A theory-based examination of the predictors of mental health help-seeking in young men

ABSTRACT

Objective

The present research seeks to understand young men’s willingness to engage in both informal (social) and formal (professional) mental health support using the lens of the Prototype Willingness Model (PWM).

Method

The main study (N = 171) examined the constructs of attitude (positive or negative appraisal of the target behaviour), subjective norm (perceived social influence related to behavioural performance), prototype similarity (degree to which the individual feels similar to the typical person who perform the behaviour) and prototype favourability (degree to which the individual identifies the typical performer as being favourable), in addition to self-stigma and emotional expressivity. Hierarchical multiple regressions assessed these constructs in predicting young men’s willingness to engage in mental health support.

Results

Results indicated attitudes, prototype similarity, self-stigma and emotional expressivity explained a significant proportion of variance for willingness to engage in both informal support and formal help-seeking. Self-stigma was significantly associated with formal help-seeking, while emotional expressivity was significantly associated with informal and formal help-seeking.

Conclusions

The current research supports the value of the PWM in understanding males’ mental health help-seeking behaviour. The study enhances awareness of key factors which may inform young men’s willingness to engage in support services for their mental health.

Key Points

What is already known about this topic:

1. The engagement of young men with therapeutic services has been broadly attributed to internal attitudes held towards mental health help-seeking.

2. The conflict between a masculine identity of being independent and emotionally stoic and the openness, trust, and emotional expressiveness needed to be open to seeking mental health help is a potential barrier to mental health help seeking among young men.

3. Men are more likely to seek support for problems which can be attributed to an external cause rather than internal difficulties. Hence, self-stigma may act as a barrier to mental health help seeking for young men.

What this topic adds:

1. The application of an established decision-making model, the Prototype Willingness Model (PWM), was useful in the prediction of engagement with both formal and informal mental health help seeking among young men.

2. The addition of self-stigma and emotional expressivity to the PWM contributed to the predictions of engagement in formal support stemming from an external cause and internal feeling states.

3. Consistent with previous research, young men appear to have more difficulty in seeking help as a result of internal phenomenon. With the increase in people’s experience of mental health distress, the current research provides greater understanding of what factors underlie young men’s willingness to seek mental health support.

KEYWORDS:

• Emotional expressivity
• help-seeking
• mental health
• prototype willingness model
• self-stigma
• young men

Seeking support for mental health difficulties is a complex and multifaceted process (Cornally & McCarthy, 2011). Mental health help-seeking behaviour involves recognising a symptom or an indicator of distress, interpreting that symptom as in need of change and deciding to take action to gain support if the distress is perceived to be greater than the individual’s internal resources and ability to cope independently (Scott & Walter, 2010). This process does not occur in isolation as a multitude of social, cultural, and individual considerations also impact an individual’s decision to seek support for mental health difficulties (Rickwood & Thomas, 2012). While the perceived social stigma of seeking mental health support has reduced, a population of individuals who have been found to be difficult to engage in mental health support is young men aged 18–34 years (Cole & Ingram, 2020; du Plessis et al., 2009; Erentzen et al., 2018; Seidler et al., 2021). Previous research has found young men are reluctant to seek professional support in comparison to young women (Affleck et al., 2018; Erentzen et al., 2018). Additionally, young men often experience psychological distress and societal disadvantage as a product of mental health difficulties at a greater rate than that experienced by young women (Degney, 2012).

Young men’s experiences of mental health

The mental health of young people has significant consequences for social, economic and cultural progress (Compton & Shim, 2015; Degney, 2012). Therefore, the experiences of young people are fundamental in developing a more effective health system which promotes greater outcomes for those experiencing mental health distress. Young men have an increasing prevalence of mental health disorders, drug use and suicide ideation and completion (Degney, 2012; Seidler et al., 2021). Burns et al. (2013) found nearly one in five men felt life was not worth continuing in a general population sample. Indeed, completed suicide is the leading cause of death in young men within Australia and occurs at a rate three times greater than female completed suicide (Australian Bureau of Statistics [ABS], 2020). While men’s mental health has become increasingly targeted by media campaigns in recent years, some young men remain disconnected from mental health services (Degney, 2012; Silver et al., 2018).

Young men’s lack of engagement with mental health support, such as psychological therapy and counselling, has been well documented within the literature (Seidler et al., 2021; Silver et al., 2018). In a national survey on the mental health and wellbeing of Australians, only 13% of young men who were experiencing clinically significant levels of mental health disorder symptomology between the ages of 16 and 24 years sought professional support compared to one-third of young women in the survey (Slade et al., 2007). More recent estimates indicate that, even if young men become engaged in therapy, the drop-out rates for men are significant with one study indicating that 44.8% men disengage from services (Seidler et al., 2021). Despite living with significant distress, many young men do not seek support and, if they do become engaged, they feel as if they cannot be supported and discontinue services. Research has begun to focus upon reasons why men feel as if they are unable to seek support.

Barriers to engagement

The level of engagement young men have with therapeutic services has been broadly attributed to internal attitudes held towards mental health help-seeking (Bathje & Pryor, 2011). The nature of therapy requires individuals to be open to support, emotionally expressive and trusting, characteristics that Reed (2009) suggests are incompatible with masculinity as viewed by some young men. These characteristics may conflict with the deeply held views of what a man may believe of himself (Affleck et al., 2018). This concept is understood as the difficulty men experience when the widely held view of masculinity and a universal understanding of what a man should be causes internal strain (Vogel et al., 2007). This form of masculinity depicts men in a traditional role; that being of muscular stature, dominant, independent, powerful and emotionally stoic (Connell & Messerschmidt, 2005). Living within the expectations of this concept may be confusing for young men (Cole & Ingram, 2020; Hodgetts & Rua, 2010), especially when the expectations of such a role conflict with their own emotional distress. This conflict has been identified within the literature as a significant barrier to young men seeking support. Indeed, Wester et al. (2007) found that the greater the levels of restrictive emotionality and affectionate behaviour, the more psychological distress the individual was experiencing, the more the isolated the individual was from support.

This argument was supported in a study among 575 men enrolled in university whereby socialised gender roles played a significant role in seeking psychological supportin particular, the interaction of self-stigmatisation and discomfort with emotional self-disclosure (Pederson & Vogel, 2007). The authors found self-disclosure (being open to another person about emotional content) and self-stigma (negative perceptions of self due to help-seeking) as mediators within the dynamic between gender role conflict and perceptions of counselling. While both of these constructs have been previously supported as predictors of male help-seeking (Deane & Chamberlain, 1994; Mahalik et al., 2003), this research prompted further exploration of what the nature of stigma and self-disclosure is in relation to men’s mental health support.

