The need to seeing oneself as socially connected 'for social belonging' is a fundamental human motivation (Walton & Cohen, 2007). Groups provide us with a sense of social identity; individuals' internalised sense of their membership in a particular group (Tajfel, 1972). Social identity has implications for multiple forms of social responses including social influence, social perception and social judgement (Haslam & Reicher, 2006; Haslam, Jetten, O'Brian & Jacobs, 2004; Patterson, Foster & Bellmer, 2001). Social identity with a group can give us a sense of grounding, enhance our self-esteem as well as buffer our well-being when threatened. It can also help people cope with the negative consequences of being a member of a devalued group (Haslam, Jetten, Postmes & Haslam, 2008).
Members of socially disadvantaged groups face a number of threats to their psychological well-being, and often report experiencing negative treatment across a wide variety of life domains (Outten, Schmitt, Garcia & Branscombe, 2009) This negative treatment can often manifest into stress. Stress is the psychological and physiological state of a person responding to the demands of strain imposed on them by stressors in the environment. A stressor is a feature of the environment that has potential to place demands on a person that is perceived by them to be threatening to the self and well-being (Haslam, 2004). Identity salience plays an important role in determining both whether and how people cope with different stressors. Specifically, if a given social identity is salient, it acts as a determinant of a person's responses to external stressors (Fischer, Haslam & Smith, 2010).
The present study is an elaboration of research carried out by Haslam et al. (2004) where university students participated in an experiment to investigate the role that social influence plays in the appraisal of a stressful situation. It was found that the appraisal of stress was affected by the group membership of the source providing the feedback, in this case a group with a positive social identity. This research draws on concepts from the social identity approach, specifically self- categorization theory (Tajfel, 1981) in explanation of results. The approach stipulates that individuals mentally place themselves in social categories that are salient for them. Self categorization motivates people to achieve a positive social identity in comparison with relevant others, thus maintaining a positive self- esteem.
This study aims to expand on these findings by investigating a population that have a negative social identity. There is an obvious gap in literature regarding research on the unemployed community, specifically investigating the role that social influence plays in the appraisal of a stressful situation. Work provides people with a positive social identity which is threatened or removed by unemployment. Many of the negative consequences people in unemployment experience may be explained by the loss of a positive social identity, leading to numerous psychological and physical effects (Cassidy, 2001). The financial crises of 2007-2008 led to the biggest global recession since the Great Depression of the 1930s, increasing unemployment rates globally (Choudhry, Marelli, & Signorelli, 2010). The impact of social identity with the unemployed community will be examined in this exploratory study. This will be conducted by examining an appraisal process that impacts on perceptions of stress focusing on concepts from the social identity approach (Tajfel, 1978; Tajfel & Turner, 1979, Tajfel, 1981) and Lazarus' (1991) cognitive relational theory of stress and coping.
The Social Identity approach
Festinger (1954) postulates that self- evaluation can rarely occur objectively, by necessity it relies heavily on comparisons between the self and others. The social identity approach was heavily influenced by Festinger's beliefs regarding social comparison. A social identity framework claims that individuals see the world from the perspective of fellow ingroup members. By doing so, relevant meanings associated with the in-group identity come to the fore (Cohen & Garcia, 2005). This forms a socially constructed sense of who and what 'we' are as well as who and what 'we' are not (Oyserman, Fryberg, & Yoder, 2007). By identifying with a social category, people evidence emotional involvement, commitment and a common fate with other members of this category (Tajfel & Turner, 1986). If a given social identity is salient, it is argued to be a powerful motivator of social perception and behaviour (Haslam & Reicher, 2006). The social identity approach therefore claims that groups are not just a passive context for individual behaviour; they mediate it to a large extent. Two theories encompass the social identity approach: Social identity theory (Tajfel, 1978; Tajfel & Turner, 1979) and self-categorisation theory (Tajfel, 1981).
