Skip to main content
NCBI home page
Wiley Open Access Collection logoLink to Wiley Open Access Collection
. 2026 May 24;117(4):1657–1672. doi: 10.1111/bjop.70085

Understanding considerations and priorities in health behaviours: A multi‐dimensional risk–benefit perspective

Yiyun Shou 1,2,3,✉, Joey Elizabeth Yeo 1, Alexander Shao‐Rong Pang 4
PMCID: PMC13655933  PMID: 42178943

Abstract

Health behaviours and decisions are complex and involve different dimensions of individuals' lives. The current study aimed to investigate considerations and their priorities in health decision and behaviour adoption via a risk–benefit framework unifying elements of mainstream models of health behaviour. We aimed to explore the core dimensions of the outcomes being considered by people and individual differences in perceived importance across dimensions. Two studies were conducted among U.S. adult residents. Study 1 (N = 295) involved eliciting qualitative responses of the health decisions or behaviours that participants encounter in their everyday life, and ratings of the associated dimensions. Study 2 (N = 400) assessed participants' ratings of the importance of outcomes and dimensions across various health decisions and behaviours. Both studies revealed that people considered dimensions beyond health risk and benefit. The dimensions varied their priorities, with health and emotional outcomes being the top considerations. Participants also showed asymmetric considerations of benefit and risk within each dimension. The latent profile analyses in Study 2 also revealed two distinct groups of individuals in risk and benefit considerations across dimensions. The findings highlight that acknowledging individual differences in the weighting and evaluation of outcomes across domains is essential for more tailored health communication and interventions.

Keywords: benefit perceptions, health behaviour, health decisions, risk perceptions

BACKGROUND

Enormous literature has demonstrated the complexity of health decisions and behaviours in the sense that factors are usually beyond one's consideration of health. A number of models have been proposed to explain one's health decision and behaviour and outline the key psychological and social determinants. Most elements across various models affect one's decision and behaviour adoption essentially as a form of risk (negative) and benefit (positive) in terms of the outcomes of decisions.

The Health Belief Model (HBM; Ritchie et al., 2021; Rosenstock et al., 1988; Skinner et al., 2015), Protection Motivation Theory (PMT; Balla & Hagger, 2025; Murgraff et al., 1999), the Integrated Behavioural Model (IBM; Conner & Sparks, 2005; McEachan et al., 2011; Morrison et al., 2002), and Social Cognitive Theory (SCT; Bandura, 1986, 2004) share many common conceptual components despite using different terminology. These models converge on risk/loss perception (i.e., perceived severity and susceptibility/vulnerability to negative health outcomes; barriers or costs that impede action) and benefit evaluation (i.e., perceived advantages of behaviour adoption, including response efficacy and outcome expectations) as central determinants of health behaviours.

For example, in HBM (Ritchie et al., 2021; Rosenstock et al., 1988; Skinner et al., 2015), perceived severity and susceptibility constitute risk perception (El‐Toukhy, 2015), while perceived benefits refer to advantages gained from specific action (typically health risk reduction), and perceived barriers include factors that impede behaviour (e.g. financial costs, time, effort). Self‐efficacy, one's belief in their ability or success likelihood to adopt the behaviour, influences their perceived likelihoods of outcomes (i.e., the probability of an outcome is the sum of the probability of the outcome given the decision is taken and the probability of the outcome given the decision is not taken). Similarly, PMT's (Balla & Hagger, 2025; Murgraff et al., 1999) threat appraisal (perceived vulnerability and severity) must outweigh the rewards (i.e., benefit) associated with the maladaptive behaviours, and response efficacy (i.e., benefit) must outweigh the response costs for protection motivation to be elicited (Conner, 2019). The IBM (Conner & Sparks, 2005; McEachan et al., 2011; Morrison et al., 2002) posits that attitudes, a function of behavioural beliefs based on individuals' evaluations of each potential consequence (Yzer, 2017), and subjective norms reflecting social consequences of compliance or deviation from group expectations serve as direct determinants of behaviour (Yzer, 2017). SCT considers personal cognitive factors (e.g., outcome expectation), behavioural factors (e.g., reinforcements including reward and punishment), and socio‐environmental factors (e.g., normative beliefs, social support, and enablers and barriers). The outcome considerations spread out of these factors.

Although various health behaviour and decision models incorporate various elements, as discussed above, most can be centred around the roles of risk (probability and value of negative outcomes such as harm and costs), and benefits of distinct types (e.g., health or social) (Augeraud‐Véron & Leandri, 2024). To substantiate, even though there are health benefits to healthy eating, social barriers to healthy eating in nurses include social pressure from colleagues, resulting in guilt, if gifted cakes are not eaten (Nicholls et al., 2017). Negative social consequences and positive health consequences were associated with lower drinking intentions, and this association is stronger for those who reported higher levels of previous drinking (Kingsbury et al., 2015). Uptake of cancer screening and coping with serious/chronic illnesses have financial implications (Head et al., 2018; Kasting et al., 2021). A public health measure of stringent control of opioids would impose legal consequences for overutilisation of opioids and sedatives in symptom management of palliative patients (Pinkerton et al., 2020; Sim et al., 2014).

The current study aims to understand individuals' considerations and priorities in health decisions and behaviour adoption by unifying the key elements across models of health behaviours. While many psychological processes and social factors can influence the assessment of outcomes of a health behaviour, one's decision relies on the final aggregated evaluation of outcomes and their likelihoods. One's weighting of an outcome depends on one's perceived importance of outcomes. Understanding how people perceive the outcomes and their importance is essential for understanding the key elements of behaviour determinants and potential individual differences in decision‐making on health behaviours.

Meanwhile, valuations of the outcomes across dimensions can vary across subgroups of the population. Younger, less educated, lower‐income, and male individuals may value lower health risks as they tend to perceive lower levels of health threats and risks compared to their counterparts (Badr et al., 2021; Bechard et al., 2021; Bhuiya et al., 2021; Karimy et al., 2021; Niño et al., 2021). This may be because older individuals face higher health risks (Karimy et al., 2021), while young people tend to value social outcomes as they are more likely to conform to social norms (Helms et al., 2014) and be affected by social comparisons, thus often underestimating health risks (Kim et al., 2018). Individuals with financial hardship are associated with unhealthy behaviours (Mendez et al., 2024), which could be due to these individuals having to prioritise financial outcomes. Middle‐aged patients, females, and those with lower education levels are also more likely to be concerned about financial barriers to healthy food and medications (Campbell et al., 2017; Kim et al., 2017). Furthermore, educated individuals may value social and health outcomes more, as they are equipped with more knowledge on health risks (Badr et al., 2021) and have stronger stigmas against unhealthy behaviours such as smoking due to social norms (Stuber et al., 2008).