Self-stigma occurs when the beliefs of the wider population become integrated within the individual, evoking shame or guilt for acting against public perception (Cole & Ingram, 2020; Corrigan et al., 2006). The more an individual perceives having a mental health difficulty as akin to weakness, the greater ambivalence the individual will experience in seeking support (Silver et al., 2018). As a result of this conflict, the decision to seek mental health support may be seen only as a final course of action (Gough, 2016). Self-stigma has been found as most prominent in topics that men perceive as most threatening to their sense of self-worth such as mental health. Interestingly, men are more likely to seek support for problems which can be attributed to an external cause, such as a relationship breakdown, rather than a difficulty which stems from an internal cause (e.g., generalised feelings of sadness) (Cole & Ingram, 2020). Cole and Ingram theorise that external causes related to help-seeking do not reflect upon the internal traits of an individual, and thus a man’s sense of self-worth is not threatened as directly. These authors further found a man’s willingness to seek support from friends and family was also limited by the presence of self-stigma.

In a qualitative examination of men’s views towards help-seeking for depression, Mahalik and Dagirmanjian (2019) found certain conditions reduce the stigma men associate with seeking support. The authors found the death of a family member was considered to be a legitimate reason to seek support for mental health and was seen as the cause of negative affect states such as depression. The authors noted that the typical reaction associated with death is sadness and feelings of depression. The external nature of the event provides a sense of legitimacy to the participant’s decision to seek help. A further condition which mitigated perceived stigma was a view that openness to help-seeking increased with age. Participants indicated the thought of seeking help in their early adulthood appeared more threatening as opposed to doing so at a later stage of life. Supporting these results is an exploration of the pathways men pursue to gain mental health support (Seidler et al., 2022). Results indicated younger men pursued support more commonly based on the encouragement of a family member, mostly in the context of anxiety (unspecified as to the origin of distress). Older men appeared to present for concerns related to relationships and workplace rather than a presenting difficulty relating to internal feeling states. Notably, this research demonstrated 40% of the total sample sought support for their mental health through self-initiation, thereby suggesting an increase in men accessing such services. Hence, the influence of self-stigma in help-seeking is not clear and may be mitigated by perceived legitimising external factors.

Difficulties with communicating emotional material is a major barrier for young men seeking mental health support (Vogel et al., 2007). Being able to share distressing feelings with others has been found to enhance protective relationships (Graham et al., 2008). As such, difficulty expressing emotions can be associated with fewer social supports, as well as a reduced intention and capacity to seek professional support. Indeed, by encouraging a young man to express emotional content may invoke feelings of distress for being unable to do so (Wilson, 2010). Young men may consider the distress or challenge in discussing affective content as not worth the potential gains of professional support, further reducing their help-seeking behaviour. Restrictive emotionality has been found to inversely predict perceptions of therapy effectiveness (Cusack et al., 2006). Compounding this dilemma is the intensity of symptoms (Vogel et al., 2007). As emotional distress increases, so does the perceived challenge presented by discussing one’s internal experience. Emotional expressivity is a core barrier which prevents young men from accessing professional support as the process of disclosure may be inherently painful and challenging for young men.

Facilitators of mental health support

While there have been many barriers identified in the literature for young men accessing mental health support, there are also facilitators for young men’s engagement in mental health services. Support and encouragement from social networks have been found to be the strongest predictor of help-seeking behaviour (Cusack et al., 2004). In a study of 2737 high school students, Pisani et al. (2012) found positive perceptions of help-seeking from adults were associated with increased disclosure of suicidal ideation and help-seeking. These findings provide support for interventions which challenge normative beliefs about help-seeking as a way of fostering likelihood of disclosure. Additionally, Pisani et al. demonstrated that the wider social environment can influence individuals’ inclination to reach out for support. Evidence for this finding has also been found in adult populations (Harding & Fox, 2015). Ultimately, the acceptance of mental health help-seeking among their social networks can facilitate men seeking mental health support. To date, however, there has been limited examination of this behaviour through a theoretical decision-making paradigm. Such a model may ultimately promote understanding more broadly of the predictors of mental health help-seeking and the underlying individual and social factors that shape the behaviour.

Decision-making models and help-seeking

There has been some previous exploration of help-seeking behaviours through predictive health models. Li et al. (2018) examined the extent to which the Theory of Planned Behaviour (TPB) (Ajzen, 1991) explained mental health help-seeking in a population of Australian university students. The Theory of Planned Behaviour is a predictive theoretical framework developed by Ajzen (1991) which proposes that behaviour is determined by people’s intentions towards the behaviour. Intention to perform a behaviour is predicted by attitude, subjective norm and perceived behavioural control (PBC). Attitude can be defined as a positive or negative evaluation of the specific behaviour. Subjective norms are the favourable or unfavourable feedback an individual perceives from others in relation to performing a behaviour (Fishbein & Ajzen, 1975). PBC is a measure of the individual’s belief in their ability to perform a specific behaviour and is also believed to predict behaviour directly (Ajzen, 1991). In the research by Li et al., their study incorporated the variables of public stigma and self-stigma in addition to the TPB constructs of attitudes, subjective norms and perceived behavioural control. Results indicated intention was a robust predictor of help-seeking. The relationship between intention and other predictors was mediated by attitudes.

While Li et al.’s (2018) study demonstrates the efficacy of theoretical models in predicting help-seeking for mental health, their study places emphasis on factors which are based on intentional action, preconceived expectations and beliefs. Much of the research has rested on the assumption that mental health help-seeking behaviour comes about through intentional or rational pathways. However, the more socially oriented and spontaneous components of the process have not been captured as thoroughly (Hammer & Vogel, 2013). Indeed, there has been limited examination of men’s experience of mental health help-seeking using a theoretical decision-making paradigm that includes more socially driven and social image-related constructs.

The prototype willingness model

The Prototype Willingness Model (PWM) (Gerrard et al., 2008; Gibbons et al., 2003) was developed to account for socially driven pathways in decision-making. This model assumes that there are parallel processes of decision-making which function independently of each other. These processes consist of a reasoned pathway which places emphasis on rational consideration of the identified behaviour prior to action. In addition to a rational pathway similar to the TPB focusing on attitudes, norms, and intentions, the social reactive pathway of the PWM assumes that some decisions regarding health behaviours are made largely in response to a supportive context and social influences. The reactive pathway consists of prototype images which is understood by the degree to which people feel similar to social others who perform the behaviour (the prototype) as well as the degree to which they favour such prototypes. This degree of similarity and favourability inform the individual’s willingness to perform an identified behaviour (Gibbons & Gerrard, 1995). Behavioural willingness is conceptualised as an individual’s openness to perform a behaviour within a specified context. This construct measures the individual’s response to risk in a set of circumstances and is predicted by several measures including attitude, subjective norm and prototype images. Drawing these parallel processes together, intention, informed by the reasoned pathway, and behavioural willingness, informed by the reactive pathway, both predict the target behaviour.