Social identity theory (Tajfel, 1978; Tajfel & Turner, 1979) maintains that an individual's social identity is an essential component in the formation of the self-concept. Social identity theory (SIT) proposes that people strive to achieve or maintain a positive social identity. This positive identity is derived from favourable comparisons made between the ingroup and relevant outgroups (Brown, 2000). This positive identity expresses itself in the form of a positive differentiation on perceptions, attitudes and behaviour (Tajfel, 1978). In terms of SIT, a disadvantaged or inferior position of one's own group leads to a negative social identity (Tajfel & Turner, 1986). The unemployed community are attributed with a negative social identity; it has been shown that society tends to blame the unemployed person for their circumstances. These actions attribute negative stereotyping of the unemployed as a group (Kelvin & Jarret, 1985; Cassidy, 2001).
With regard to understanding how and when particular social identities become salient and the consequences of this, one must look to self-categorization theory (SCT) for answers (Wharton, 1992). According to SCT (Tajfel, 1981), the self can be defined at different categorical levels depending on the social context. Consequences of social comparison with another ingroup member will depend on the context; whether it encourages self- construal at the individual or group level. Social identity and self-categorization theories illustrate how individuals come to define themselves through the social groups to which people belong, forming social identities which are essential parts of the self-concept (Tajfel, 1978). The negative stereotype associated with unemployment suggests that it for most people, it is undesirable to be categorised as unemployed, thus it is not an easily chosen self-categorization (Cassidy, 2001). McFadyen (1995) argues that self-categorization as unemployed can impact on the ways people cope with life during unemployment.
Stress and its effects on the unemployed
Psychological stress "is a particular relationship between the person and the environment that is appraised by the person as taxing or exceeding his or her resources and endangering his or her well-being" (Lazarus & Folkman, 1984, p.19). Strain arises when imposed demands are perceived to exceed ability to cope. The literature on the psychological and physical effects of unemployment has consistently shown that the unemployed are more distressed than their employed counterparts; it has a predominantly negative effect on the person as a whole.
Research has indicated that the individuals face devaluation and rejection on the premises of their disadvantaged group membership (Schmitt & Branscombe, 2002). Outten (2005) has noted that incidences of mistreatment as well as perceptions of disadvantage may be conceptualised as stressors; events that produce tension in peoples' lives. The consequences of stressors relating to unemployment are deep and far reaching. Research indicates that unemployment can lower individuals' moral and lead to increased personality instability (Eisenberg & Lazarsfeld, 1938), physical and mental ill-health (Stokes & Cochrane, 2010) as well as issues such as family instability and violence (Hanisch, 1999). Furthermore, unemployment can cause increased depression and lower self- esteem (Waters & Moore, 2001). In Ireland, research has shown that high unemployment rates contribute to drug use, crime and frustrations connected to political conflict (Carlson, Fellows & Maslach, 1989). Moreover, it is argued that the consequences of stressful circumstances are determined by individuals' appraisals of these circumstances as well as coping mechanisms they choose to engage with (Outten et al., 2009).
In 1966, Lazarus first presented a comprehensive stress theory which suggested that stress was organizing concept for understanding a wide range of phenomena, consisting of many variables and processes (Lazarus & Folkman, 1984). He conducted pioneering studies demonstrating that cognitive evaluation of noxious stimuli (e.g. films of accident victims) determined the nature of emotional and physiological responses (Tomaka, Blascovch, Kibler & Ernst, 1997). Lazarus (1991) is the latest revision of the Lazarus stress theory; it is a cognitive relational theory that can be applied to all aspects of a personal life. 'Relational' is defined in terms of emotions being always about person environment relationships that involve harms and benefits. Therefore, psychological stress and emotions are transactions between individuals and their environments that change over time and circumstance.
Two concepts are central to Lazarus' psychological stress theory; appraisal and coping. These concepts mediate the relationship between a person's environment and his/her adaptational outcomes. Adaptation refers to the continual interplay between appraisal and coping where people manage their environment to maintain an optimum level of physical, psychological and social well-being (Hart & Cooper, 2001). Appraisal is an individual's evaluation of what is happening for their well-being, coping is the individuals' efforts to manage specific situations in thought and action (Krohne, 2002).
