Risky business: Risk information and the moderating effect of message frame and past behaviour on women’s perceptions of the Human Papillomavirus vaccine

Abstract

The effect of response cost information, message framing and past behaviour on women’s coping appraisal and motivation to be vaccinated against the Human Papillomavirus (HPV) were investigated using a 2 Frame × 2 Response Cost × 2 Pap Status design. Women (N = 286) read one of four messages about the vaccine. Women who received high-risk information perceived the vaccine as having higher response cost and were less motivated to be vaccinated compared to women who received low-risk information. The deleterious effects of risk information on specific aspects of women’s coping appraisal may be mitigated by appropriately framed messages.

Keywords

health communication HPV vaccines message framing protection motivation theory (PMT)risk

Introduction

According to protection motivation theory (PMT; Rogers, 1975) information about a health threat spurs two cognitive processes, a threat appraisal process and a coping appraisal process. In the threat appraisal process people consider the severity of the threat (perceived severity) and their vulnerability to this threat (perceived vulnerability). In the coping appraisal process, people consider the effectiveness of an adaptive response to alleviate the threat (response efficacy) and the negative implications of enacting the response (response cost). In this process people also consider their ability to enact successfully the adaptive response (self-efficacy). Together, threat and coping appraisals determine an individual’s intentions (protection motivation) to engage in the adaptive response.

PMT captures many of the cognitions affecting women’s decision to obtain the HPV vaccine (Brewer and Fazekas, 2007). Specifically, women’s perceptions of the vaccine’s effectiveness and side-effects and women’s perceived likelihood of contracting HPV all predict whether they get vaccinated (Brewer and Fazekas, 2007). The purpose of this study was to examine factors affecting women’s threat and coping appraisals of the HPV vaccine and ultimately their protection motivation for HPV vaccination. The three factors that we examined included: the content of messages about the vaccine; the frame of messages about the vaccine; and the sexual health status of women receiving information about the vaccine. We examined these factors specifically to gain insight into theory-based strategies to improve the content and the delivery of information associated with the release of new and potentially controversial vaccines in the future.

The initial release of the HPV vaccine was surrounded in controversy. The vaccine was conveyed in some media reports as being unsafe, leading many people to perceive the vaccine as having health risks (i.e. high response cost; Abdelmutti and Hoffman-Goetz, 2009). It has been speculated that these false reports affected women’s willingness to be vaccinated (Abdelmutti and Hoffman-Goetz, 2009). Using an experimental manipulation of message content, we examined the impact of information exaggerating the negative effects of the vaccine on women’s coping appraisal and protection motivation to be vaccinated. We hypothesized that women who receive inaccurate information conveying the response cost of the vaccine as risky, comparable to the information conveyed in some media reports, would have a less favourable coping appraisal of the vaccine and would be less motivated to be vaccinated than women who received accurate information presenting the vaccine as low risk (Hypothesis 1).

In addition to manipulating information about the response cost of the HPV vaccine, we also varied the frame of the message components encouraging vaccination. These components either emphasized the benefits of being vaccinated (gain-framed) or the consequences of not being vaccinated (loss-framed). Varying the tone of a message in this way can systematically affect aspects of an individual’s coping appraisal and motivation to engage protective health behaviours (e.g. Meyerowitz and Chaiken, 1987). Indeed, Rothman and Salovey (1997) suggest that when individuals are considering engaging in a health behaviour with risky or uncertain outcomes, loss-framed messages are most persuasive and that when individuals are considering engaging in a health behaviour with low-risk or certain outcomes, gain-framed messages are most persuasive. For example, loss-framed messages have been shown to be particularly effective for behaviours such as prostate examinations (Cherubini et al., 2005) and breast self-exams (Williams et al., 2001), while gain-framed messages are especially effective in promoting behaviours such as physical activity (Latimer et al., 2008) and the use of sunscreen (Hevey et al., 2010).