In the context of men’s help-seeking behaviour, it is thought that the typical prototypes men draw from may be consistent with traditional definitions of masculinity and hence may be instrumental in the internal conflict young men may experience in seeking help. Prototype images have been identified as a consistent predictor of willingness to perform health behaviours as documented by a meta-analytic review (Todd et al., 2016). The findings of this meta-analysis supported the reactive pathway as a strong predictor of the variance in the examined health behaviours. In comparison to the reasoned pathway, the reactive pathway explained 4.9% of variance beyond the contribution of the reasoned pathway. Based upon the reviewed literature and reported barriers to engagement, the current study explored the predictors underlying young men’s decision to access mental health support using the reactive pathway of the PWM which incorporates the impact of images of typical help-seekers. As this model allows for the incorporation of additional variables (Gerrard et al., 2008), self-stigma and emotional expressiveness were included within the predictive framework.

The PWM has also been established as a model which predicts help-seeking behaviour. Hammer and Vogel (2013) examined the dual pathway model for psychological help-seeking among college students to determine the utility of the reasoned pathway in addition to reactionary processes. The results of their study found willingness independently contributed a significant proportion of variance in willingness to engage in help-seeking beyond what was contributed by the reasoned pathway. These findings indicate the social reactive pathway accounted for the behaviour more so than intentional and reasoned processes. Further, the authors noted an inverse relationship with prototype favourability; the more favourable the prototype appeared, the less willing the individual was to seek support. It was concluded that the prototype image constructs for help-seeking function in a more unique manner as opposed to other health behaviours. The explanation for this relationship stems from the type of person who would typically attend therapy; in this context, an individual who was experiencing distress. In seeking professional support for mental health, an individual may want to distance themselves from an unfavourable prototype if they perceive the prototype as possessing unfavourable traits.

The current study

Past research indicates that men are experiencing an increase in mental health distress.

In addition to this increase, young men appear to engage in mental health help-seeking at lower rates than young women (Watsford & Rickwood, 2014). Many young men are left in a position where their subjective distress may be high; however, many do not seek support and therefore suffer in isolation (Ogrodniczuk et al., 2021). The current research aims to understand the underlying predictors which inform men’s decisions to seek mental health support using the theoretical decision-making framework of the Prototype Willingness Model (PWM).

The current study focused on help-seeking in two forms: informal help-seeking, support sought from social others within the individual’s life, and formal support, support of a professional within the mental health industry. Drawing from the literature which suggests young men may be more willing to seek help for externally caused distress rather than for internal distress (Cole & Ingram, 2020; Mahalik & Dagirmanjian, 2019; Seidler et al., 2022), the current research also included help-seeking in response to an external cause (e.g., a relationship break up) and internal distress (e.g., general feelings of sadness).

This research focused on the reactive pathway of the PWM as this pathway has been demonstrated as a consistent predictive model for help-seeking behaviours, suggesting that help-seeking may be more reactive in nature and vulnerable to relevant social images (Hammer & Vogel, 2013; Todd et al., 2016). As research clearly indicates that stigma and difficulty with emotional expressivity impact on young men’s willingness to engage in support (Corrigan et al., 2006), these constructs were also examined.

The application of this theory-based model will address a gap in the literature by identifying what underlies young men’s willingness to engage either in formal or informal mental health support for distress caused by external or internal factors. It is hypothesised that young men’s willingness to engage in mental health support will be predicted by the PWM constructs of attitude, subjective norm, prototype similarity and prototype favourability. Additionally, it is hypothesised that the constructs of self-stigma and emotional expressiveness will be significantly associated with young men’s willingness to seek help. Willingness to seek help will be examined for both informal (e.g., friend or family) and formal (e.g., mental health professional) mental health support and in the case of both external (e.g., relationship breakdown) and internal (e.g., generalised feelings of sadness) causes of distress.

Method

Participants

A total of 10 young male participants (M age = 25.8, SD = 3.10) completed the pilot study, and 171 young male participants (M age = 24.31, SD = 5.17) completed the main survey. Participants for the pilot survey were recruited via snowballing from existing networks. Participants for the pilot study were men aged between 18 and 34 years, residing in Australia.

Recruitment for the main survey occurred through a university study participant sign-up system which provides access to a cohort of undergraduate psychology students. This platform enables participants to be identified through an anonymous code to protect individual confidentiality. Additionally, participants were recruited through paid advertising from online media platforms such as Facebook. All participants were offered entry into a prize draw to win one of two $50 gift vouchers, with students offered the option of course credit rather than prize draw entry. No identifying information was retained from participants to protect individual privacy. Before starting the survey, all participants were taken to a participant information sheet which provided detailed information as to the nature of the study. At the end of this information sheet, they indicated their consent by agreeing to begin the survey. The participants were young men between the ages of 18 and 34 years old as individuals in this population are reported to have high levels of disengagement from mental health services (ABS, 2020; Seidler et al., 2021). The current research obtained ethical approval from the Queensland University of Technology Human Research Ethics Committee (ethics approval number: 2000001024).

Measures

Pilot study

To understand men’s perceptions of mental health and help-seeking, the pilot study included open-ended questions asking participants to describe what they would do if they were experiencing mental health difficulty. To inform the willingness scenarios presented in the main survey, open-ended questions which asked participants to describe events or situations where they would be encouraged to seek support for a mental health problem from friends and family as well as from mental health professionals. The examples provided by participants served as the basis for the willingness scenarios in the main survey.

Main survey

The main survey assessed participant demographics, PWM predictive factors (attitude, subjective norm, prototype favourability and similarity) (see Table 1), which were separated into formal and informal mental health help-seeking. The additional factors of self-stigma and emotional expressivity were also assessed. Items were predominantly assessed on a seven-point Likert-type scale with the exception of attitude which was assessed on a semantic differential scale. The scales were based on the PWM with items preceded by the phrase “If I were concerned about my mental health more than usual …” The self-stigma of help seeking (SSOSH) was used to capture the construct of self-stigma. Previous research has demonstrated high reliability for this scale (α =.91) (Vogel et al., 2006). The emotional expressivity scale (EES) was used to capture emotional expressivity construct (Kring et al., 1994). This scale has been well validated with previous research reporting high internal consistency (α =.83) (Mehak et al., 2021). For both informal and formal help-seeking, participants were provided with a willingness scenario informed by the pilot study. Each scenario described a situation where they were required to choose whether they would seek mental health support by either (1) talking with friends and family (informal support) or (2) seeing a mental health professional (formal support). Participants were then asked to rate their level of willingness to engage and not engage in formal and informal support. The first scenario focused on external events which may prompt an individual to seek mental health support (i.e., “Imagine if you were experiencing some problems with your partner and you found it difficult to focus on anything else”). The second scenario focused on internal feeling states of sadness which may affect quality of life (i.e., ‘Imagine you were feeling very sad and did not feel like going to work, spending time with friends or doing things you would usually). Negatively worded items were included to reduce response bias which were later recoded to be scored in a positive direction.