Lazarus (1991) distinguished two forms of appraisal; primary and secondary. Primary appraisal concerns whether something of relevance to the individual occurs in the encounter, there is no potential for an emotion. Secondary appraisal concerns the options and prospects for coping (Krohne, 2002; Lazarus, 1991). Results from Lazarus' (1966) original stress experiments concluded that different cognitive appraisals mediated the relationship between the stressor and stress reactions. As mentioned previously, coping is a central concept in Lazarus' psychological stress theory; hence is a key factor in stress-relevant person- environment transactions (Krohne, 2002). Coping has been defined by Lazarus and Folkman (1984) as the cognitive and behavioural process of managing internal or external demands that are appraised as taxing or exceeding the resources of a person.
Self-Categorization model of stress: Coping
The role of perceived social support in coping with stress has been well established in literature. Research implies that the emotional benefits of social support are important in the effective management of stressors. A social support network can provide people with emotional support, instrumental support, companionship and informational support (Haslam, 2004). The experience of stress is bound up with the facts of group life; moreover the social identity approach has shown that group processes can serve to lessen the impact of stress through the provision of support (Haslam, 2004). A self-categorization model of stress argues that secondary appraisal of one's ability to cope with a stressor should vary as a function of social identity salience. Haslam (2004) argues that self-categorisation processes should have a major impact on the dynamics of social support. Aspinwell and Taylor (1997) suggest that informational and appraisal supports are very important in the proactive stage of coping for the individual, as opposed to emotional support. This can increase their understanding of a situation (primary appraisal) as well as their ability to cope with it (secondary appraisal). The exchange of information within a social network enables individuals to acquire new interpretations, clarify their understanding of a situation and exert influence over their response to stressful situations (Haslam et al,, 2004; Haslam, 2004). Haslam et al., (2004) illustrated that shared social identity provides a basis for effective social support. Social identity stipulates that the effects of a stressful experience can be buffered by identification with a meaningful ingroup. This buffering occurs as a result of increased perceptions of social support from the ingroup (Fritsche, Janas & Fankhänel, 2008).
Several studies have advocated the findings by Haslam et al., (2004). A study by Tarrant, Dazeley and Cottom (2009) investigated the effects of social categorization on the experience of empathy. University students reported stronger empathy and helping intentions when the student in question was attending the ingroup university, as opposed to an outgroup university. Furthermore, it was found that bias against the outgroup members could be overcome by activating an ingroup norm which prescribes empathy for others. This study demonstrates the importance of social context in the development of positive attitudes towards outgroups. Haslam, Jetten, Postmes and Haslam (2009) examined the relationship between social identity, health and well-being. They found that social identity theorising has huge capacity to inform practical strategies aimed at maintaining and enhancing well-being, particularly among at risk populations as well as determining who people are and what they are capable of. This cements the notion that social identities are a fundamental part of our lives and are central to our well-being.
Importance of Group Membership
There is a large amount of research to suggest that among disadvantaged groups, the impact of stress may be reduced through the provision of social support. The decline in one's psychological well-being caused by unemployment is affected by the social support available to the unemployed individual (Lev-Wiesel & Kaufman, 2004). Schmitt and Branscombe (2002) examined the relationship between perceived discrimination and the positions of one's group in the social structure. For members of disadvantaged groups, attributions of prejudice were found to be internal, stable, uncontrollable, in turn conveying widespread exclusion and devaluation of one's group. Members of privileged groups attribute prejudice more locally. Therefore, attributions to prejudice are more harmful for members of disadvantage groups. By increasing identification with their disadvantaged social group, members cope with experiences of prejudice. Social support may be provided in the form of shared time and resources, countering the psychological cost of feeling rejected by the dominant culture by provision of a sense of belonging and acceptance. Therefore, by increasing psychological investment with the ingroup, members of disadvantaged groups feel accepted, thus enhancing their mental well-being.