Thus, in accordance with message framing postulates (Rothman and Salovey, 1997), an effective strategy to motivate vaccination among women who were exposed to negative media reports conveying the vaccine as having risky outcomes might have been to provide them with supplemental loss-framed information emphasizing the consequences of not being vaccinated. In the vaccine literature there generally is an advantage for loss-framed messages (Abhyankar et al., 2008; Gerend and Shepherd, 2007; Gerend and Sias, 2009; Gerend et al., 2008; McCaul et al., 2002). This advantage has been attributed to people perceiving the response cost of vaccination as risky (Abhyankar et al., 2008). Within our experimental manipulation of message content and frame we hypothesized that information conveying the response cost of the HPV vaccine as risky would have fewer deleterious effects on women’s coping appraisals and motivation to be vaccinated when accompanied by loss-framed information encouraging vaccination than with gain-framed information (Hypothesis 2a). As a corollary, we also hypothesized that information conveying the response cost of the HPV vaccine as low risk and outlining the benefits of being vaccinated would positively affect women’s coping appraisals and motivation to be vaccinated compared to a low-risk, loss-framed message (Hypothesis 2b).

We also examined individual characteristics of the message recipients as a third factor influencing responses to HPV-related information. There is a growing body of evidence documenting individual difference variables as factors that moderate the effects of framed messages. For example, gender has been shown to moderate the effect of framed messages encouraging smoking cessation (Toll et al., 2008), income has been found to moderate the effect of framed messages promoting breast screening (Consedine et al., 2007) and body consciousness as well as gender have been shown to moderate the effect of framed message promoting the use of sunscreen (Hevey et al., 2010).

The majority of information about the HPV vaccine has been conveyed through mass media and does not tailor to individual characteristics of the message recipient (Caskey et al., 2009). However, Gerend and Shepherd (2007) demonstrated that women’s sexual history moderates the effects of HPV vaccination information. They found that women who engaged in risky sexual behaviours were persuaded to be vaccinated especially when they received loss- versus gain-frame messages. In the current study we considered the influence of women’s Pap test history, a potential global indicator of women’s sexual activity and health status, on their response to high- and low- risk gain- and loss-framed messages. Because women who are Pap tested are more likely to be sexually active (Lee et al., 1998) we hypothesized that women who had been Pap tested would have stronger perceptions of HPV vulnerability than women who had not been Pap tested (Hypothesis 3a). Moreover, we predicted that women who had not been Pap tested (i.e. had low perceptions of vulnerability) would hold especially positive coping appraisals and would be motivated to get vaccinated after reading low-risk, gain-framed messages because of the congruency between their perceptions of vulnerability to HPV and the tone of the information being presented about the vaccine (Hypothesis 3b). We did not have a specific hypothesis for women who had been Pap-tested because we anticipated that multiple types of messages might persuade these women; some may be especially responsive to high-risk, loss-framed information in accordance with message framing postulates, while others may discount the high-risk information recognizing it as inaccurate and thus be persuaded by low-risk, gain-framed messages.

Method

Participants

A convenience sample of 286 women attending a mid-size, Canadian university participated in the study in exchange for a chance to win gift certificates. The study was approved by the university’s general research ethics board and all participants consented to participate in the study. Women could not have been previously vaccinated against HPV. The mean age was 21.77±2.29 years. The majority of the participants were white (80%); 46% of the women had not previously been Pap tested and 75% were sexually active. A Fisher’s exact test of independence indicated that women who were sexually active were more likely to have had a Pap test than women who were not active, p < .001.

Procedure and materials

The experiment occurred at one time point and was completed online. Participants were randomly assigned to one of four message conditions according to the 2 (Message Frame: Gain vs Loss) × 2 (Response Cost: High Risk vs Low Risk) experimental design. After participants completed the pre-intervention questionnaire, they read their assigned message and then completed the post-intervention questionnaire.

The Centers for Disease Control and Prevention’s (2010) report about adverse events following the HPV vaccine was used as a reference to manipulate perceptions relating to the response cost of the vaccine. The wording of the report was modified to emphasize either inaccurate high-risk information or accurate low-risk information about the HPV vaccine, clinical trials and side-effects. An example gain-framed message stated, ‘Women who receive the HPV vaccine are protected against HPV as well as cervical cancer and genital warts’, whereas, a loss-framed message stated, ‘Women who are not vaccinated are at risk of contracting HPV as well as cervical cancer and/or or genital warts’. An example high-risk response cost message stated, ‘Before the HPV vaccine was licensed it was only tested in seven clinical trials which only included 21,000 women ages 9 through 26’, whereas a low-risk message stated, ‘Before the HPV vaccine was licensed it was heavily tested in seven clinical trials and on over 21,000 women ages 9 through 26’ (Table 1).