Table 1. Mental health help-seeking: scale items for the model constructs.

Variable	Item Example	Scoring or Rating	Scale Detail and Reliability
Attitude (2 items for each support type)	For me, if I were concerned about my mental health more than usual, seeking formal mental health support would be …	[1] favourable [7] unfavourable	Pearson’s r = .77 (Informal) Pearson’s r = .63 (Formal)
Subjective norm (2 items for each support type)	If I were concerned about my mental health more than usual, most of my friends would think I should seek informal mental health support if needed.	[1] strongly disagree to [7] strongly agree	Pearson’s r = .63 (Informal) Pearson’s r = .63 (Formal)
Prototype similarity (1 item for each support type)	In general, how similar do you feel like to the type of person your age who seeks formal mental health support.	[1] strongly dissimilar, [7] strongly similar	N/A
Prototype favourability (1 item for each support type)	In general, how favourable do you feel about someone your age who seeks formal mental health support.	[1] strongly unfavourable, [7] strongly favourable	N/A
Behavioural Willingness (2 items for each support type)	Imagine you were feeling very sad and did not feel like going to work, spending time with friends or doing things you would usually enjoy. How willing would you be to seek informal support (i.e., friends and family).	[1] not willing at all [7] very willing	Pearson’s r = .77 (Informal scenario 1) r = .60 (Informal scenario 2) Pearson’s r = .80 (Formal scenario 1) r = .57 (Formal scenario 2)
Self-stigma of Seeking Help (10 items)	I would feel inadequate if I went to a therapist for psychological help.	[1] strongly disagree to [7] strongly agree	Cronbach’s α =.86
Emotional Expressivity Scale (17 items)	Even if I am feeling very emotional, I don’t let others see my feelings.	[1] strongly disagree to [7] strongly agree	Cronbach’s α =.94

Scale reliability is shown for two behaviours: Informal = Informal help-seeking behaviours; Formal = formal help-seeking behaviours.

Results

Descriptive statistics

Of the total sample, 78% identified as heterosexual, 13% identified as homosexual, and 9% identified as other (including pansexual). The majority (50.9%) were not in a relationship, 30.6% were dating, 8.3% were married and 10.2% were in a de-facto relationship. Participants were mostly educated with 40% reporting a high school certificate, 31% a bachelor degree, 15.5% a diploma and 5.2% had postgraduate qualifications. Nearly half the participants (47.7%) reported working part-time or on a casual basis, 25.5% were employed full time and 27.4% were unemployed.

The majority of participants (66%) reported they had engaged in some form of informal support when they had concern about their mental health in the past, with 82% of this sub-sample indicating they had sought informal support in the last year. Additionally, 60.6% of participants had previously engaged in formal mental health support, with 69% of those who had engaged in this support having accessed formal support in the past year. When asked what participants would usually do to manage their mental health, 38.7% of participants reported they manage their mental health independently and 28.5% reported talking to friends and/or family. A minority (12.7%) reported that they would see a mental health professional, and 22.1% reported doing nothing to actively manage their mental health. It was noted that participants could report multiple coping strategies.

Bivariate correlations

Informal mental health help-seeking

Participant willingness to engage in this form of help-seeking was significantly and positively correlated with the PWM variables of attitude, subjective norm and prototype similarity for both informal and formal help-seeking. Prototype favourability was moderately correlated with willingness for informal help-seeking and significantly correlated with willingness for formal help-seeking. Emotional expressivity was significantly and positively correlated, whereas self-stigma was significantly and negatively correlated with willingness for both help-seeking conditions.

Predicting willingness to seek help

Predicting willingness with PWM and additional variables

Informal mental health seeking behaviours, external events

A hierarchical regression was conducted to predict willingness to seek informal support based on an external event which occurs in the context of a relationship. The PWM variables of attitude, subjective norm, prototype similarity and prototype favourability were entered in the first step of the regression. These constructs accounted for a significant proportion of the variance (39.4%) of willingness to seek informal mental health support, R² = .39, F (4, 142) = 23.04, p < .001. As seen in Table 2, attitude and prototype similarity have emerged as significant predictors of willingness. Self-stigma and emotional expressivity were entered in the second step of the regression but did not significantly contribute to the variance in willingness ΔR² = .02, F (2,140) = 2.65, p =.07. The full model accounted for 41.6% of the variance in willingness to seek informal support, R² = .42, F (2, 140) = 25.69, p < .001. Only attitude and prototype similarity emerged as significant positive predictors of the model in the final step. When the analysis was repeated with key demographics (i.e., age, sexual orientation and relationship status) entered in the first step, the same pattern of results was found (see Appendix A).

Table 2. Hierarchical multiple regression with PWM and additional constructs predicting willingness to engage in informal mental health behaviours for external event and internal feeling state scenarios (N = 171).

Step	Variable	B	SE	β	95% CI	sr^2	R^2	R^2 (Δ)
External Event
Step 1	Attitude	.30	.07	.31***	[.16, .43]	.09	.39***
	Subjective Norm	.10	.12	.06	[−.13, .41]	.00
	Prototype Similarity	.36	.07	.39***	[.21, .50]	.09
	Prototype Favourability	−.06	.08	−.05	[−.22, .10]	.01
Step 2	Attitude	.25	.07	.26***	[.10, .40]	.05	.41**	.02
	Subjective Norm	.09	.12	.05	[−.14, .32]	.00
	Prototype Similarity	.33	.07	.36***	[.19, .48]	.07
	Prototype Favourability	−.04	.08	−.08	[−.26, .06]	.02
	Self-Stigma	−.24	.15	−.12	[−.55, .06]	.01
	Emotional Expressivity	.13	.13	.08	[−.12, .39]	.03
Internal Feeling State
Step 1	Attitude	.31	.13	.30***	[.15, .48]	.09	.30***
	Subjective Norm	.14	.13	.08	[−.13, .41]	.00
	Prototype Similarity	.31	.08	.31***	[.14, .47]	.09
	Prototype Favourability	−.13	.09	−.10	[−.32, .05]	.01
Step 2	Attitude	.23	.08	.23***	[.06, .48]	.05	.35**	.04**
	Subjective Norm	.11	.13	.06	[−.15, .37]	.00
	Prototype Similarity	.27	.08	.28***	[.11, .44]	.07
	Prototype Favourability	−.18	.09	−.15	[−.37, .00]	.02
	Self-Stigma	−.22	.17	−.10	[−.57, .12]	.01
	Emotional Expressivity	.33	.14	.18*	[.03, .72]	.03

Β = standardised regression coefficient; sr² = squared semi-partial correlations. Statistical significance: *p < .05. **p < .01. ***p < .001.