Gore (1978) hypothesized that social support modifies the relationship between unemployment stress and health responses. She found that the rural unemployed evidenced a significantly higher level of social support than did the urban unemployed. Significantly, the unemployed participants who had less social support experienced more ill health effects. This study demonstrates the exacerbation of life stress caused by a low sense of social support, resulting in health effects.
Â In a longitudinal, comparison study, Bolton and Oatley (1987) examined the relationship between social support and depression. It was found that unemployed participants were significantly more depressed then the employed ones, with further analysis showing a strong interaction between employment and social support. Depression was found to be more likely when a source of social interaction that is important for a sense of worth is lost, and when there is no alternative means of expressing this worth in other relationships. The work environment is seen to provide people with a positive social identity; the removal of work therefore results in that positive social identity being threatened or removed.
Kokko, Pulkkinen and Puustinen (2000) examined longitudinally, selection into unemployment on the basis of those individuals' characteristics which reflect poor emotional regulation and psychological distress. They found that low self-control of emotions, especially aggression at young age directly predicted long-term unemployment in adulthood. Passive and anxious behaviour also predicted long-term unemployment indirectly; through low educational achievement.
Another mediating factor in the stress process is the perception of control the individual feels he or she has. Cassidy (2001) states that individuals who perceive control of events as beyond their group experience more stress than individuals who perceive themselves as being in control of events. Moreover, individuals who have a positive problem- solving style cope better with stress. As noted previously, the unemployed community struggling to cope with a lost social identity, as well as stigma associated with the unemployed community. One way to reduce levels of stress experienced may be to engage in proactive coping mechanisms. Literature regarding proactive coping highlights its purpose in avoiding potential stressors. Proactive coping efforts are undertaken to modify or prevent a potentially stressful event before it occurs. This coping mechanism may be of particular importance for members of the unemployed community. Unemployment can result in stress of long term duration, by engaging in proactive efforts one may avoid or offset stressful events- keeping chronic stress at low levels (Aspinwell & Taylor, 1997). Personal control has shown to be associated many positive outcomes including with lower reactivity to stressors in daily life (Diehl & Hay, 2010).
Overview of this study
This exploratory study will examine if members of the unemployed community, a stigmatised group, are influenced by informational support as demonstrated in Haslams' (2004) study. It is imperative that the gap in research regarding the unemployed community and self-categorization is lessened, as unemployment rates have raised significantly in the recent years. Previous research gives a clear indication that members of the unemployed community are more vulnerable then employed counterparts to stressors in life. By examining the role that shared negative group membership plays in the appraisal process that impact perceptions of stress for this vulnerable community, the researcher hopes to advance knowledge in this largely non-researched paradigm.
The literature reviewed provides ample evidence to suggest that a shared social identity provides a foundation for shared social support. A stressful experience can be buffered by identification with a meaningful ingroup. Mirroring Haslam et als' (2004) findings, it is hypothesised in this study that social identification as unemployed will be associated with perceptions of increased social support, and reduced stress. Secondly, the appraisal of stress will be affected by the group membership of a source that provides stress related feedback. Thirdly, in line with research carried out in this area (Cassidy, 2001) it is predicted that there will be differences in levels of identification and perceived stress between those who have been unemployed less than one year and those in long term unemployment.
Participants were 171 unemployed men and women, recruited from Limerick city Social Welfare Office. 20 cases were omitted from the data set as a manipulation key to the study was not comprehended correctly. The final sample was 151 unemployed individuals, 91 (60.3%) men and 60 (39.7%) women. The age range was between 20 and 64 years (Mage=35.45 years, SD=10.24). 68 participants were long term unemployed (>12 months) and 83 were short term unemployed (<12 months). Occupations of participants immediately prior to unemployment were primarily in the business and financial service (25.8%), building and construction (22.5%) and shops and retail sectors (20.5%). Educational demographics were as follows; one had no schooling, 16 had some secondary education, 36 completed secondary education, 45 had some third level education, 40 had completed third level and 17 participants had completed postgraduate studies. One third of participants were married. The majority of the samples' ethnicity was Irish. All partook voluntarily in the study in compliance with UL EHS Ethics Board.
