, with non-respondents citing time constraints.

Design and materials

The questionnaire began with a pamphlet outlining the benefits of earthquake preparation, adapted from McClure et al.’s (2009) study. This pamphlet began with a short statement emphasizing the likelihood of a major earthquake occurring in Wellington at some point in the future. The pamphlet also listed examples of actions that people could perform to prepare for an earthquake. The pamphlet contained a 2 (outcome framing: positive or negative) x 2 (behavior framing: positive or negative) manipulation. However, the wording was simplified from that in McClure et al.’s study to eliminate any confusion due to conflicting references to likelihood and chance for positive and negative outcomes. The wording for the four conditions thus read as follows:

Positive behavior valence, positive outcome. If you are well prepared for a major earthquake you are more likely to survive the event in good shape.

Negative behavior valence, positive outcome. If you are poorly prepared for a major earthquake, you are less likely to survive the event in good shape.

Positive behavior valence, negative outcome. If you are well prepared for a major earthquake, you are less likely to experience harm in the event.

Negative behavior valence, negative outcome. If you are poorly prepared for a major earthquake, you are more likely to experience harm in the event.

The main manipulation sentence was repeated in bold at the end of the scenario. We added a second sentence presenting the manipulation in terms of an example of a relatively recent earthquake in New Zealand: ‘For example, in the 1987 Edgecombe earthquake, people who attached [did not attach] ornaments, etc. to shelves found that most [few] of them were intact after the earthquake.’

The measures of preparation were based on the items used by McClure et al. (2009) and derived from Spittal, Walkey, McClure, Siegert, and Ballantyne’s (2006) preparation scale. Two general items assessing the ‘importance of preparing’ and the ‘likelihood of preparing’ and were rated on a 7 point Likert scale. There were five specific preparation actions: “Obtain a supply of tinned food that could be used in an emergency”, “put aside spare plastic bags and toilet paper for use as an emergency toilet”, “accumulate enough tools to make minor repairs to the house following a major earthquake”, “obtain a working battery radio”, and “purchase a first aid kit”. Participants rated the importance that they themselves carry out each action on a 7 point Likert scale with the anchor points: 'Not at all important ', and 'extremely important'. Items were averaged to create scale scores assessing attitudes as to the importance of general and specific preparation.

We also added a question measuring participants’ length of residence in Wellington: ‘Less than a year, 1-5 years, over 5 years’.

Procedure

Participants were recruited in public places (e.g., parks and squares) in Wellington city; the researchers checked that the respondents were residents not tourists. Respondents were given a chocolate bar after completing the questionnaire.

Results

Correlations between the variables are shown in Table 1. Length of time spent in Wellington correlated positively with judgments of the general importance of preparation, but not with the specific preparation actions.

Table 1 : Correlations between variables.

n = 200, * p < .05; ** p < .01

A 2 (action framing: positive or negative) x 2 (outcome framing: positive or negative) MANOVA was performed on general attitudes toward earthquake preparation and on the perceived importance of survival actions.

On the general importance items, there was a significant effect for outcome frame, F(1, 196) = 3.73, p = .05, partial η² = .02. Participants who were exposed to a negative outcome frame that referred to experiencing harm rated preparation as more important (M = 5.11, SD = 1.19) than participants exposed to a positive outcome frame that referred to surviving unharmed (M = 4.76, SD = 1.11). There was no interaction of outcome and preparation frames, F(1, 196) = 0.41, p = .52, partial η² < .01, and the main effect of action frame was non-significant, F(1, 196) = 0.12, p = .73, partial η² < .001. These results indicate that the effects of framing on general attitudes toward preparation occurred because of differences in framing of the outcomes of preparation, rather than the preparation itself.

With regard to the perceived importance of survival actions, there was an interaction between outcome and action frames when predicting the perceived importance of survival actions, F(1, 196) = 4.73, p = .03, partial η² = .02 (see Figure 1). Participants rated survival actions as more important when exposed to a negative outcome paired with negatively framed action, (M = 5.60, SD = 1.35), than with a positively framed action (M = 5.11, SD = 1.16), whereas there was no difference when a positively framed outcome was paired with a positively framed action (M = 5.30, SD = 1.20) and a negative framed action (M = 5.04, SD = .96). The main effect for outcome frame was non-significant, F(1, 196) = .42, p = .52, partial η² < .01, as was the main effect for action frame, F(1, 196) = 1.10, p = 30, partial η² < .01.