Informal mental health seeking behaviours, internal states

A hierarchical regression was performed to predict willingness to engage in informal mental health support based on negative internal feeling states which impact life quality. Table 2 shows that the PWM variables entered at step 1 accounted for a significant amount of variance in willingness, R² = .30, F (4, 142) = 15.71, p < .001. In this step, attitude and similarity emerged as significant and unique predictors of willingness to engage in informal mental health support. Self-stigma and emotional expressivity were entered in the second step and also contributed to a significant proportion of variance ΔR² = .04, F (2, 140) = 4.89, p < .01, with emotional expressivity emerging as a significant predictor. The overall model predicted 35% of the variance in willingness to seek informal support R² = .35, F (2, 140) = 20.60, p < .01. Attitude, similarity and emotional expressivity emerged as significant predictors of willingness in the final step of the regression. When the analysis was repeated with key demographics (i.e., age, sexual orientation and relationship status) entered in the first step, the same pattern of results emerged (see Appendix A).

Formal mental health seeking behaviours, external events

A hierarchical regression was conducted to predict willingness to engage in formal mental health support in the context of an external event. The PWM constructs contributed to a significant proportion of variance in the first step, R² = .30, F (4, 142) = 15.59, p < .001. Attitude, subjective norm and prototype similarity emerged as significant predictors of willingness in this step (see Table 3). Self-stigma and emotional expressivity were then entered in the second step and significantly contributed to the variance, ΔR² = .03, F (2, 140) = 4.09, p < .05. Self-stigma emerged as a significant unique predictor in this step. The overall model explained a significant amount (34%) of variance in willingness, R² = .34, F (2, 140) = 15.59, p < .05. Attitude, similarity and self-stigma emerged as significant and unique predictors of willingness to seek formal mental health support in the context of an external event. When the analysis was repeated with key demographics (i.e., age, sexual orientation and relationship status) entered in the first step, the same pattern of results occurred with the exception of self-stigma which was now slightly beyond the stated threshold for significance (see Appendix B).

Table 3. Hierarchical multiple regression with PWM and additional constructs predicting willingness to engage in formal mental health behaviours for external event and internal feeling state scenarios (N = 160).

Step	Variable	B	SE	β	95% CI	sr^2	R^2	R^2 (Δ)
External Event
Step 1	Attitude	.27	.08	.25***	[.10, .43]	.04	.30***
	Subjective Norm	.30	.11	.19**	[.06, .53]	.06
	Prototype Similarity	.29	.07	.30***	[.15, .43]	.10
	Prototype Favourability	.02	.08	.02	[−.14, .19]	.00
Step 2	Attitude	.23	.08	.20**	[.05, .40]	.04	.34**	.03**
	Subjective Norm	.22	.11	.14	[−.00, .46]	.02
	Prototype Similarity	.25	.07	.25***	[.10, .39]	.07
	Prototype Favourability	−.04	.08	−.03	[−.21, .13]	.00
	Self-Stigma	−.38	.11	.18*	[−.55, .06]	.03
	Emotional Expressivity	.12	.13	.07	[−.13, .39]	.00
Internal Feeling State
Step 1	Attitude	.26	.08	.23***	[.09, .43]	.05	.29***
	Subjective Norm	.29	.11	.19**	[.06, .52]	.04
	Prototype Similarity	.29	.07	.30***	[.14, .43]	.10
	Prototype Favourability	.03	.08	.03	[.14, .43]	.00
Step 2	Attitude	.19	.08	.18*	[.02, .36]	.03	.40***	.10***
	Subjective Norm	.16	.11	.10	[−.05, .38]	.01
	Prototype Similarity	.21	.06	.22***	[.07, .34]	.06
	Prototype Favourability	−.06	08	−.05	[−.23, .09]	.00
	Self-Stigma	−.52	.16	−.26***	[−.85, −.20]	.06
	Emotional Expressivity	.31	.12	.19*	[.03, .72]	.04

Β = standardised regression coefficient; sr² = squared semi-partial correlations. Statistical significance: *p < .05. **p < .01. ***p < .001.

Formal mental health seeking behaviours, internal feeling states

The PWM constructs of attitude, subjective norm, prototype similarity and favourability contributed to a significant amount of variance (29%) in willingness to seek formal mental health support, R² = .29, F (4, 142) = 15.08, p < .001. As seen in Table 3, attitude, subjective norm and similarity emerged as significant unique predictors of the variance of willingness. Self-stigma and emotional expressivity were entered in the second step and contributed significantly to the proportion of variance explained in willingness, ΔR² = .40, F (4, 140) = 12.86, p < .001. Both self-stigma and emotional expressivity emerged as unique predictors. The overall model accounted for ΔR² = .40, F (4, 140) = 27.95, p < .001. Attitude, similarity, self-stigma and emotional expressivity emerged as significant, unique predictors at the final step. When the analysis was repeated with key demographics (i.e., age, sexual orientation and relationship status) entered in the first step, the same pattern of results obtained for the internal state analysis reoccurred (see Appendix B).

Discussion

The present research examined young men’s informal and formal mental health help-seeking in response to an external or internal stressor through the lens of the reactive pathway of the Prototype Willingness Model, an established decision-making model for engaging in health behaviours. The role of self-stigma and emotional expressivity were also examined. The results of the current study generally supported the hypothesis that the PWM constructs would be significant and positive predictors of willingness to engage in formal and informal mental health help-seeking. The results also supported the hypothesis that the additional constructs would significantly account for a proportion of variance in both formal and informal mental health help-seeking. However, differences were observed in the analysis of external events and internal feeling states. Overall, the results were generally supportive of the application of the PWM in predicting willingness in mental health help-seeking among young men, with some inconsistencies in relation to subjective norms and prototype favourability.

Predicting young men’s mental help-seeking willingness with the PWM

For both informal and formal mental health help-seeking, only attitude and prototype similarity emerged as significantly uniquely associated with willingness to seek help in both the external condition (i.e., relational breakdown) and the internal feeling state condition (e.g., general feelings of sadness). In all scenarios examining informal mental health help-seeking, subjective norms and prototype favourability were not significant predictors of willingness to seek informal help. This result suggests that men’s willingness to seek informal support is informed by the degree to which performance of the behaviour is positively valued and the degree to which men feel similar to others who seek mental health support. In contrast, the lack of support for subjective norms indicates that seeking informal support for mental health is not influenced by perceived social pressure to engage in the behaviour. A possible explanation of this finding may be there is limited exposure and promotion of men’s informal mental health help. Despite an increase in advertising and media campaigning targeting men’s mental health (Robertson et al., 2018), such efforts may not have been widely accepted. A normalisation of men’s mental health and engagement with support services may have yet to impact social pressures for men in this regard.