This research followed a 2 x 2 x 2 between subjects design among 151 participants. They were approached while queuing outside the Limerick city Social Welfare Office. Participants were randomly assigned to a two (source: ingroup vs. outgroup) x two (message: stressful vs. challenging) factorial design. There were 34 in the (outgroup: stressful) condition with 39 in each of the remaining conditions (outgroup: challenging), (ingroup: stressful), (ingroup: challenging). The independent variables were the message source either ingroup or outgroup, whether the exam was stressful or challenging and level of identification either high or low. The dependant variable was the level of stress experienced, measured by responses to the three stress scales. Participants completed the survey in one sitting.
The questionnaire began with basic demographic questions including information relating to the participants' length of unemployment and previous occupation. Following this, they indicated their level of group identification with the unemployed community using Doosje, Ellemers and Spears (1995) global measure of identification. The succinct nature of the scale makes it suitable as a measure of both social identification and social identity salience. This four item scale is pre-validated as it had a reliability of Î±=0.89 when used in previous similar research (Haslam et al. 2004). In the current study however, one original item ('I am pleased to be a member of the unemployed community') was omitted as it was not appropriate for the population. An example of a question used is 'I see myself as a member of the unemployed community'. An acceptable level of internal consistency was found for the three items (Î±=0.80). Haslam, Oakes, Reynolds and Turners' (1999) single-item measure of social identity was included, previous research indicated that this item is highly correlated with other measures of social identification (e.g. Mael, 1998). This item 'Being a member of the unemployed community is important to me' did not correlate with the previous three items and brought the reliability down to (Î±=0.72). Accordingly, on analysis this item was omitted as a measure of social identification. Items are scored on a seven point likert scale; with options ranging from 'Not at all' to 'Very much'.
Participants then responded to five original items measuring participants perceived experience of stress regarding the exam (Î±=0.77). Items included 'How difficult would you find the exam?' and 'How pleasant would you find the exam?'. Three items on this scale were reverse coded. All were measured on a seven point likert scale, where 1= Not at all and 7= Very much.
Participants then responded to 19 items, requiring them to rate how they felt after completion of the exam. The first nine items were adapted from the Beck Anxiety Inventory (e.g. Beck, Epstein, Brown & Steer, 1988) with the final ten items taken from the Stait-Trait Anxiety scale (Speilberger, Gorsuch & Lushene, 1983). Both scales are multiple choice self report inventories that are commonly used for measuring the severity of an individual's anxiety (e.g. Haslam et. al 2004). The Beck Anxiety Inventory is measured on a four-point likert scale, ranging from 'Not at all' to 'Severely', with Î±=0.93. The scale includes common symptoms of anxiety, e.g. 'feeling hot' and 'heart pounding'.
A short (nine-item) form of the State Anxiety subscale The Stait- Trait Anxiety Inventory (Speilberger, Gorsuch & Lushene, 1983) was presented to measure levels of stress resulting from hypothetical completion of the exam. Such short forms of the STAI have been shown elsewhere to provide valid and reliable measures of state anxiety (e.g. O'Neil, Spielberger & Hansen, 1969). Stait anxiety is defined as an unpleasant emotional state characterized by tension, worry, fear and unease (Lev-Weisel & Kaufman, 2004). Participants were instructed to rate the intensity of their feelings on a four- point likert scale, ranging from 1 (not at all) to 4 (very much). Items in this scale include 'I feel calm' and 'I am worried'. Four items on this scale were reversed. The overall internal reliability of the State Anxiety scale was Î±=0.85.
Similar to previous research (Haslam et al. 2004), the five items of perceived stress and combined with the 10 items from the Beck Anxiety Inventory and the nine items from The State Trait Anxiety Inventory to form a single measure of stress. The resulting aggregated measure of these 24 items had a good reliability Î±=0.93.