Table 1.

Example messages

Gain-frame	Loss-frame
Women who choose to be vaccinated are protected against HPV	Women who choose not to get the vaccine will not be protected against HPV
High risk	Low risk
Before the HPV vaccine was licensed it was only tested in 7 clinical trials which only included 21,000 women ages 9 through 26	Before the HPV vaccine was licensed it was heavily tested in 7 clinical trials and on over 21,000 women ages 9 through 26

Measures

Pre-intervention questionnaire

Demographic questionnaire. Participants indicated their sex, age, ethnicity and aboriginal status.

Sexual health questionnaire. Participants indicated ‘yes’ or ‘no’ as to whether they had already received the HPV vaccine and whether they had been Pap tested. Participants also indicated whether they had beliefs against the HPV vaccine and whether they were sexually active on a seven-point Likert scale (1 = strongly disagree; 7 = strongly agree).

Post-intervention questionnaires

All scales demonstrated acceptable internal reliability with a Cronbach’s alpha of .74 or greater (Nunnally, 1978).

Coping appraisal. A three-item measure adapted from Ajzen (2002), Courneya and Hellsten (2001) and Neuwirth et al. (2000) was used to assess participants’ perceptions of the response efficacy of the HPV vaccine for preventing HPV infection, cervical cancer and genital warts. Items were rated on a seven-point Likert scale.

A nine-item response cost measure adapted from Ajzen (2002), Courneya and Hellsten (2001) and Neuwirth et al. (2000) was used to assess participants’ perceptions of the benefits (three items) and costs (six items) of HPV vaccine. The stem ‘Being vaccinated against HPV…’ preceded statements such as ‘leads to serious side-effects’ and ‘leads to feelings of relief’. Items were rated on a seven-point Likert scale. Benefit items were reverse scored such that a higher score on the scale indicates higher perceived cost of being vaccinated.

To assess participants’ self-efficacy to obtain the HPV vaccine in the next three months, six items adapted from Rhodes and Courneya (2003) were used. Items were rated on a seven-point scale and included items such as ‘I believe that I have the ability to get vaccinated against HPV in the next three months (definitely do not/definitely do)’.

Threat appraisal. To assess participants’ perceptions of vulnerability to HPV, a two-item measure adapted from Neuwirth et al. (2000) was used. Participants responded using a scale that ranged from 1 (very likely) to 7 (very unlikely). An example item is: ‘Considering your personal circumstances, how likely is it that you will get genital warts?’

To assess participants’ perceptions of severity of HPV, a three-item measure adapted from Courneya and Hellsten (2001) was used. Participants responded using a scale that ranged from 1 (very likely) to 7 (very unlikely). An example item included is: ‘I feel that HPV is a serious infection for me to contract.’

Protection motivation. To assess participants’ motivation to get vaccinated against HPV, a three-item intentions measure adapted from Ajzen’s (2002) recommendations was used. Items began with one of three stems (‘I intend’, ‘I will’ and ‘I plan’) and captured intentions, within the next three months, to be vaccinated against HPV within the next three months. Responses were rated using a seven-point Likert scale. In the vaccination literature, these items are occasionally considered measures of vaccine acceptability (Brewer and Fazekas, 2007).

Knowledge comprehension. Participants answered four multiple choice questions that were designed to emphasize further the content of the message and to assess comprehension. Responses to the four questions were totalled. Correct responses were given a value of one and incorrect responses were given a value of zero (see Appendix).

Manipulation check. To assess whether participants attended to the message frame and the risk information conveyed in the passage, participants responded to four items using a seven-point Likert scale. An example item is: ‘The passage heavily focused on the benefits of getting the HPV vaccine.’