This research applies a multi‐dimensional risk–benefit framework integrating the key elements of mainstream models of health behaviours. We define risk or risky outcome as the negative or undesirable outcome of a behaviour, while benefit is defined as the positive or desirable outcome. The first goal of this research was to explore the core dimensions of the outcomes being considered by people in outcome evaluation and their importance in health behaviours/decisions. This goal will be examined in a qualitative study (Study 1) and a quantitative study (Study 2). The second goal is to explore if there are distinct groups of individuals in terms of their risk and benefit considerations across dimensions.

STUDY 1

Study 1 aimed to explore the outcome dimensions being considered by participants in health decision‐making and to elicit the specific health decisions or behaviours they encounter in their daily lives. Traditional health behaviour frameworks are usually applied to specific health behaviours determined by researchers or health experts. However, they may not fully reflect participants' views nowadays in the rapidly evolving society, especially since the COVID‐19 pandemic.

Method

Participants and procedures

The inclusion criteria were U.S. residents aged 18 or above, whose first language is English. The exclusion criteria were below aged 18 or self‐reported cognitive impairment. A total of 300 participants completed the survey. Five participants did not provide meaningful responses (e.g., irrelevant or nonsense) and were excluded. The remaining 295 participants included 146 (49%) females and were aged between 18 and 79 (M = 40.62, SD = 15.49). Of the participants, 221 (75%) were White, 26 (9%) were African American, 31 (11%) were Asian, and 17 (5%) were mixed ethnicities/other; 176 (60%) had a Bachelor's or above education; 117 (40%) had a current or history of physical health conditions. The study was programmed using the Qualtrics survey platform. Participants were recruited from Prolific. Participants were invited and accessed to the survey via Prolific. Participants were reimbursed with approximately 1 pound (approx. 1.3 USD) for their participation. The study was approved by the Institutional Review Board of National University of Singapore (NUS‐IRB‐2022‐699).

Materials

Participants were asked to describe a situation that involved the uptake of health behaviours or decisions. To encourage participants to think more about the decisional aspect, especially benefit and risk evaluation, participants were asked to think about a situation that one could be faced with uncertainty. In the description, participants were required to describe the potential negative and positive outcomes of the respective behaviour or decision. Participants provided their description in an open text response box. After providing the description, participants were asked to select the aspects of the situation's potential negative outcomes (as risks) from 11 options listed in Table 1. They were also asked to select the aspects for potential positive outcomes (as benefits). Then, participants were asked to list the top outcomes they would consider when they were deciding whether to take the behaviour or decision. They could list a maximum of four specific outcomes ranked by their level of importance in four text boxes and all other outcomes in the fifth text box. They listed negative and positive outcomes separately.

TABLE 1.

Descriptive statistics of dimensions selected and recalled outcomes in Study 1.

Variables Risk or benefit Benefit Risk
Proportion N Proportion Mean rank SD N Proportion Mean rank SD
Domain selected
Career/Education 0.38 65 0.22 86 0.29
Emotions 0.72 165 0.56 170 0.58
Health 0.97 238 0.81 272 0.92
Finance 0.48 58 0.20 124 0.42
Time/Effort 0.50 67 0.23 132 0.45
Legality 0.10 12 0.04 25 0.08
Morality 0.23 35 0.12 53 0.18
Relation‐F 0.53 110 0.37 115 0.39
Relation‐O 0.38 82 0.28 71 0.24
Leisure 0.34 69 0.23 69 0.23
Other 0.07 16 0.05 8 0.03
Domain recalled
Career/Education 0.17 28 0.09 3.21 1.42 37 0.13 3.19 1.17
Emotions 0.49 110 0.37 3.85 1.11 86 0.29 3.45 0.97
Health 0.93 213 0.72 4.86 0.46 253 0.86 4.79 0.56
Finance 0.41 39 0.13 3.56 0.99 110 0.37 3.45 1.11
Time/Effort 0.23 25 0.08 3.20 1.12 56 0.19 3.16 1.07
Legality 0.03 2 0.01 4.00 1.41 10 0.03 3.30 1.16
Morality 0.16 26 0.09 3.77 1.03 27 0.09 3.22 1.25
Pain/Suffering 0.08 16 0.05 4.31 0.87 15 0.05 3.60 1.06
Relation‐F 0.29 49 0.17 3.63 0.95 70 0.24 3.51 1.02
Relation‐O 0.11 17 0.06 3.71 1.21 18 0.06 3.17 1.47
Leisure 0.09 24 0.08 3.46 1.22 11 0.04 2.45 1.04
Mixed/Other 0.21 33 0.11 4.15 0.97 47 0.16 3.98 0.97
Open in a new tab

Note: Career/Education = Employment/Career/Education, Relation‐F = Relationship with families/friends, Relation‐O = Relationships with people other than families/friends; Mixed/other: outcomes that did not have a clear dimension or had a mixture of several dimensions. The scoring of the rank: 5 = if listed in the first outcome box (as most important), 1 = if listed in the fifth, last outcome box, missing if not listed in the outcome boxes.

The key health behaviour or decision described in the context was extracted by the second author and was reviewed by the first author. The top important outcomes participants provided were coded into a list of categories as listed in Table 1. The coding was done by two independent coders (the second author and the third author) with an agreement of about 85% of the initial coding of all recalled outcomes. The coding results were reviewed by a third coder (the first author), revised and resolved before data analysis.

Results and discussion

The health behaviours or decisions in the situations described by respondents were categorised into 19 types of health behaviour (see Table S1 for the list). The top‐mentioned health behaviour was deciding on taking a treatment (23% of participants mentioned), followed by taking medication (16%), vaccination (11%), visiting medical professionals/seeking health (10%), pregnancy/having a child/birth control/abortion (5%), and diet (5%). The recipients of these health decisions are mostly self (i.e., the patients themselves) for approximately 92% of health decisions, while the rest (8%) involved others.

Table 1 presents the frequency of dimensions selected as risk and benefit outcomes. Health and emotion were most often selected dimensions for risk outcomes at 92% and 58% respectively, and for benefits at 81% and 56%, respectively. When comparing risk and benefit outcomes, the frequency with which participants selected finance and time/effort chosen as dimensions for risk outcomes was 22% higher than for benefits. Overall, the health dimension was chosen in 97% of the scenarios either as risk or benefit consideration, followed by emotion (72%), relationships with family and friends (53%), time/effort (50%), finance (48%), and career/education (38%). Table 1 also summarises the recalled outcomes. Health, emotion, and finance were the most commonly recalled dimensions overall. Health, emotion, and family/friends were mostly recalled dimensions for beneficial outcomes, while health, emotion, and finance were mostly recalled dimensions for risk outcomes. In addition to the predefined dimensions in Table 1, we also observed frequent recalls of pain/suffering, which can be both health‐ and emotion‐related. When being recalled as a consideration, health, emotion, and pain/suffering had the highest priority in benefit consideration, while health, pain/suffering, and family/friends had the highest priority in risk consideration.