A prototype is defined as a mental representation of the aggregate traits which a typical person who would perform the behaviour possesses. Prototype similarity emerged as a consistent predictor of willingness whereas prototype favourability did not. It appears that men identify their own characteristics as being related to or similar to these typical others who engage in help-seeking. However, prototype favourability did not emerge as a significant predictor of willingness to seek informal help. As the representation of the prototypical other is informed by both individuals and groups (Hammer & Vogel, 2013), young men may perceive typical others who perform help-seeking behaviours as being negative and unfavourable. Thus, they may have the desire to distance themselves from an unfavourable prototype. Elliott et al. (2017) proposes that comparison to social others is more influential in the prediction of behaviour than the evaluation of others performing the behaviour. Therefore, the recognition of similarities between the individual and typical others who engage in informal help-seeking may have served as a stronger predictor of willingness than perceiving the behaviour as unfavourable.

The influence of self-stigma and emotional expressivity on willingness to seek mental health help

Self-stigma and emotional expressivity were included within the analysis to understand the impact these variables have on mental health help-seeking. For informal mental health help-seeking in external conditions (i.e., relationship conflict), self-stigma and emotional expressivity were not significant. For the informal mental health help-seeking condition stemming from an internal cause (i.e., sadness and depressive symptoms), only emotional expressivity emerged as significant. For formal help-seeking in the external condition (i.e., relationship conflict), only self-stigma emerged as significant. Finally, for formal help-seeking in the internal condition (i.e., sadness and depressive symptoms), both additional factors significantly contributed to behavioural willingness. Self-stigma became nonsignificant when demographic variables were introduced into the analysis. However, examination of confidence intervals indicated this construct was now slightly beyond the stated threshold for significance, suggesting this was a statistical artefact due to the introduction of more variables into the model rather than a theoretically meaningful finding. A more comprehensive account of the results and theoretical implications are detailed below.

Informal mental health help-seeking: external

For this willingness scenario, neither additional construct significantly predicted participant willingness to seek informal support. This result indicates men’s sense of self-stigma, their belief that help-seeking is unacceptable for the individual, does not become expressed in an individual’s willingness to perform informal help-seeking for external causes of distress. This result could be explained by the partner dynamic stated in the external scenario. In the context of a relationship conflict, a man may feel gaining support through informal means as more normative and, therefore, does not reflect their perception of self-stigma. This reasoning could be further compounded by the informal nature of the supports. A young man may not feel a sense of stigma by utilising informal supports. Murphy and Hennessy (2017) suggested a peer group can be a place which provides normality to express distressing experiences such as partner conflicts. This finding diverges from the research undertaken by Cole and Ingram (2020) who found self-stigma to be a barrier to young men connecting with informal supports in the context of experiencing depression. A reason for this discrepancy may relate to the internal nature of the distressing problem. Young men may find seeking support of peers or family for a relationship issue more normative and more acceptable because the issue, by its nature, is an externalised event which does not reflect as clearly on the individual’s characteristics. Consequently, it does not as directly pose a threat to a man’s sense of who he is (Cole & Ingram, 2020).

Similarly, an individual’s ability to communicate emotional material (emotional expressivity) does not manifest in their willingness to perform the behaviour in response to an external event. A possible reason for this outcome may again be related to the partner dynamic. Within a relationship, previous research has indicated men feel deficient in relational power, the ability to express their influence (Garfield, 2010). In the context of conflict, exerting influence in the relationship would involve the sharing of emotional content, a position that may be difficult for a man. As such, their lack of willingness to perform this behaviour may be reflective of the experience they have in their relationships whereby the male may not discuss their emotions as readily or with as much ease as his female partner (Garfield, 2010).

Informal mental health help-seeking: internal

In this willingness scenario, emotional expressivity emerged as a significant factor. This finding is consistent with previous research (Cole & Ingram, 2020) which identified that men are more comfortable seeking help for external reasons rather than internal feeling states. This finding can be further understood by a man’s sense of comfort with his own emotions (Berke et al., 2018). Indeed, some research suggests fear of emotion can become conditioned for men due to experiences of negative feedback or gendered norms. It is reasonable, therefore, to assume that in an environment where emotional expressivity is not merely encouraged but expected (Scott & Walter, 2010), a man’s ability and willingness to communicate their emotions would be a barrier. As indicated by the significance of emotional expressivity, the more comfort one has with communicating their emotional experience, the more willing the individual would be to engage in an informal support process. Consequently, when faced with an internal stressor such as a depressed mood state, a young man’s ability to communicate their emotional content would be a significant, yet negative predictor of willingness to engage in informal help-seeking.

Self-stigma did not emerge as a significant predictor of willingness to pursue informal help-seeking. Again, this result is a departure from previous research undertaken by Cole and Ingram (2020) who found high self-stigma decreases informal help-seeking, which was consistent across a number of domains such as family, friends and anonymous self-help avenues (e.g., online support). Taken with the significance of emotional expressivity, it appears young men in this sample find it more difficult to express their emotions to informal supports but do not perceive the act of doing so to be threatening to such a degree that it would activate their perception of self-stigma. This finding may be reflective of an evolving view of mental health and the degree to which informal help-seeking is becoming more normalised (Pisani et al., 2012; Seidler et al., 2022). Additionally, it appears internal characteristics such as an individual’s capacity to articulate emotion is more hindering to young men but an external factor such as self-stigma (perceived judgement originating from outside the individual) does not appear as influential. This discrepancy with past research does raise queries about how the quality of informal support networks influences the presence of self-stigma and whether a greater perception of safety in relationships can prevent against the presence of self-stigma.

Formal mental health help-seeking: external

For this willingness scenario, only self-stigma emerged as a significant predictor of willingness. This finding is consistent with the wider societal perceptions related to mental health (Gough, 2016) and its internalisation within men. It appears as if the thought of seeking formal mental health support for external situation conflicts with the belief that such a course of action is unacceptable (Owen et al., 2018). As such, a perception of stigma is elicited from the individual when posed with this scenario.

Formal mental health help-seeking: internal

Within this willingness scenario, both the additional variables emerged as significant predictors of willingness to seek mental health support in response to internal feeling states. This finding is in line with previous literature (Cole & Ingram, 2020), supporting the notion men’s willingness to seek formal mental health support is significantly affected by the perceived self-stigma involved in performing the behaviour as well as the degree to which they feel comfortable in communicating emotional content. A further compounding issue may relate to the additive impact of these two constructs. An individual’s perceived ability to put language to their emotional experiences may be compounded by their sense of self-stigma surrounding formal mental health help-seeking. Thus, a man experiencing distressing internal feelings may not be able to articulate this experience. In addition, when confronted with the thought of psychotherapy, the individual’s perception of stigma becomes enacted and the belief that a man should inhabit qualities of stoicism and emotional distance becomes strengthened. The man, therefore, is left in a position of not knowing how to put language to their internal experience as well as believing that this is an appropriate quality for a man to have. He may not have the ability to express emotion but also believes he should not. These dual pressures may ultimately inhibit their ability to disclose their pain, particularly in a formal therapeutic process (Silver et al., 2018).