Results

We tested our hypotheses concurrently using 2 Frame (Gain vs Loss) × 2 Response Cost (High Risk vs Low Risk) × 2 Pap Status (Yes vs No) ANCOVA for each dependent variable. Participants’ beliefs against the HPV vaccine served as a covariate. Cohen’s d was calculated to determine the effect sizes. Effect sizes of .20, .50 and .80 were considered small, medium and large effects respectively (Cohen, 1992).

Protection motivation

Consistent with Hypothesis 1, women who received a low-risk message were more motivated to be vaccinated compared to women who received a high-risk message, F(1, 278) = 4.47, d = .20, p = .04 (Table 2). Contrary to Hypotheses 2 and 3b, there were no significant interactions.

Table 2.

Main effects

	Conveyed risk		p	d	Pap test		p	d	Frame		p	d
	High	Low			Yes	No			Gain	Loss
Protection motivation	20.00±11.23	22.27±11.37	.05	.20	21.49±10.80	20.75±11.99	.28	.06	21.40±11.59	20.81±11.03	.81	.05
Self-efficacy	29.42±8.44	29.47±8.28	.78	.01	29.95±8.31	28.86±8.36	.12	.13	29.52±8.42	29.35±8.26	.81	.02
Response efficacy	15.52±4.39	15.39±4.83	.95	.03	15.21±4.64	15.74±4.58	.42	.12	15.93±4.33	14.84±4.90	.21	.24
Response cost	33.33±8.50	29.48±8.35	.00	.46	30.92±8.36	31.91±8.93	.15	.12	31.69±8.99	30.98±8.16	.12	.08
Perceived vulnerability	5.18±2.72	5.20±2.91	.91	.01	5.71±3.08	4.56±2.32	.00	.42	5.21±2.74	5.16±2.90	.82	.02
Perceived severity	17.50±4.26	17.06±4.48	.30	.10	17.63±4.17	16.86±4.58	.04	.18	17.46±4.25	17.04±4.53	.63	.10

Note: *p < .05.

Coping appraisal

Response efficacy

A Frame × Pap Status interaction was observed, F(1, 281) = 6.98, p = .01. To decompose this interaction, we conducted separate ANCOVAs for each Pap status comparing response efficacy across the gain- and loss-framed conditions. Among women who were Pap tested, women who received a gain-framed message perceived the response efficacy of the vaccine to be greater than women who received a loss-framed message F(1, 154) = 8.47, p = .001, d = .49. Contrary to Hypothesis 3b, no framing effects were found among women who had not been Pap tested (p > .05).

Response cost

Consistent with Hypothesis 1, women who received a low-risk message perceived greater response cost of the vaccine compared to women who received a high-risk message, F(1, 277) = 17.70, d = .46, p = .001 (Table 3). A three-way interaction, also was detected, F(1, 277) = 5.93, p = .02. To decompose the interaction we conducted ANCOVAs for each Pap status group comparing perceived response cost across the four experimental conditions. Significant effects for condition were followed up with Bonferroni adjusted pairwise comparisons.

Table 3.

Three-way interaction response efficacy

	Pap tested				Not pap tested
	HRG	HRL	LRG	LRL	HRG	HRL	LRG	LRL
Response cost	32.12±7.63	34.70±8.29*	29.75±8.66*	27.95±7.80*	35.25±10.06*	31.52±7.54	30.00±8.97*	30.67±7.75
Protection motivation	21.53±10.77	18.65±9.70	21.58±11.65	23.40±10.64	19.36±11.90	19.76±12.71	22.87±12.90	20.43±10.77

Note: HRG = high-risk gain; HRL = high-risk loss; LRG = low-risk gain; LRL = low-risk loss; *p < .05.

Condition effects emerged for both groups of women, F_{not Pap test}(1, 127) = 3.16, p = .03; F_{Pap tested}(1, 149) = 4.91, p = .01. In partial support of Hypothesis 3b, among women who had not been Pap tested, the low-risk, gain-framed message resulted in lower perceived response cost than the high-risk, gain-framed message (p = .05, d = .29). Contrary to Hypothesis 3b, the low-risk, gain-framed message did not result in lower perceived response cost than the high-risk, loss-framed message or the low-risk, loss-framed message (Table 3).