STUDY 2

The dimension ratings and outcome recall in Study 1 were contingent on the scenarios described by the participants. Study 2 aimed to systematically examine the importance of various outcome dimensions as participants navigated health behaviours or decisions. We expected that participants would rate importance highly on dimensions including health, emotion, finance, and time based on Study 1's results and the mainstream health behaviour models. Study 2 also addressed the second goal of this research by investigating the patterns of the importance ratings among participants and seeing if there are distinct groups of participants.

Method

Participants

The inclusion criteria of participants were (1) U.S. residents, (2) aged 18 or above, and (3) possess English as their first language. The exclusion criteria were non‐U.S. residents, aged below 18, and English not being their first language. A total of 435 participants who met the inclusion criteria were recruited from Prolific to complete the survey. Thirty‐five of them were excluded due to not passing the attention check and were excluded from the analysis. A total of 400 participants were included in the analysis. The participants were aged between 18 and 81 (M = 39.40, SD = 13.44). About 205 (51%) were females and 247 (61.75%) had Bachelor's or above education; 189 (47.25%) were White, 149 (37.25%) were African American, 34 (8.5%) were Asian, and 26 (6.5%) were Other; 140 (35%) had a current or history of physical health conditions.

Materials

A total of 30 medical and health scenarios were generated from an ongoing systematic review on health behaviours in the literature and scenarios generated by the participants in Study 1. These scenarios covered diverse types of medical and health decision scenarios including treatment decisions, preventive/protective behaviours, lifestyle behaviours, risky health behaviours, and preventive behaviours (see Supporting Information S2). Participants were asked to indicate how important the outcomes in each dimension were to them when they were considering the decisions in those scenarios. A total of 11 dimensions were provided for ratings based on Study 1's results. The outcomes are separated into benefit and risk outcomes. Participants rated the level of importance for each type of outcome (benefit or risk) and for each dimension on a 5‐point scale from 1 = not important at all to 5 = extremely important.

Procedure

Participants were invited from Prolific to the study hosted on the Qualtrics survey platform. They provided consent, completed demographic background variables, and completed the ratings for the 30 scenarios. Participants were reimbursed with 4.06 pounds (approx. 5.30 USD) for their participation. The study was approved by the Institutional Review Board of National University of Singapore (NUS‐IRB‐2023‐515).

Data analytic methods

Average item correlations and Cronbach's alphas were calculated to assess the consistency of the importance ratings across the 30 decisional scenarios for each outcome type in each dimension. Mixed‐effects ordinal regression with random intercept for participants was run to compare benefit outcome and risk outcome for each dimension and to assess the order of the overall priority of the 18 outcomes (2 types × 9 dimensions). Mixed effects ordinal regression was estimated using the ‘ordinal’ package.

We then applied latent profile analysis (LPA) using participants' ratings on the benefit and risk across nine domains to identify possible subgroups of participants. LPA is based on the Gaussian mixture models and estimated using the ‘Mclust’ package (Version 6.1.1; Scrucca et al., 2023) in the R program. We evaluated 14 types of models that varied in their assumptions of variance and covariance structures within and between clusters. Bayesian regularisation using a normal prior for means based on data mean and covariances, and inverse Wishart for the covariance matrix (Fraley et al., 2012) was implemented to avoid singularity and shrinkage in complex models (Scrucca et al., 2016). BIC was used to identify the best fit models that successfully converge. The model selection and determination of the number of profiles are based on (1) BIC values (higher is better in the ‘Mclust’ package), (2) Entropy ≥0.8 (as an index to determine the cluster separate) (Clark & Muthén, 2009), wherein a higher value indicates better cluster separation, (3) average posterior classification probabilities >0.7 (Nylund‐Gibson & Choi, 2018) wherein a higher value indicates better cluster separation, and (4) no single cluster has fewer than 25 cases (Lubke & Neale, 2006).

Results

Table 2 displays the average importance ratings of the 18 outcome types across all health behaviours/decisions among participants. Participants rated negative/risky outcomes as being of more importance for finance and time aspects, while rating beneficial outcomes as of more importance than the negative outcomes for all the other seven domains (see Table 3). We conducted ordinal regression models with random intercepts for participants, outcome types, and behaviours nested within outcome types to examine the rating variability across health behaviours within outcome types. There were some variabilities across scenarios within outcome types for both the risky outcomes (random intercept variance = 0.24) and beneficial outcomes (random intercept variance = 0.20). Nevertheless, the high average correlations and ICCs of the ratings for each outcome type across the 30 health behaviours/decisions show that the aggregated ratings for each outcome type can be a reliable indication of participants' average importance view.

TABLE 2.

Summary of the ratings of the nine dimensions in Study 2.

Aspect Benefit Risk Risk vs. benefit
M (SD) AIC M (SD) AIC b p
Career 2.78 (1.46) 0.53 2.65 (1.45) 0.53 −0.25 <.001
Emotions 3.47 (1.32) 0.42 3.29 (1.37) 0.43 −0.30 <.001
Finance 2.97 (1.47) 0.48 3.11 (1.41) 0.39 0.23 <.001
Health 4.04 (1.20) 0.36 3.83 (1.33) 0.39 −0.36 <.001
Leisure 2.75 (1.44) 0.50 2.60 (1.42) 0.54 −0.26 <.001
Morality 2.71 (1.52) 0.58 2.59 (1.48) 0.56 −0.22 <.001
Social close 2.95 (1.44) 0.50 2.80 (1.42) 0.50 −0.28 <.001
Social other 2.61 (1.42) 0.55 2.49 (1.40) 0.54 −0.24 <.001
Time 3.08 (1.38) 0.49 3.13 (1.31) 0.43 0.09 <.001
Open in a new tab

Note: Cronbach's alphas for all domains and outcome types were ≥0.94. Comparisons between benefit and risk ratings were based on mixed effects ordinal regression with subject ID as random intercept.

Abbreviation: AIC, average inter‐item correlations between 30 scenarios.

TABLE 3.

Comparison of the ratings between adjacent pairs of dimensions.