Practical implications

The practical implications of these results relate primarily to the engagement of young men in supports for mental health distress and difficulty. The current study provides evidence for several factors which may serve as barriers to young men for mental health help-seeking. However, the finding that attitudes were significantly associated with willingness to seek both informal and formal help suggests that marketing campaigns encouraging young men to engage in an active appraisal of help-seeking could be effective in promoting help-seeking behaviour. Bringing men’s attention to the positive aspects of seeking support either informally or formally and that they can outweigh the disadvantages could be beneficial.

The current results also suggest young men carry a strong perception of stigma about accessing therapy, which is arguably consistent with the internalised views of the community regarding mental health support. Despite an increase in media targeting of men’s mental health (Seidler et al., 2021), a strong sense of stigma continues to be a barrier for young men. This finding is further supported by the non-significance of prototype favourability in the analysis. In addition to the conflict posed by acting against this stigma (i.e., to choose to enter therapy despite these beliefs), young men do not feel a strong sense of

favourability towards typical others who perform these behaviours. However, their sense of similarity to this typical other is an underlying contributor to willingness. Future media campaigns and advocacy may be more effective if greater focus were placed upon normalising the experience of mental health distress and the shared similarities people with mental health experiences have (i.e., having negative thoughts and anxiety). This promotion of a relatable image that focuses upon inclusion could assist men in overcoming these barriers to professional support.

For clinicians, understanding the degree to which young male clients may hold strong perceptions of self-stigma may assist with engagement. Psychometric tools such as the Self-Stigma of Help-Seeking Scale (Vogel et al., 2006) may be beneficial to understand a male client’s sense of self-stigma so it can be addressed in the process of therapy to forestall potential disengagement. Additionally, gaining an understanding of young men’s ability to communicate their emotions would also be useful. Measures such as the Emotional Expressivity Scale would give the clinician an understanding of their client’s capacity to communicate their emotional content (Kring et al., 1994). This measure may guide treatment and prompt the clinician to support the client in fostering their abilities to put words to their emotions.

Strengths, limitations and future directions

The application of the established predictive framework PWM to this important health behaviour is a strength of the study. The inclusion of self-stigma and emotional expressivity ensured that the current research built upon existing findings. The incorporation of pilot study to inform the behavioural scenarios is also a key strength ensuring the survey design was based upon participants’ lived experience to inform relatable scenarios as well as preferred terminology. There are also several limitations that require noting. This study is limited by the lack of generalisability of findings to other participant populations. A high percentage of participants had a history of seeking professional mental health support suggesting a potential self-selection bias; these men may be more open to seeking this avenue of support when confronted with conflict or distress. Furthermore, the sample size did not allow meaningful age-based comparisons among the various ages which comprise this cohort. While scale reliability was good for the study’s multi-item measures, the study is limited by the use of single-item scales for prototype similarity and favourability. A further limitation was the lack of condition checks within the measures to gauge participants’ understanding of the different foci for the distinct sets of measures (i.e., formal/informal support; internal/external feeling states). Due to the number of different conditions within the design of this study, future research may benefit from this inclusion to ensure accurate reflection on and responses to each condition. Future research could also incorporate constructs from the reasoned pathway such as intention to conduct a dual pathway analysis. Additionally, the Health Action Process Approach (Schwarzer, 2016) could also assist in future research endeavours. This model incorporates the constructs of action planning which would assist in identifying a participant’s direct planning for the behaviour, as well as coping planning, which captures an element of problem solving in the event that barriers impede engagement in the behaviour. Future research which includes a follow-up study to measure subsequent help-seeking behaviour would also provide stronger evidence for the impact of these variables on young men’s subsequent help-seeking behaviour.

Another direction future studies could take is the incorporation of other constructs in addition to the PWM variables. While self-stigma and emotional expressivity were shown to be predictive of willingness to seek formal support, other constructs which target a man’s sense of identity and masculinity may be useful to understand how seeking support affects their sense of self. This avenue of research may be particularly relevant for an older male cohort. As seeking support, particularly in the form of psychotherapy, has been demonstrated as inconsistent and counter to some men’s sense of identity (Affleck et al., 2018), a construct which taps into these aspects may have a significant impact in a man’s willingness to seek support. Further noteworthy constructs of exploration may include mental health literacy and help-seeking attitudes which may also inform the experience of men in seeking support for their mental health. In this vein, more thorough examination of the internal and external causes of distress could yield a greater clarification of how the origin of the presenting problem (i.e., internally identified or identified as a product of the individual’s environment) reflects in young men’s willingness to seek mental health support. In doing so, a clear delineation between what constitutes an internally identified symptom and an external cause of distress is recommended to avoid examination of the same construct.

To account for the lack of generalisability to a broader population of men, more thorough examination of participants’ past help seeking is strongly recommended. Including an accurate reflection of past behaviour could provide greater context to how past help-seeking interacts with both the constructs of the model but also any additional variables that may be included (e.g., self-stigma). This design may also capture whether the length of time or rate of frequency a man pursues help-seeking predicts their willingness to do so in the future (i.e., seeing a therapist for one session as opposed to 10, seeking support from family multiple times in the past as opposed to a few).

Conclusion

Through application of an established predictive model, the PWM, to predict engagement with formal and informal mental health support among young men, this study has uniquely contributed to the literature surrounding men’s mental health. The presence of two additional variables, self-stigma and emotional expressivity, within the analysis also contributed to the understanding of these behaviours. The findings of the study partially supported the application of the PWM to this behaviour, with willingness being significantly and consistently predicted by attitude and prototype similarity. Both the additional variables also influenced the decision to engage in formal support stemming from an external cause and internal feeling states. Self-stigma also influenced the decision to engage in informal support stemming from internal feeling states. The primary difference in these decision-making processes appears to be related to the cause of distress. Consistent with previous research, young men appear to have more difficulty in seeking help as a result of internal phenomenon. With the increase in mental health distress, the current research provides greater understanding of what factors underlie young men’s willingness to seek mental health support, an aspect of life which is fundamental both to the wellness of the individual and the functioning of the greater community in which they live.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Appendices Appendix A.

Hierarchical Multiple Regression with Demographic Variables, Prototype Willingness Model, and Additional Constructs Predicting Willingness to Engage in Informal Mental Health Behaviours for External Event (N =145) and Internal Feeling State Scenarios (N =145).