Among women who had been Pap tested, Bonferroni pairwise comparisons revealed that the low-risk, loss-framed message (p = .01, d = .84) and the low-risk (p = .05, d = .58), gain-framed message resulted in lower perceptions of response cost than the high-risk loss-framed message.

Self-efficacy

A Response Cost × Frame interaction was observed for self-efficacy, F(1, 277) = 5.77, p = .02. However, contrary to Hypothesis 2, Bonferroni adjusted pairwise comparisons did not reveal any signficant differences between the groups (Fig. 1).

Perceived vulnerability

Consistent with Hypothesis 3a, women who had been Pap tested had greater perceived vulnerability to HPV F(1, 281) = 13.82, d = .42, p < .001 (Table 2).

Perceived severity

Consistent with Hypothesis 3a, women who had been Pap tested perceived HPV as more severe than women who had not been Pap tested, F(1, 280) = 4.13, d = .17, p = .04 (Table 2).

Manipulation checks

Knowledge comprehension

Results indicated good understanding of the message with 96.6% of the sample scoring 6 or above. A Frame × Response Cost × Pap Status ANOVA indicated that there were no significant differences across groups.

Framing and risk

A series of Frame × Response Cost × Pap Status multivariate analyses of variance (MANOVAs) with follow-up ANOVAs were conducted to confirm the effectiveness of the framing and risk manipulations.

For the framing manipulation check, only the main effects of Frame, F(1, 286) = 7.13, p < .01, Pillai’s Trace = .05 and response cost F(1, 286) = 21.75, p < .001, Pillai’s Trace = .13, were significant. Compared to the loss-framed message, the gain-framed message was considered to have a more positive tone, F(1, 287) = 11.91, p < .01, d = .38 but it was not perceived to have a stronger focus on the benefits of the HPV vaccine (p > .05). Both the tone, F(1, 287) = 4.65, p = .03, d = .30 and the focus, F(1, 287) = 43.00, p < .001, d = .81 of the low-risk message were considered more positive than the high-risk message.

The MANOVA testing the response cost manipulation indicated a significant main effect, F(1, 286) = 50.41, p < .01, Pillai’s Trace = .26. Participants who received the high-risk message believed that the side-effects of the vaccine were riskier, F(1, 28) = 96.88, p < .01, d = .17, and that the passage focused more heavily on the risks of the vaccine, F(1, 287) = 23.09, p < .001, d = .55.

Discussion

The purpose of this study was to examine factors affecting women’s coping and threat appraisals as well as their motivation to be vaccinated against HPV infection. The three factors we examined were selected specifically to provide insight into the effects of the types of messages that were circulated when the vaccine was first released. As we predicted, women who received inaccurate high-risk messages similar to those included in some media reports perceived the vaccine as having greater response cost and were less motivated to be vaccinated compared to women who received accurate low-risk information (Hypothesis 1). These findings provide insight into the deleterious effects the negative media coverage may have had on vaccine uptake. Evidently it is critical that organizations disseminating information about the HPV vaccine (e.g. the media) be educated about the vaccine so that the information released is factual and accurate.

With the goal of providing insight into strategies to potentially mitigate the effects of inaccurate reports about a vaccine and to improve communication strategies for promoting vaccination, we examined whether supplementing information about the vaccine with persuasive gain- or loss-framed messages differentially affected how women perceived the vaccine. We also examined whether women’s receptiveness to these messages was affected by individual differences in Pap test status. We found some evidence, although weak, to support our hypothesis that supplementing high-risk information with loss-framed messages and low-risk information with gain-framed messages would affect women’s coping appraisal and motivation to be vaccinated (Hypothesis 2). Specifically, the interaction between vaccine risk information conveyed and message frame was only confirmed for the self-efficacy component of women’s coping appraisals. While inspection of the mean self-efficacy scores reveals a pattern consistent with our hypothesis, our analyses were underpowered to detect significant differences. These findings must be replicated with a larger sample before we can put forth a definitive recommendation regarding the role of framed messages in buffering the effects of negative reports in promoting vaccination and specifically in fostering confidence to get vaccinated.