Figure 1. Ratings of importance of specific preparation actions as a function of valence of preparedness (well or poorly prepared)
and outcome frame (positive or negative outcome).
(Note that ratings of specific importance ranged from a minimum score of 1 to a maximum of 7.)

Discussion

The results clarify the effects of outcome frames and action frames. Consistent with McClure et al.’s (2009) findings, judgments of the general importance of preparation are affected solely by the outcome frame, whereas for survival actions there is an interaction between the action frames and the outcome frame. However, whereas McClure et al. found that messages emphasizing negative outcomes had more beneficial effects on intentions to perform survival actions when paired with a positive action (being well prepared), in the present results the negative outcome has more effect when paired with a negative action (being poorly prepared). Thus negative framing of outcomes effected judgments when paired with negative actions but not positive actions. This result is consistent with the predictions of the majority of researchers who suggest that negative information in general, not just information about outcomes, is weighted more strongly than positive information (Levin et al., 1998; Meyerowitz & Chaiken, 1987). It is possible that McClure et al.’s results were affected by the wording of the scenarios that combined contrasting directions of likelihood and chance for the outcome that might occur: e.g., “you are likely to have a lesser chance of experiencing harm”.

The finding that action framing qualified (i.e., moderated) the effect of outcome framing on specific actions but not general preparedness emphasizes an important distinction between specific and more general intentions to act and prepare. Turner, Nigg, and Paz (1986) found that when citizens focused on specific actions to prepare for hazards, such as getting a battery radio, they were more willing to prepare than when they focused on earthquake preparation in general. Similarly, with framing effects on intergroup attitudes, Sibley, Liu, and Kirkwood (2006) found that the framing of affirmative action policies influenced support for specific aspects of an affirmative action program, but did not alter global attitudes toward resource allocations for minority groups. On this basis, they argued that the framing of a social issue may influence isolated cognitions related to specific aspects of that issue, but is less likely to influence general attitudes, as these attitudes are anchored in social representations that have strong and coherent evaluative associations with societal values. Despite the difference between general and specific attitudes, the results in both cases show higher ratings of intentions to prepare with negative outcome frames.

Research on disaster preparedness has shown that well-designed communications enhance preparedness (e.g., Cowan, McClure, & Wilson, 2002; Flynn, Slovic, Mertz, & Carlisle, 1999; Lindell & Perry, 2000; McClure, Allen, & Walkey, 2001, McClure, Sutton, & Sibley, 2007). However, a limitation that should be acknowledged is that risk communications have little effect on mitigation actions, which are affected more by an internal locus of control (Russell, Goltz, & Bourque, 1995; Spittal, McClure, Walkey, & Siegert, 2008; Spittal, Siegert, McClure, & Walkey, 2002 ; Yoshida & Deyle, 2005). This limitation may also apply to framing effects.

In addition, survey responses indicating participants’ intentions do not always predict subsequent behavior. However, Banks et al. (1995) and Ganzach and Karsahi (1995) followed up their participants several months after their framing experiments, and found that participants who had been exposed to a negative outcome frame were more likely to have performed the recommended behavior than those exposed to a positive frame. It could be expected, therefore, that a follow up of the current study would find similar results, but this remains to be shown. The perceived importance of preparation on its own may not increase earthquake preparation, but it is likely to be one factor in preparation. For people who are more responsive to risk, framing effects on their own may lead to people taking action. However, for others, any change in cognitions due to framing is likely to be overridden by other distractions and pressures over time. We recommend that regular reminders and incentives be included to reinforce the effects produced by framing.

Most research on goal framing has focused on either health (Banks et al., 1995; Krishnamurthy et al., 2001; Maheswaran & Meyers-Levy, 1990; Robberson & Rogers, 1988; Rothman et al., 1993), or marketing (Ganzach & Karsahi, 1995), rather than disaster preparedness. Although preparing for major earthquakes can greatly reduce the resultant harm (Smith, 1996), many vulnerable people remain under-prepared. The current research on framing effects gives a clearer indication about how these communications can be made more effective, and these effects are likely to generalize to other domains such as health and marketing.