Rank groups Paired domain comparison b p Adjusted p
Last Morality risk vs. social other risk 0.10 <.001 <.001
5 Leisure risk vs. morality risk 0.05 .004 .072
Social other benefit vs. leisure risk 0.01 .749 1
Career risk vs. social‐other benefit 0.04 .018 .324
Morality benefit vs. career risk 0.05 .004 .072
4 Leisure benefit vs. morality benefit 0.07 <.001 <.001
Career benefit vs. leisure benefit 0.05 .004 .072
Social close risk vs. career benefit 0.01 .424 1
3 Social close benefit vs. social close risk 0.20 <.001 <.001
Finance benefit vs. social close benefit 0.02 .271 1
2 Time benefit vs. finance benefit 0.14 <.001 <.001
Finance risk vs. time benefit 0.04 .024 .432
Time risk vs. finance risk 0.01 .488 1
Top Emotions risk vs. time risk 0.19 <.001 <.001
Emotions benefit vs. emotions risk 0.22 <.001 <.001
Health risk vs. emotions benefit 0.49 <.001 <.001
Health benefit vs. health risk 0.25 <.001 <.001
Open in a new tab

As shown in Figure 1, participants rated health benefits and risks as most important, followed by emotional benefits and risks. Time and financial outcomes are the next most important considerations. Potential risks to other people in society were considered least important, followed by morality, leisure, and benefit to people in society.

FIGURE 1.

FIGURE 1

Open in a new tab

Importance ratings of the risk and benefit domains.

Latent profile analysis

The BIC values for solutions of 1 to 8 profiles across 14 model types suggested that the best‐fit model is a two‐profile VVV (varying volumes, variances, and covariances between clusters) model (BIC = 8310; AIC = 6797), which had a substantially better fit than the second best model (two‐profile VEV [equal variances, varying volumes, and covariances between clusters] model, BIC = 8877; AIC = 7433). The two‐profile VVV model suggested about 25% of the cases (n = 96) in one profile and 75% (n = 304) in the second profile. The average posterior classification probabilities were 0.99 (0.86–1) for profile 1 and 1 for profile 2, indicating a clear profile separation.

Figure 2 illustrates the mean ratings for the two profiles across the 18 outcomes, and Table 4 shows the mean difference tests between the two profiles. Both profiles rated similarly high importance for health risks, health benefits, and emotional risks. Participants in profile 2 rated all other outcomes significantly of lower importance, suggesting that they prioritise health and emotional outcomes significantly higher than other outcomes. On the other hand, participants in profile 1 had more balanced considerations of all outcomes.

FIGURE 2.

FIGURE 2

Open in a new tab

Mean ratings across risk and benefit outcomes.

TABLE 4.

Comparison of two profiles in outcome ratings and demographic backgrounds.

Profile 1 (N = 96) Profile 2 (N = 304) p OR
Benefit Career 3.65 (0.90) 2.51 (0.95) <.001
Emotions 3.68 (0.88) 3.40 (0.83) .006
Finance 3.64 (0.86) 2.75 (0.97) <.001
Health 4.00 (0.72) 4.05 (0.67) .520
Leisure 3.67 (0.88) 2.46 (0.89) <.001
Morality 3.67 (0.91) 2.40 (1.06) <.001
Social‐close 3.64 (0.87) 2.73 (0.96) <.001
Social‐other 3.63 (0.87) 2.29 (0.89) <.001
Time 3.69 (0.88) 2.89 (0.92) <.001
Risk Career 3.39 (0.99) 2.41 (0.97) <.001
Emotions 3.46 (0.98) 3.24 (0.86) .058
Finance 3.43 (0.95) 3.00 (0.81) <.001
Health 3.83 (0.89) 3.83 (0.82) .956
Leisure 3.44 (1.01) 2.34 (0.91) <.001
Morality 3.41 (1.01) 2.34 (1.00) <.001
Social‐close 3.41 (0.96) 2.60 (0.93) <.001
Social‐other 3.38 (0.97) 2.21 (0.87) <.001
Time 3.45 (0.99) 3.03 (0.81) <.001
Age 39.9 (13.0) 39.2 (13.6) .656 1.00
Gender Male 53 (57.0%) 132 (44.4%) .046 REF
Female 40 (43.0%) 165 (55.6%) 1.37
Ethnicity African 57 (59.4%) 92 (30.3%) <.001 REF
Other 9 (9.38%) 53 (17.4%) 2.37
White 30 (31.2%) 159 (52.3%) 2.13**
Income a 3.19 (1.01) 2.67 (1.04) <.001 0.72*
Education Bachelor's & above 78 (81.2%) 169 (56.0%) <.001 REF
Other 18 (18.8%) 133 (44.0%) 2.13*
Employment Full‐time 73 (76.0%) 164 (53.9%) <.001 REF
Other 23 (24.0%) 140 (46.1%) 1.79 b
Health condition No 66 (69.5%) 187 (62.8%) .285 REF
Yes 29 (30.5%) 111 (37.2%) 1.37
Mental health condition No 60 (63.2%) 183 (61.6%) .882 REF
Yes 35 (36.8%) 114 (38.4%) 0.67
Open in a new tab

Abbreviations: OR, multiple logistics odds ratio; REF, reference category.

a

Income was measured by five levels from ‘low’ (=1) to ‘upper’ (=5).

b

Employment and income were not in the same model due to multi‐collinearity.

*

p < .05.

**

p < .01.

Table 4 shows the comparison of demographic variables between the two profiles. The two profiles did not differ significantly in age and health or mental health status. Individuals of profile 2 were more likely to be females, have a lower income, be not in full‐time employment, and be less likely to attain a Bachelor's or above education than individuals in Profile 1. In the multivariate analysis, gender was no longer a significant factor differentiating the two profiles. The results suggest that income and education level were the two primary factors that differentiate the two profiles.

Discussion

Consistent with Study 1's results, Study 2 demonstrated that health and emotion were the top considerations when participants were deciding on health behaviours or decisions, followed by financial and time/logistics, and close social networks (i.e., family, friends). Study 2's results also showed that people distinguished positive and negative outcomes when prioritising outcomes of a dimension. Echoing Study 1's finding that risk outcomes were two times more likely to be selected than benefit outcomes for finance and time/logistics, participants in Study 2 also rated risk as being more important than benefit for these two dimensions. On the other hand, while risk and benefit for health and emotion were selected at a similarly high rate in Study 1, participants in Study 2 rated benefit of the two dimensions as of more importance than risk. These findings align with many health behaviour models that include the core element of perceived benefits that usually refer to health benefits (including reduction in health risk) and barriers that commonly relate to finance and logistics.

In contrast to the literature suggesting certain demographic groups would value health risks less (Badr et al., 2021; Bechard et al., 2021; Karimy et al., 2021; Niño et al., 2021), the two profiles that emerged in the LPA had similar levels of perceived importance of health outcomes. The two profiles mainly differed in their prioritisation of other outcomes in relation to health outcomes. The first profile had a more balanced consideration across all dimensions, while the second profile significantly prioritised a few outcomes, including health and emotion, as well as risks in finance and logistics, over all other dimensions. The two profiles did not differ in age, gender and existing health conditions after controlling all demographic variables in the current study. The second profile seemed to have lower socio‐economic status (SES; lower education, not in full‐time employment and lower income) than the first profile. The outcomes in the prioritised dimensions among participants in the second profile are, in fact, often the more immediate (e.g., direct financial cost, immediate relief of discomfort) and salient in health behaviour and decision‐making. Participants from lower SES backgrounds have limited resources and thus have to attend to and allocate resources to limited dimensions, which are usually the immediate threats and needs.