Our findings highlight the importance of considering the influence of women’s past experience on their thoughts about vaccination. Women who had not been Pap tested perceived themselves to be less vulnerable to HPV than women who had been Pap tested. Furthermore, they considered the consequences of HPV less severe than women who had been Pap tested (Hypothesis 3a). HPV educators must make an effort to ensure that women who have not been Pap tested are aware of their risk of contracting HPV as well as the implications of HPV. It might also be beneficial to inform these women of the importance of a Pap smear for HPV and cancer detection as research indicates that women may be unaware that a Pap smear tests for cervical cell abnormalities that are indicative of HPV (Tiro et al., 2007).

Our findings also indicate that HPV educators should be conscious of a women’s Pap status when communicating about the HPV vaccine and when addressing negative media reports. In partial support of our hypothesis, women who had not been Pap tested and received the low-risk gain-framed message perceived fewer costs of vaccination than women who received the high-risk gain-framed message. In fact, non-Pap tested women who received the high-risk gain-framed message perceived greater costs of vaccination compared to women in all of the other message conditions. Thus, providing a gain-framed message to women who have not been Pap-tested may be contraindicated when attempting to counter the deleterious effects of high-risk messages on perceptions of response cost.

Among women who have been Pap-tested, there may be an advantage to using gain-framed messages to create perceptions of vaccine effectiveness regardless of the level of associated risk conveyed. It is also important for HPV educators to be aware that among Pap-tested women, framed messages may not be enough to counter the effects of negative reports about the vaccine on perceptions of response cost. Women who had been Pap tested and received a high-risk message considered the costs of vaccination high regardless of the frame of the additional vaccine information they received.

Determining that past behaviour moderates the effect of framed HPV vaccination messages is consistent with research by Gerend and Shepherd (2007) and further reinforces the importance of measuring variables that may act to moderate message framing effects (Latimer et al., 2007). Our findings extend the HPV vaccination messaging research by providing evidence supporting an interaction between response cost information in addition to message frame and past behaviour. These findings emphasize that perceptions of personal risk of disease (i.e. threat appraisal) and perceptions of the risk associated with a target behaviour (i.e. response cost) are distinct factors that differentially influence when gain- and loss-framed messages are most effective. To explore further this distinction, future research should investigate how Pap status affects women’s threat appraisal of HPV as well as the underlying reasons as to why Pap status acts as a moderator of framing effects. This investigation is necessary because receiving a Pap test does not necessarily indicate that women understand their personal risk for disease. Indeed, in message framing research generally, moderating variables should be measured carefully and thoroughly to provide a clear conceptual understanding of their moderating effects.

Future research also should address the limitations of the current study by including a behavioural measure of vaccination and by recruiting a large, heterogenous sample of participants. With response cost information emerging as the only moderator of women’s motivation to receive the HPV vaccine in the current study, more research is needed to determine strategies to motivate women to get the HPV vaccination within the context of mixed reports related to vaccine safety.

In conclusion, reports inaccurately conveying the HPV vaccine as risky have the potential to undermine women’s motivation to obtain the vaccine. Messages aiming to mitigate the effects of these negative reports may be more effective if they are tailored to characteristics of the message recipient. Taking individual differences into account also will help to identify groups needing more information and education about HPV in general.

Footnotes

Appendix

Acknowledgements

The research project was supported by an Operating Grant from the Social Sciences and Humanities Research Council of Canada (SSHRC). Preparation of the manuscript was supported by a Canada Research Chair award (AEL) and a Doctoral Fellowship (HLG) from the Canadian Institutes of Health Research.

Competing Interests

None declared.

References

Abdelmutti

Hoffman-Goetz

(2009) Risk messages about HPV, cervical cancer, and the HPV vaccine Gardasil: A content analysis of Canadian and U.S. national newspaper articles. Women & Health 49(5): 422–440.

Abhyankar

O’Connor

Lawton

(2008) The role of message framing in promoting MMR vaccination: Evidence of a loss-frame advantage. Psychology, Health & Medicine 13(1): 1–16.

Ajzen

(2002) Constructing a TPB questionnaire: Conceptual and methodological considerations. Available at: http://people.umass.edu/aizen/pdf/tpb.measurement.pdf.

Brewer

Fazekas

(2007) Predictors of HPV vaccine acceptability: A theory-informed, systematic review. Preventive Medicine 45(2–3): 107–114.