The correlations between time in Wellington and preparation intentions show that people who have lived longer in Wellington judged the importance of general preparation higher. It is interesting that the correlation of time in Wellington with specific actions with preparation does not reach significance, which suggests that although people who have lived longer in Wellington have come to recognise the importance of preparation, they do not have stronger intentions to take specific steps. Future research could explore this further, as it is specific steps taken toward preparing for an event, rather than more general attitudes about whether or not it is important to prepare, that is the end result that needs to occur to save lives.

Conclusions

The present study extends previous findings in showing that intentions to act are affected not only by the framing of the outcome but also by the framing of the action. These results highlight the need to take account of the distinction between outcome framing and action framing when designing research on framing, to clarify which part of the message is shaping responses.

With regard to hazard preparation, these findings will be useful for agencies such as Civil Defence who promote preparation for natural disasters. Negative outcome frames led to higher judgments of the general importance of general earthquake preparation and the importance of the participants undertaking specific survival actions. If disaster agencies wish to increase intentions to prepare for earthquakes or other hazards, they are more likely to be effective if they use a message that refers to the harm that may follow inaction and also frame the relevant actions negatively.

References

Banks, S., Salovey, P., Greener, S., Rothman, A., Moyer, A., Beauvais, J., et al. (1995). The effects of message framing on mammography utilization. Health Psychology, 14, 178-184.

Block, L., & Keller, P. (1995). When to accentuate the negative: The effects of perceived efficacy and message framing on intentions to perform a health-related behavior. Journal of Marketing Research, 32, 192-203.

Cowan, J., McClure, J., & Wilson, M. (2002). What a difference a year makes: How immediate and anniversary media reports influence judgments about earthquake. Asian Journal of Social Psychology, 5, 169-185.

Druckman, J. N., & McDermott, R. (2008). Emotion and the framing of risky choice. Political Behavior, 30, 297-321.

Flynn, J., Slovic, P., Mertz, C. K., & Carlisle, C. (1999). Public support for earthquake risk mitigation in Portland, Oregon. Risk Analysis, 19, 205-217.

Ganzach, Y., & Karsahi, N. (1995). Message framing and buying behavior: A field experiment. Journal of Business Research, 32, 11-17.

Hasseldine, J., & Hite, P. (2003). Framing, gender and tax compliance. Journal of Economic Psychology, 24, 517-533.

Hastie R., & Dawes, R. M. (2001). Rational choice in an uncertain world: The psychology of judgment abs decision making. Thousand Oaks: Sage.

Hayton, J. C., Allen, D. G., Scarpello, V. (2004). Factor retention decisions in exploratory factor analysis: A tutorial on parallel analysis. Organizational Research Methods, 7, 191-205.

Judd, C.M., Smith, E.R., & Kidder, L.H. (1991). Research methods in social relations (6 th ed.). Forth Worth, TX: Harcourt Brace Jovanovich.

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

Krishnamurthy, P., Carter, P., & Blair, E. (2001). Attribute framing and goal framing effects in health decisions. Organizational Behavior and Human Decision Processes, 85, 382-399.

Kuhberger, A. (1998). The influence of framing on risky decisions: A meta-analysis. Organizational Behavior and Human Decision Processes, 75, 23-55.

Lee, A. Y., & Aaker, J. L. (2004). Bringing the frame into focus: The influence of regulatory fit on processing fluency and persuasion. Journal of Personality and Social Psychology, 86, 205-218.

Levin, I., & Gaeth, G. (1988). How consumers are affected by the framing of attribute information before and after consuming the product. Journal of Consumer Research, 15, 374-378.

Levin, I. P., Gaeth, G. J., Evangelista, F., Albaum, G., & Schreiber, J. (2001). How positive and negative frames influence the decisions of persons in the United States and Australia. Asia Pacific Journal of Marketing and Logistics, 13, 64-71.

Levin, I. P., Gaeth, G. J., Schreiber, J., & Lauriola, M. (2002). A new look at framing effects: Distribution of effect size, individual differences, and independence of types of effects. Organizational Behavior and Human Decision Processes, 88, 411-429.

Levin, I.P., Schneider, S.L., & Gaeth, G.J. (1998). All frames are not created equal: A typology and critical analysis of framing effects. Organizational Behavior and Human Decision Processes, 76, 149-188.

Lindell, M. K., & Perry, R.W. (2000). Household adjustment to earthquake hazard. Environment and Behavior, 32, 461-501.