GENERAL DISCUSSION

We investigated the considerations and priorities individuals apply to health decisions and behaviour adoption using a multi‐dimensional risk–benefit framework that unifies mainstream models while accounting for common, everyday dimensions and types of outcomes. Both Study 1 and Study 2 demonstrate that, when navigating health behaviours and decision‐making, individuals consider factors that extend well beyond traditional health‐related risks and benefits. This aligns with the modern biopsychosocial approach to health behaviours (Hoffman & Driscoll, 2000) and the mainstream models (Conner & Sparks, 2005; Ritchie et al., 2021). The dimensions and types of outcomes varied in their priorities, with health and emotional outcomes being the top considerations. Emotion is not a key element in many models of health behaviours. Emotions in the decision literature are usually considered as either a moderator of evaluation (e.g., emotions as cues to evaluate risks) or a byproduct of a decision (e.g., negative feeling due to bad outcomes) (George & Dane, 2016; Lerner et al., 2015). However, both the current studies and recent studies (Coifman et al., 2021; Ferrer & Mendes, 2018; Ofori‐Ansah et al., 2022) have suggested that emotions are important outcomes and have their own utilities in decision‐making.

There has been a globally increasing recognition and valuation of emotions, especially those relating to mental wellbeing, over the past few decades (Chen et al., 2025; Sun et al., 2024). Pursuing happiness and avoiding unpleasant feelings, such as pain and anxiety, are important goals for people, especially those in high‐income countries (Morgan et al., 2015). In the context of health behaviours, people exhibit ‘unhealthy’ behaviours (e.g., eating ‘unhealthy’ food, self‐medication, and smoking) to minimise unpleasant feelings (e.g., stress, anxiety, and pain) and maximise positive emotions (e.g., excitement and happiness). The subjective value of health‐related negative outcomes associated with unhealthy behaviour can be surpassed by the value of perceived reduction in stress and anxiety. Another compelling example is found in cancer treatment decisions, where people may avoid the recommended treatment if they perceive the emotional suffering involved as outweighing the health consequences of refusing treatment. Consequently, integrating emotions into existing health behaviour models is essential for a comprehensive understanding of, and meaningful response to, the complexities of human health behaviour and decision‐making.

Both Study 1 and Study 2 also demonstrated the contrast between different layers of interpersonal and social factors of health behaviours. Many mainstream models treat social factors of health via a broad concept of social norms and support without differentiating ‘whom’ or ‘which group’ the social norms and/or support come from. Interventions of health behaviours addressing social factors usually focus on changes to norms at the societal level, and the effectiveness strongly depends on whether such changes reach individuals' personal networks (Dempsey et al., 2018; Miller & Prentice, 2016; Zhang et al., 2010). The present studies highlight that one's close social ties (e.g., family and friends) hold significantly more importance than non‐primary groups (e.g., others in the same society). Potential risks to people outside these close social networks were least important to the current participants. This echoes findings that collective framing (societal norms) and value‐based appeals (protecting others) in messages during the COVID‐19 pandemic generated less engagement than messages that directly addressed health facts (Pattison et al., 2022). Similarly, peer social networks are important for adolescents' health behaviours (Montgomery et al., 2020) and interventions targeting peer norms can be effective in changing those behaviours (Stok et al., 2016).

Another important finding was the asymmetric considerations of benefit and risk within each dimension. Such asymmetry is not simply due to the attribute framing effect or loss aversion as the directions of asymmetry varied across dimensions. This finding suggests that the benefit and risk considered within each dimension are not two dichotomous states of a single outcome (e.g., recovering from an illness versus not recovering) but two separate outcomes with different stakes. For example, the health benefit of taking a treatment option is recovering from an illness, while the risk is the side effect. The financial benefit of having fast or unhealthy food is the money one can immediately save from buying more expensive healthy food, while financial risk is a potential medical cost in the unknown future. This finding may also explain the inconsistent effects of goal framing (focusing on the outcomes) and attribute framing (focusing on dichotomous states of a single outcome) across health behaviours (Akl et al., 2011), as the type of risks and benefits and their stakes can vary substantially across domains.

Finally, the findings from the LPA highlight that the considerations of various dimensions and outcomes vary significantly across individuals. Those in low‐resource settings may have limited resources to account for all dimensions of a health behaviour or decision, forcing them to prioritise immediate needs (e.g., finance, time at work) (Nour et al., 2006). This could explain why health behaviour interventions often yield varying effects on individuals with different socio‐economic profiles (Van Den Bekerom et al., 2024). This highlights the need of understanding health behaviours and developing interventions via the lens of health equity (Gallagher et al., 2025).

Implications

Current health behaviour models lack synergy across elements with varying terminology and no clear information on weighting or relative importance of different elements. These models typically present elements in parallel, implicitly suggesting they are linearly and equally associated with a given behaviour or decision. This lack of synergy, weighting, and aggregation logic limits the systematic design of interventions, as well as the determination of the prioritisation of certain aspects or dimensions in low‐resource settings. By systematically accounting for outcome dimensions and their relative importance, the multi‐dimensional risk–benefit perspective provides an avenue and framework for designing targeted, effective, and tailored interventions for health behaviours interventions. These findings also highlight that acknowledging individual differences in the weighting and evaluation of the outcomes across domains is essential for developing more tailored health communication and intervention. The two profiles demonstrate that while factors beyond health, such as career, finances, and social relationships, all influence health decisions, the extent of this impact varies across subgroups and demographic profiles in the population. These results imply that health behaviour models should account for population heterogeneity, particularly regarding the social determinants of health behaviours.

Furthermore, these findings further reinforce the inherent complexity of real‐world health decision‐making. While the outcome types tested here are commonly discussed in the literature, they may not fully capture the intricate array of considerations involved in actual health behaviours and decisions. This is also reflected by the prevalence of ‘mixed/other’ outcomes recalled in Study 1, which either involved several life domains or went beyond the major domains considered in the present study.