A pilot study was carried out on a group of twenty participants in the Limerick social welfare office in the weeks prior to final collection. Participants were asked for their feedback on the questionnaire, questions like was it difficult to fill out, how it could be improved, was it easy to read were asked. As a result of this pilot study, the questionnaire was appropriately adapted. Data collection took place on social welfare collection days, Tuesday 15th and Wednesday 16th February 2011. Participants were approached while queuing outside the Limerick city Social Welfare Office before it opened. They were verbally informed that the study was designed to examine how people experience stress.
Questionnaires were administered by the researcher to participants. Each participant was given an information sheet and was asked to sign an informed consent sheet. The questionnaire comprised of six sections, firstly the demographic information was completed. Following this, participants indicated their level of identification with the unemployed community using three items from Doosje, Ellemers, & Spears (1995) social identification scale as well as Haslam, Oakes, Reynolds and Turners' (1999) single-item measure of social identity.
Next, participants read a hypothetical scenario from a leaked report informing them that a new, mandatory english and mathematics exam would be introduced for job seekers with non-compliance resulting in a stop to unemployment benefit payments. Depending on condition received, participants were introduced to Frances (in-group-, an unemployed person) or Frances (out-group, a civil servant). They were asked to rate how much they identified with her on a seven point likert scale. Participants then read Frances' account of the exam. The message conveyed Frances' experience as either stressful or challenging.
Following a manipulation check to ensure participants comprehended the advice, they were then asked to physically imagine themselves in the exam scenario. They read a detailed procedure of what it entailed, time limits involved and how results would be returned. This was designed to induce stress for the participant. Participants rated the immediate effect of the exam by responding to five statements regarding their subjective experience of stress on a seven point likert scale.
Finally, participants indicated their agreement with 19 items, nine adapted from the Beck Anxiety Inventory (e.g. Beck, Epstein, Brown & Steer, 1988) with the final ten items adapted from the Stait-Trait Anxiety scale (Speilberger, Gorsuch & Lushene, 1983). Both standardised stress measures used four point likert scales. All participants were given a thorough verbal debriefing along with a detailed written one, stating that the scenarios they had just read about were imaginary. Included in the debriefing were contact details of sources of support for people in unemployment for participants who felt that they had been affected by the sensitive nature of the research.
To ensure the message source and the message content information were read thoroughly and considered, participants were asked two questions relating to Frances' advice. The first question was a yes/no answer to the question 'Is Frances unemployed?'. The second asked participants to respond to the question 'Did Frances find this test a positive experience?'. They responded on a likert scale ranging from 1(not at all) to 7(very much). If the participant responded incorrectly to either question, their data was omitted from the data set. For the second question, if the participants responded incorrectly below or above the median value their data was omitted, resulting in 20 cases being deemed void.
Participants' experiences of subjective stress following the exam were analysed using a 2x2x2 univariate analysis of variance (ANOVA), in which the Message Source (ingroup/outgroup), the Message Content (stressful/challenging), and level of Identification with unemployment (above median=high/below median=low) were included as between- participants' factors.
The three-way interaction between Message Source, Message Content, and Identification was not significant, F(1, 142) = .197, p = .442, Î·p2 = .004. See table 1. There was a significant two-way interaction between Message Content and Message Source, F(1, 142) = 2.55, p<.01, Î·p2 = .05. See figures 1 and 5. Simple effects analysis shows that for stressful messages, receiving the message from the ingroup results in more reported stress than from the outgroup F(1,146) = 2.60, p <0 .01. But, for the challenging messages, there was no effect of source on reported stress F(1,146) = .67, p = .167. For messages from the ingroup, the content of the message has an effect, such that stressful messages are reported as more stressful than challenging messages F(1,146) = 4.29, p<0.01. But, for messages from the outgroup; message content does not affect reported stress F(1,146) = .17, p = .492.
There was no significant main effect of Message Source F(1, 142) = .369, p =.29, Î·p2 = .08. The main effect of Message Content was approaching significance, F(1, 142) = 1.23, p = .06, Î·p2 = .02. When participants were exposed to the stressful advice, in general their stress was greater (M = 2.38) than when they were told it was challenging (M = 2.18). See figure 2.