Linville, P., Fischer, G., & Fischhoff, B. (1993). AIDS risk perceptions and decision biases. In J. Pryor & G. Reeder (Eds.), The social psychology of HIV

(pp. 5-38). Hillsdale, NJ: Erlbaum.

Maheswaran, D., & Meyers-Levy, J. (1990). The influence of message framing and issue involvement. Journal of Marketing Research, 27, 361-367.

Mandel, D. R. (2001). Gain-loss framing and Choice: Separating outcome formulations from descriptor formulations. Organizational Behavior and Human Decision Processes, 76, 56-76.

Maule, J., & Villejoubert, G. (2997). What lies beneath: Reframing framing effects. Thinking and Reasoning, 13, 25-44.

Meyerowitz, B. E., & Chaiken, S. (1987). The effect of message framing on breast self-examination attitudes, intentions, and behavior. Journal of Social Psychology & Personality, 52, 500-510.

McClure, J. L., Allen, M. W. & Walkey, F. H. (2001). Countering fatalism: Causal information in news reports affects judgments about earthquake damage. Basic and Applied Social Psychology, 23, 109-121.

McClure, J., Sutton, J. M., & Sibley, C. (2007). Listening to reporters or engineers: How different messages about building design affect earthquake fatalism. Journal of Applied Social Psychology, 37, 1956-1973.

McClure, J., White, J., & Sibley, C. G. (2009). Framing effects on preparation intentions: Distinguishing actions and outcomes. Disaster Prevention and Management, 18, 187-199.

McKenna, F., & Myers, L. (1997). Illusory self-assessments. Can they be reduced? British Journal of Psychology, 88, 39-51.

Paton, D. Kelly G., Burgelt, P. T., & Doherty, M. (2006). Preparing for bushfires: understanding intentions. Disaster Prevention and Management, 15, 566-575.

Pinon, A., & Gambara, H. (2005). A meta-analytic review of framing effect: risky, attribute and goal framing. Psithema, 17, 325-331.

Pratto, F., & John, O.P. (1991). Automatic vigilance: The attention-grabbing power of negative social information. Journal of Personality and Social Psychology, 61, 380-391.

Robberson, M., & Rogers, R. (1988). Beyond fear appeals: Negative and positive persuasive appeals to health and self-esteem. Journal of Applied Social Psychology, 18, 277-287.

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

Rothman, A., Salovey, P., Antone, C., Keough, K., & Martin, C. (1993). The influence of message framing on intentions to perform health related behaviors. Journal of Experimental Social Psychology, 29, 408-433.

Russell, L. A., Goltz, J. D., & Bourque, J. B. (1995). Preparedness and hazard mitigation actions before and after two earthquakes. Environment and Behavior, 27, 744-770.

Sibley, C. G., Liu, J. H., & Kirkwood, S. (2006). Toward a social representations theory of attitude change: The effect of message framing on general and specific attitudes toward equality and entitlement. New Zealand Journal of Psychology, 35, 3-13.

Smith, K. (1996). Environmental hazards: Assessing risk and reducing disaster (2 nd ed.). London: Routledge.

Spittal, M. J., McClure, J., Siegert, R. J., & Walkey F. H. (2005). Optimistic bias in relation to preparedness for earthquakes. Australasian Journal of Disaster and Trauma Studies, 2005-1, 1-10.

Spittal, M., McClure, J., Walkey, F., & Siegert, R. (2008). Psychological predictors of earthquake preparation. Environment and Behavior, 40, 798-817.

Spittal, M., Siegert, R. S., McClure, J., & Walkey, F. H. (2002). The spheres of control scale; The identification of a clear replicable three-factor structure. Personality and Individual Differences, 32, 121-131.

Spittal, M. J., Walkey, F. H., McClure, J., Siegert, R. J., & Ballantyne, K. E. (2006). The Earthquake Readiness Scale: The development of a valid and reliable unifactorial measure. Natural Hazards, 39, 15-29.

Turner, R.H., Nigg, J.M., & Paz, D.H. (1986). Waiting for disaster: Earthquake watch in California. Los Angeles: University of California Press.

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

Yoshida, K., & Deyle, R. E. (2005). Determinants of small business mitigation. Natural Hazards Review, 5, 1-12.

Acknowledgements

We thank the Friday Research Group for their many useful comments. This research was funded by a Foundation of Research Science and Technology (FRST) subcontract from GNS Science: CO5X0402.

Copyright