Limitations and future research

We focused primarily on outcome considerations and importance ratings. The next phase of this research will assess actual behaviours and decisions, aiming to better understand how various dimensions aggregate to shape individual choices. While we focused on the initial adoption of health behaviours, more complex processes, such as behaviour maintenance and habit change, may require expanding the current framework. Furthermore, certain psychological factors underlying health behaviours cannot be simply linked to a single outcome or dimension. For example, some important components for mainstream health behaviour models include self‐efficacy, knowledge of behaviour, and perceived control, which relate to the probability of the occurrence of all outcomes and dimensions via influencing the probability of the occurrence of the behaviour or decision. Future research should integrate these elements to better understand how they moderate individual perceptions of outcomes across different dimensions. Moreover, the responses in Study 1 were based on a diverse range of the behaviours and decisions. The average ratings may lean toward more commonly recalled scenarios while not adequately addressing other diverse scenarios. Different medical and health contexts may involve distinct decisional structures, such as variations in clinical severity, uncertainty, time sensitivity, and reversibility, which could differentially shape the salience and weighting of outcome dimensions. Meanwhile, although all participants rated the same set of 30 scenarios in Study 2, there was variability across scenarios in outcome importance ratings. This variability suggests that aggregating across heterogeneous scenarios may obscure scenario‐specific patterns of outcome prioritisation. Future research should address the heterogeneity across diverse behaviours.

A further limitation is that the results were based on U.S. samples and did not distinguish different health conditions. The importance of ratings across dimensions and profiles of rating patterns may not be the same in other countries that have different healthcare systems, health beliefs, health conditions, interpersonal dynamics, as well as social outcomes of health behaviours. Future research should expand to other populations using our framework. Additionally, both studies were self‐reported, and Study 2 was based on hypothetical scenarios. While hypothetical scenarios allow systematic comparison across a broad range of health decisions, participants' reported priorities may not fully capture the emotional intensity, contextual constraints, and real‐world trade‐offs present during actual health decision‐making, which may limit generalisability to real‐life behaviours. Future research should consider capturing actual, real‐time considerations using methods such as diaries or momentary assessments and investigate the broader consequences of prioritisation in health decision‐making. Finally, the identification of the profile numbers in the current study could be restricted to the current sample size. Future research should examine profiles in a larger number of participants to identify more subgroups, especially the minority in the population.

AUTHOR CONTRIBUTIONS

Yiyun Shou: Conceptualization; investigation; methodology; data curation; formal analysis; supervision; funding acquisition; writing – original draft; writing – review and editing; visualization. Joey Elizabeth Yeo: Methodology; data curation; investigation; formal analysis; project administration; writing – original draft; writing – review and editing. Alexander Shao‐Rong Pang: Investigation; data curation; writing – review and editing.

FUNDING INFORMATION

The study was funded by the National University of Singapore Start‐Up Grant.

ETHICS STATEMENT

The ethical aspect of this project was approved by the Institutional Review Board of the National University of Singapore (NUS‐IRB‐2022‐699 and NUS‐IRB‐2023‐515). Informed consent was obtained from all individual participants included in the study.

PREREGISTRATION

The preregistration protocol of the first study is on osf at: https://osf.io/5vwat.

Supporting information

Table S1. Categories of health behaviour described by participants in Study 1.

BJOP-117-1657-s001.docx (20.5KB, docx)

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.