Caskey

Lindau

Alexander

(2009) Knowledge and early adoption of the HPV vaccine among girls and young women: Results of a national survey. Journal of Adolescent Health 45(5): 453–462.

Centers for Disease Control and Prevention, National Center for Emerging and Zoonotic Infectious Diseases and Division of Healthcare Quality Promotion (2010) Reports of health concerns following HPV vaccination. Available at: http://www.cdc.gov/vaccinesafety/vaccines/hpv/gardasil.html (accessed March 2009).

Cherubini

Rumiati

Rossi

Nigro

Calabro

(2005) Improving attitudes toward prostate examinations by loss-framed appeals. Journal of Applied Social Psychology 35(4): 732–744.

Cohen

(1992) Statistical power analysis. Current Directions in Psychological Science 1(3): 98–101.

Consedine

Horton

Magai

Kukafka

(2007) Breast screening in response to gain, loss, and empowerment framed messages among diverse, low-income women. Journal of Health Care for the Poor and Underserved 18(3): 550–566.

10.

Courneya

Hellsten

(2001) Cancer prevention as a source of exercise motivation: An experimental test using protection motivation theory. Psychology, Health & Medicine 6(1): 59–64.

11.

Gerend

Shepherd

(2007) Using message framing to promote acceptance of the Human Papillomavirus vaccine. Health Psychology 26(6): 745–752.

12.

Gerend

Sias

(2009) Message framing and color priming: How subtle threat cues affect persuasion. Journal of Experimental Social Psychology 45(4): 999–1002.

13.

Gerend

Shepherd

Monday

(2008) Behavioural frequency moderates the effects of message framing on HPV vaccine acceptability. Annals of Behavioural Medicine 35: 221–229.

14.

Hevey

Pertl

Thomas

Maher

Craig

Chuinneagain

(2010) Body consciousness moderates the effect of message framing on intentions to use sunscreen. Journal of Health Psychology 15(4): 553–559.

15.

Latimer

Salovey

Rothman

(2007) The effectiveness of gain-framed messages for encouraging disease prevention behaviour: Is all hope lost? Journal of Health Communication 12(7): 645–649.

16.

Latimer

Rench

Rivers

. (2008) Promoting participation in physical activity using framed messages: An application of prospect theory. British Journal of Health Psychology 13(Pt 4): 659–681.

17.

Lee

Parsons

Gentleman

(1998) Failing short of Pap test guidelines. Health Reports 10(1): 9–19.

18.

McCaul

Johnson

Rothman

(2002) The effects of framing and action instructions on whether older adults obtain flu shots. Health Psychology 21(6): 624–628.

19.

Meyerowitz

Chaiken

(1987) The effect of message framing on breast self-examination attitudes, intentions, and behaviour. Journal of Personality and Social Psychology 52(3): 500–510.

20.

Neuwirth

Dunwoody

Griffin

(2000) Protection motivation and risk communication. Risk Analysis 20(5): 721–734.

21.

Nunnally

(1978) Psychometric Theory. New York: McGraw-Hill.

22.

Rhodes

Courneya

(2003) Self-efficacy, controllability and intention in the theory of planned behaviour: Measurement redundancy or causal independence? Psychology and Health 18(1): 79–91.

23.

Rogers

(1975) A protection motivation theory of fear appeals and attitude change. Journal of Psychology 91: 93–114.

24.

Rothman

Salovey

(1997) Shaping perceptions to motivate health behaviour: The role of message framing. Psychological Bulletin 121(1): 3–19.

25.

Tiro

Meissner

Kobrin

Cholette

(2007) What do women in the U.S. know about Human Papillomavirus and cervical cancer? Cancer Epidemiology Biomarkers Prevention 16(2): 288–294.

26.

Toll

Salovey

O’Malley

Mazure

Latimer

McKee

(2008) Message framing for smoking cessation: The interaction of risk perceptions and gender. Nicotine Tobacco Research 10(1): 195–200.

27.

Williams

Clarke

Borland

(2001) Effects of message framing on breast-cancer-related beliefs and behaviours: The role of mediating factors. Journal of Applied Social Psychology 31(5): 925–950.