Table

Step	Variable	B	SE	β	95% CI	sr^2	R^2	R^2 (Δ)
External Event
Step 1	Age	.25	.30	.08	[−.35, .85]	.00	.00
	Sexual Orientation	−.05	.32	−.01	[−.69, .58]	.00
	Relationship Status	−.08	.26	−.02	[−.61, .43]	.00
Step 2	Age	.33	.24	.01	[−.21, .80]	.00	.41***	.41***
	Sexual Orientation	.29	.25	.07	[−.21, .80]	.00
	Relationship Status	−.19	.21	.06	[−.61, .22]	.00
	Attitude	.38	.07	.35***	[.18, .48]	.12
	Subjective Norm	.08	.12	.05	[−.16, .32]	.00
	Prototype Similarity	.35	.07	.39***	[.21, .50]	.14
	Prototype Favourability	−.05	.08	−.05	[−.22, .10]	.00
Step 3	Age	.31	.23	.10	[−.16, .78]	.00	.44***	.02***
	Sexual Orientation	.28	.25	.05	[−.70, .16]	.00
	Relationship Status	−.27	.21	−.08	[−.70, .16]	.01
	Attitude	.27	.07	.29**	[.12, .43]	.08
	Subjective Norm	.05	.12	.03	[−.19, .29]	.00
	Prototype Similarity	.34	.07	.36***	[.19, .47]	.14
	Prototype Favourability	−.09	08	−.08	[−.25, .07]	.00
	Self-Stigma	−.20	.15	.12	[−.07, −.47]	.01
	Emotional Expressivity	.20	.13	−.10	[−.51, .10]	.01
Internal Feeling State
Step 1	Age	.04	.32	.01	[−.60, .69]	.00	.02
	Sexual Orientation	−.64	.34	−.15	[−.13, .04]	.00
	Relationship Status	.02	.28	.00	[−.53, .58]	.00
Step 2	Age	.11	.23	.03	[−.43, .65]	.00	.31***	.28***
	Sexual Orientation	−.31	.30	.07	[−.90, .28]	.00
	Relationship Status	−.04	.24	−.01	[−.53, .44]	.00
	Attitude	.31	.08	.30***	[.13, .48]	.08
	Subjective Norm	.14	.14	.08	[−.13, .48]	.01
	Prototype Similarity	.31	.08	.31***	[.14, .48]	.08
	Prototype Favourability	−.13	.09	−.10	[−.32, .06]	.01
Step 3	Age	.08	.27	.02	[−.45, .62]	.00.36 **	.05**
	Sexual Orientation	−.41	.29	−.10	[−.98, .16]	.01
	Relationship Status	−.19	.25	−.05	[−.69, .30]	.01
	Attitude	.21	.09	.21**	[.03, .39]	.04
	Subjective Norm	.08	14	.05	[−.19, .38]	.08
	Prototype Similarity	.27	.08	.28***	[.11, .44]	.07
	Prototype Favourability	−.17	.09	−.14	[−.36, .01]	.02
	Self-Stigma	−.23	.18	−.10	[−.59, .12]	.01
	Emotional Expressivity	.38	.15	.20*	[.05, .68]	.03

Β = standardised regression coefficient; sr² = squared semi-partial correlations.

Statistical significance: *p < .05. **p < .01. ***p < .001.

Appendix B.

Hierarchical Multiple Regression with Demographic Variables, Prototype Willingness Model, and Additional Constructs Predicting Willingness to Engage in Formal Mental Health Behaviours for External Event (N =154) and Internal Feeling State Scenarios (N =154).

Table

Step	Variable	B	SE	β	95% CI	sr^2	R^2	R^2 (Δ)
External Event
Step 1	Age	.48	.30	.16	[−.10, .98]	.01	.02
	Sexual Orientation	−.19	.32	.05	[−.43, .83]	.00
	Relationship Status	−.35	.29	−.11	[−.36, .67]	.00
Step 2	Age	.33	.22	.10	[.20, −.17]	.01	.33***	.30***
	Sexual Orientation	.42	.27	.11	[−.11, .96]	.01
	Relationship Status	.04	.22	.01	[−.39, .48]	.00
	Attitude	.27	.07	.25***	[.10, .41]	.06
	Subjective Norm	.39	.11	.19*	[.06, .53]	.04
	Prototype Similarity	.29	.07	.30	[.15, .42]	.10
	Prototype Favourability	.03	.08	.03	[.13, .21]	.00
Step 3	Age	.35	.35	.11	[−.14, .82]	.01	.36*	.03*
	Sexual Orientation	.33	.26	.08	[−.20, .86]	.01
	Relationship Status	.00	.22	.01	[−.44, .45]	.00
	Attitude	.23	.09	.21**	[.05, .39]	.04
	Subjective Norm	.20	.12	.14	[−.01, .45]	.01
	Prototype Similarity	.25	.07	.26**	[.10, .39]	.08
	Prototype Favourability	−.02	.08	−.20	[−.20, .15]	.02
	Self-Stigma	−.35	.18	−.17	[−.73, −.01]	.01
	Emotional Expressivity	.14	.11	.08	[−.13, .44]	.00
Internal Feeling State
Step 1	Age	−.08	.30	−.02	[−.68, .50]	.00	.01
	Sexual Orientation	−.05	.31	−.10	[−.68, .54]	.00
	Relationship Status	−.35	.26	.11	[−.16, .45]	.01
Step 2	Age	.26	.25	−.08	[−.75, .26]	.00	.31***	.30***
	Sexual Orientation	−.16	.23	.04	[−.36, .70]	.00
	Relationship Status	−.24	.22	.08	[−.19, .44]	.00
	Attitude	.25	.08	.23***	[.08, .45]	.05
	Subjective Norm	.30	.11	.19**	[.07, .54]	.04
	Prototype Similarity	.29	.07	.31***	[.15, .46]	.10
	Prototype Favourability	.03	.08	.03	[−.13, .21]	.01
Step 3	Age	.19	.24	.06	[−.67, .27]	.00.41***	.09***
	Sexual Orientation	.02	.24	.00	[−.47, .53]	.00
	Relationship Status	.15	.21	.05	[−.47, .53]	.01
	Attitude	.18	.08	.17*	[.01, .35]	.03
	Subjective Norm	.18	.11	.11	[−.04, .40]	.01
	Prototype Similarity	.22	.07	.23***	[.08, .35]	.06
	Prototype Favourability	−.06	.07	−.05	[−.22, .54]	.01
	Self-Stigma	−.53	.17	−.26**	[−.87, .20]	.06
	Emotional Expressivity	.28	.13	.18*	[.02, .54]	.03

Β = standardised regression coefficient; sr² = squared semi-partial correlations.

Statistical significance: *p < .05. **p < .01. ***p < .001.