REFERENCES

  1. Akl, E. A. , Oxman, A. D. , Herrin, J. , Vist, G. E. , Terrenato, I. , Sperati, F. , Costiniuk, C. , Blank, D. , & Schünemann, H. (2011). Framing of health information messages. Cochrane Database of Systematic Reviews, 2011, CD006777. 10.1002/14651858.CD006777.pub2 [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Augeraud‐Véron, E. , & Leandri, M. (2024). Optimal self‐protection and health risk perceptions: Exploring connections between risk theory and the health belief model. Health Economics, 33(7), 1565–1583. 10.1002/hec.4826 [DOI] [PubMed] [Google Scholar]
  3. Badr, H. , Oluyomi, A. , Woodard, L. , Zhang, X. , Raza, S. A. , Adel Fahmideh, M. , El‐Mubasher, O. , & Amos, C. A. (2021). Sociodemographic and health belief model factors associated with nonadherence to COVID‐19 mitigation strategies in the United States. Annals of Behavioral Medicine, 55(7), 677–685. 10.1093/abm/kaab038 [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Balla, J. , & Hagger, M. S. (2025). Protection motivation theory and health behaviour: Conceptual review, discussion of limitations, and recommendations for best practice and future research. Health Psychology Review, 19(1), 145–171. 10.1080/17437199.2024.2413011 [DOI] [PubMed] [Google Scholar]
  5. Bandura, A. (1986). Social foundations of thought and action: A social cognitive theory (p. 617). Prentice‐Hall, Inc. [Google Scholar]
  6. Bandura, A. (2004). Health promotion by social cognitive means. Health Education & Behavior, 31(2), 143–164. 10.1177/1090198104263660 [DOI] [PubMed] [Google Scholar]
  7. Bechard, L. E. , Bergelt, M. , Neudorf, B. , DeSouza, T. C. , & Middleton, L. E. (2021). Using the health belief model to understand age differences in perceptions and responses to the COVID‐19 pandemic. Frontiers in Psychology, 12, 609893. 10.3389/fpsyg.2021.609893 [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Bhuiya, T. , Klares, R. , Conte, M. A. , & Cervia, J. S. (2021). Predictors of misperceptions, risk perceptions, and personal risk perceptions about COVID‐19 by country, education and income. Journal of Investigative Medicine, 69(8), 1473–1478. 10.1136/jim-2021-001835 [DOI] [PubMed] [Google Scholar]
  9. Campbell, D. J. T. , Manns, B. J. , Weaver, R. G. , Hemmelgarn, B. R. , King‐Shier, K. M. , & Sanmartin, C. (2017). Financial barriers and adverse clinical outcomes among patients with cardiovascular‐related chronic diseases: A cohort study. BMC Medicine, 15(1), 33. 10.1186/s12916-017-0788-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Chen, M. S. , Cai, Q. , Omari, D. , Sanghvi, D. E. , Lyu, S. , & Bonanno, G. A. (2025). Emotion regulation and mental health across cultures: A systematic review and meta‐analysis. Nature Human Behaviour, 9, 1176–1200. 10.1038/s41562-025-02168-8 [DOI] [PubMed] [Google Scholar]
  11. Clark, S. L. , & Muthén, B. (2009). Relating latent class analysis results to variables not included in the analysis .
  12. Coifman, K. G. , Disabato, D. J. , Aurora, P. , Seah, T. H. S. , Mitchell, B. , Simonovic, N. , Foust, J. L. , Sidney, P. G. , Thompson, C. A. , & Taber, J. M. (2021). What drives preventive health behavior during a global pandemic? Emotion and worry. Annals of Behavioral Medicine, 55(8), 791–804. 10.1093/abm/kaab048 [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Conner, M. (2019). Models of health behavior. In Llewellyn C., Ayers S., McManus C., Newman S., Petrie K. J., Revenson T. A., & Weinman J. (Eds.), Cambridge handbook of psychology, health and medicine (3rd ed., pp. 55–60). Cambridge University Press. 10.1017/9781316783269 [DOI] [Google Scholar]
  14. Conner, M. , & Sparks, P. (2005). Theory of planned behaviour and health behaviour. In Conner M. & Norman P. (Eds.), Predicting health behaviour (2nd ed., pp. 121–162). Open University Press. [Google Scholar]
  15. Dempsey, R. C. , McAlaney, J. , & Bewick, B. M. (2018). A critical appraisal of the social norms approach as an interventional strategy for health‐related behavior and attitude change. Frontiers in Psychology, 9, 2180. 10.3389/fpsyg.2018.02180 [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. El‐Toukhy, S. (2015). Parsing susceptibility and severity dimensions of health risk perceptions. Journal of Health Communication, 20(5), 499–511. 10.1080/10810730.2014.989342 [DOI] [PubMed] [Google Scholar]
  17. Ferrer, R. A. , & Mendes, W. B. (2018). Emotion, health decision making, and health behaviour. Psychology & Health, 33(1), 1–16. 10.1080/08870446.2017.1385787 [DOI] [PubMed] [Google Scholar]
  18. Fraley, C. , Raftery, A. E. , Murphy, T. B. , & Scrucca, L. (2012). Mclust Version 4 for R: Normal mixture modeling for model‐based clustering, classification, and density estimation (Vol. 597). Technical Report. [Google Scholar]
  19. Gallagher, K. S. , Stojanovski, K. , Ogarrio, K. , Wright, L. , Fuster, M. , & Bell, C. N. (2025). Applications of equity frameworks in theory‐based health behavior interventions: A scoping review. International Journal for Equity in Health, 24(1), 79. 10.1186/s12939-025-02438-x [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. George, J. M. , & Dane, E. (2016). Affect, emotion, and decision making. Organizational Behavior and Human Decision Processes, 136, 47–55. 10.1016/j.obhdp.2016.06.004 [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Head, B. , Harris, L. , Kayser, K. , Martin, A. , & Smith, L. (2018). As if the disease was not enough: Coping with the financial consequences of cancer. Supportive Care in Cancer, 26(3), 975–987. 10.1007/s00520-017-3918-y [DOI] [PubMed] [Google Scholar]
  22. Helms, S. W. , Choukas‐Bradley, S. , Widman, L. , Giletta, M. , Cohen, G. L. , & Prinstein, M. J. (2014). Adolescents misperceive and are influenced by high‐status peers' health risk, deviant, and adaptive behavior. Developmental Psychology, 50(12), 2697–2714. 10.1037/a0038178 [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Hoffman, M. A. , & Driscoll, J. M. (2000). Health promotion and disease prevention: A concentric biopsychosocial model of health status. In Handbook of counseling psychology (3rd ed., pp. 532–567). John Wiley & Sons, Inc. [Google Scholar]
  24. Karimy, M. , Bastami, F. , Sharifat, R. , Heydarabadi, A. B. , Hatamzadeh, N. , Pakpour, A. H. , Cheraghian, B. , Zamani‐Alavijeh, F. , Jasemzadeh, M. , & Araban, M. (2021). Factors related to preventive COVID‐19 behaviors using health belief model among general population: A cross‐sectional study in Iran. BMC Public Health, 21(1), 1934. 10.1186/s12889-021-11983-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Kasting, M. L. , Haggstrom, D. A. , Lee, J. L. , Dickinson, S. L. , Shields, C. G. , & Rawl, S. M. (2021). Financial hardship is associated with lower uptake of colorectal, breast, and cervical cancer screenings. Cancer Causes & Control, 32(10), 1173–1183. 10.1007/s10552-021-01465-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Kim, T. J. , Vonneilich, N. , Lüdecke, D. , & Von Dem Knesebeck, O. (2017). Income, financial barriers to health care and public health expenditure: A multilevel analysis of 28 countries. Social Science & Medicine, 176, 158–165. 10.1016/j.socscimed.2017.01.044 [DOI] [PubMed] [Google Scholar]
  27. Kim, Y. , Park, I. , & Kang, S. (2018). Age and gender differences in health risk perception. Central European Journal of Public Health, 26(1), 54–59. 10.21101/cejph.a4920 [DOI] [PubMed] [Google Scholar]
  28. Kingsbury, J. H. , Gibbons, F. X. , & Gerrard, M. (2015). The effects of social and health consequence framing on heavy drinking intentions among college students. British Journal of Health Psychology, 20(1), 212–220. 10.1111/bjhp.12100 [DOI] [PubMed] [Google Scholar]
  29. Lerner, J. S. , Li, Y. , Valdesolo, P. , & Kassam, K. S. (2015). Emotion and decision making. Annual Review of Psychology, 66(1), 799–823. 10.1146/annurev-psych-010213-115043 [DOI] [PubMed] [Google Scholar]
  30. Lubke, G. , & Neale, M. C. (2006). Distinguishing between latent classes and continuous factors: Resolution by maximum likelihood? Multivariate Behavioral Research, 41(4), 499–532. 10.1207/s15327906mbr4104_4 [DOI] [PubMed] [Google Scholar]
  31. McEachan, R. R. C. , Conner, M. , Taylor, N. J. , & Lawton, R. J. (2011). Prospective prediction of health‐related behaviours with the theory of planned behaviour: A meta‐analysis. Health Psychology Review, 5(2), 97–144. 10.1080/17437199.2010.521684 [DOI] [Google Scholar]
  32. Mendez, I. , Strassle, P. D. , Ponce, S. , Le, R. , Stewart, A. L. , & Nápoles, A. M. (2024). Age‐related differences in the association between financial hardship and weight change during the COVID‐19 pandemic. Heliyon, 10(10), e30917. 10.1016/j.heliyon.2024.e30917 [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Miller, D. T. , & Prentice, D. A. (2016). Changing norms to change behavior. Annual Review of Psychology, 67(1), 339–361. 10.1146/annurev-psych-010814-015013 [DOI] [PubMed] [Google Scholar]
  34. Montgomery, S. C. , Donnelly, M. , Bhatnagar, P. , Carlin, A. , Kee, F. , & Hunter, R. F. (2020). Peer social network processes and adolescent health behaviors: A systematic review. Preventive Medicine, 130, 105900. 10.1016/j.ypmed.2019.105900 [DOI] [PubMed] [Google Scholar]
  35. Morgan, J. , Robinson, O. , & Thompson, T. (2015). Happiness and age in European adults: The moderating role of gross domestic product per capita. Psychology and Aging, 30(3), 544–551. 10.1037/pag0000034 [DOI] [PubMed] [Google Scholar]
  36. Morrison, D. M. , Mar, C. M. , Wells, E. A. , Gillmore, M. R. , Hoppe, M. J. , Wilsdon, A. , Murowchick, E. , & Archibald, M. E. (2002). The theory of reasoned action as a model of children's health behavior. Journal of Applied Social Psychology, 32(11), 2266–2295. 10.1111/j.1559-1816.2002.tb01863.x [DOI] [Google Scholar]
  37. Murgraff, V. , White, D. , & Phillips, K. (1999). An application of protection motivation theory to riskier single‐occasion drinking. Psychology & Health, 14(2), 339–350. 10.1080/08870449908407332 [DOI] [Google Scholar]
  38. Nicholls, R. , Perry, L. , Duffield, C. , Gallagher, R. , & Pierce, H. (2017). Barriers and facilitators to healthy eating for nurses in the workplace: An integrative review. Journal of Advanced Nursing, 73(5), 1051–1065. 10.1111/jan.13185 [DOI] [PubMed] [Google Scholar]
  39. Niño, M. , Harris, C. , Drawve, G. , & Fitzpatrick, K. M. (2021). Race and ethnicity, gender, and age on perceived threats and fear of COVID‐19: Evidence from two national data sources. SSM ‐ Population Health, 13, 100717. 10.1016/j.ssmph.2020.100717 [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Nour, K. , Laforest, S. , Gauvin, L. , & Gignac, M. (2006). Behavior change following a self‐management intervention for housebound older adults with arthritis: An experimental study. International Journal of Behavioral Nutrition and Physical Activity, 3(1), 12. 10.1186/1479-5868-3-12 [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Nylund‐Gibson, K. , & Choi, A. Y. (2018). Ten frequently asked questions about latent class analysis. Translational Issues in Psychological Science, 4(4), 440–461. 10.1037/tps0000176 [DOI] [Google Scholar]
  42. Ofori‐Ansah, S. , Evans, M. , Jones, J. , & Thomas, N. (2022). Decision‐making experiences of young adults with long‐term conditions. Journal of Renal Care, 48(1), 24–40. 10.1111/jorc.12367 [DOI] [PubMed] [Google Scholar]
  43. Pattison, A. B. , Reinfelde, M. , Chang, H. , Chowdhury, M. , Cohen, E. , Malahy, S. , O'Connor, K. , Sellami, M. , Smith, K. L. , Stanton, C. Y. , Voets, B. , & Wei, H. G. (2022). Finding the facts in an infodemic: Framing effective COVID‐19 messages to connect people to authoritative content. BMJ Global Health, 7(2), e007582. 10.1136/bmjgh-2021-007582 [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Pinkerton, R. , Mitchell, G. , & Hardy, J. (2020). Stringent control of opioids: Sound public health measures, but a step too far in palliative care? Current Oncology Reports, 22(4), 34. 10.1007/s11912-020-0900-5 [DOI] [PubMed] [Google Scholar]
  45. Ritchie, D. , Van Den Broucke, S. , & Van Hal, G. (2021). The health belief model and theory of planned behavior applied to mammography screening: A systematic review and meta‐analysis. Public Health Nursing, 38(3), 482–492. 10.1111/phn.12842 [DOI] [PubMed] [Google Scholar]
  46. Rosenstock, I. M. , Strecher, V. J. , & Becker, M. H. (1988). Social learning theory and the health belief model. Health Education Quarterly, 15(2), 175–183. 10.1177/109019818801500203 [DOI] [PubMed] [Google Scholar]
  47. Scrucca, L. , Fop, M. , Murphy, T. B. , & Raftery, A. E. (2016). mclust 5: Clustering, classification and density estimation using Gaussian finite mixture models. The R Journal, 8(1), 289–317. [PMC free article] [PubMed] [Google Scholar]
  48. Scrucca, L. , Fraley, C. , Murphy, T. B. , & Raftery, A. E. (2023). Model‐based clustering, classification, and density estimation using mclust in R. Chapman and Hall/CRC. 10.1201/9781003277965 [DOI] [Google Scholar]
  49. Sim, S. , Ho, S. , & Kumar, R. K. L. (2014). Use of opioids and sedatives at end‐of‐life. Indian Journal of Palliative Care, 20(2), 160–165. 10.4103/0973-1075.132654 [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Skinner, C. S. , Tiro, J. , & Champion, V. L. (2015). The health belief model. In Glanz K., Rimer B. K., & Viswanath K. (Eds.), Health behavior: Theory, research, and practice (5th ed., pp. 75–94). Jossey‐Bass/Wiley. [Google Scholar]
  51. Stok, F. M. , De Vet, E. , De Ridder, D. T. D. , & De Wit, J. B. F. (2016). The potential of peer social norms to shape food intake in adolescents and young adults: A systematic review of effects and moderators. Health Psychology Review, 10(3), 326–340. 10.1080/17437199.2016.1155161 [DOI] [PubMed] [Google Scholar]
  52. Stuber, J. , Galea, S. , & Link, B. G. (2008). Smoking and the emergence of a stigmatized social status. Social Science & Medicine, 67(3), 420–430. 10.1016/j.socscimed.2008.03.010 [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Sun, R. , Balabanova, A. , Bajada, C. J. , Liu, Y. , Kriuchok, M. , Voolma, S.‐R. , Đurić, M. , Mayer, C.‐H. , Constantinou, M. , Chichua, M. , Li, C. , Foster‐Estwick, A. , Borg, K. , Hill, C. , Kaushal, R. , Diwan, K. , Vitale, V. , Engels, T. , Aminudin, R. , … Sauter, D. (2024). Emotional experiences and psychological well‐being in 51 countries during the COVID‐19 pandemic. Emotion, 24(2), 397–411. 10.1037/emo0001235 [DOI] [PubMed] [Google Scholar]
  54. Van Den Bekerom, L. , Van Gestel, L. C. , Schoones, J. W. , Bussemaker, J. , & Adriaanse, M. A. (2024). Health behavior interventions among people with lower socio‐economic position: A scoping review of behavior change techniques and effectiveness. Health Psychology and Behavioral Medicine, 12(1), 2365931. 10.1080/21642850.2024.2365931 [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Yzer, M. (2017). Theory of reasoned action and theory of planned behavior. In Rössler P., Hoffner C. A., & Zoonen L. (Eds.), The international encyclopedia of media effects (1st ed., pp. 1–7). Wiley. 10.1002/9781118783764.wbieme0075 [DOI] [Google Scholar]
  56. Zhang, X. , Cowling, D. W. , & Tang, H. (2010). The impact of social norm change strategies on smokers' quitting behaviours. Tobacco Control, 19(Suppl_1), i51–i55. 10.1136/tc.2008.029447 [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Table S1. Categories of health behaviour described by participants in Study 1.

BJOP-117-1657-s001.docx (20.5KB, docx)

Data Availability Statement

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.

Articles from British Journal of Psychology are provided here courtesy of Wiley

RESOURCES