There was a significant main effect of Identification, F(1, 142) = 3.57, p<0.005, Î·p2 = .07. High identifiers with the unemployed community reported a mean level of stress at (M = 2.44) for low identifiers the mean level of stress reported was (M = 2.12). Specifically, high identifiers report being more stressed than low identifiers regardless of the message they were exposed to. See figures 3 and 4.
In sum, when advice of a stressful nature is received from an ingroup member (unemployed person) it is perceived as more stressful than from an outgroup member (civil servant). The content of ingroup messages have an effect, in that stressful messages from the ingroup are reported as more stressful by participants. However, for out group members, stressful message content does not affect reported stress. When the advice is of a challenging nature, reported stress is not affected regardless of the message source (unemployed person or civil servant). Participants who identified highly with the unemployed community experienced more stress overall then low identifiers.
In this exploratory study, social identity salience with the unemployed community was examined to determine if the appraisal of a potentially stressful situation can be determined by the group membership of a source that provides stress related feedback. In line with previous findings, it was hypothesised that social identification as unemployed will be associated with perceptions of increased social support, and reduced stress. Results verify that that this interactive hypothesis was not supported. In fact, an opposite effect was found to be true. When participants reported high levels of social identification with the unemployed community, their stress was actually greater than low identifiers. Social support did not mediate this effect as the source of the message did not affect reported level of stress.
As the second hypothesis predicted, the appraisal of stress was found to be affected by the group membership of the source providing stress related feedback. When participants received a message from an unemployed person, telling them the exam was stressful, participants in turn reported high levels of stress. If they were told that the exam was stressful by a person in fulltime employment, reported stress was lower. Interestingly, when the content of the message was relying information of a challenging nature, there was no effect on reported stress for both the ingroup and outgroup. These results do not comply with the findings of the original research by Haslam et al (2004), where both stressful and challenging messages were appraised in accordance with the ingroup provider. Instead, in this study it appears that the capacity for informational support to affect appraisal is contingent on the group membership of the support provider only when the content is stressful.
Self categorization theory only partially supports this finding, in that stressful information form a member of the unemployed community has more of an impact then that provided by a member of the employed community. The word partially is used with strength as the same impact was not found when the information was of a challenging nature. If the participant shares the same social perspective of the source, they are seen as more qualified to inform him or her about social reality (Turner, 1987). The source of the message alone did not affect the levels of stress experienced by participants. The content of the message participants were exposed to, was found not to be statistically significant, although it was approaching it. A larger sample size may have brought it up to significant levels.
The third and final hypothesis suggested that there would be differences in levels of identification and perceived stress between those who have been unemployed less than one year and those in long term unemployment. No relationship was found to exist regarding level of identification, length of unemployment and perceived stress. Unexpectedly though, level of identification people reported with the unemployed community impacted on how they experienced stress. Specifically, people who identified highly with the unemployed community reported higher levels of stress in than low identifiers. It is imperative to note that this difference remained, regardless of the content of the message or the message source.
Results of the current research challenge the traditional self-categorization model of stress. This model implies that the experience of stress may be reduced by the provision of support (Haslam, 2004). In the current research the provision of support by a member of the unemployed community increased reported stress when the perceiver received a message telling them the exam was stressful. In Haslam et al (2004) that consequence was qualified when the perceiver appraised an ingroup message relying that the exam was challenging and positive, resulting in lower levels of reported stress. However, in that study a population with a positive social identity (university students were investigated. In the current study, the population under investigation have a negative social identity did not react in the same way as the university students. Moreover, the social identity approach postulates that identification with a meaningful ingroup can buffer the effects of a stressful experience. That said, even identification with a negatively stigmatized group is shown to produce benefits through the enhancement of a sense of belonging (Schmitt & Branscombe, 2002). In addition, secondary appraisal of one's ability to cope with a stressor should vary as a function of social identity salience. This buffering effect did not occur when unemployed individuals were presented with a challenging message about the proposed exam. One may incur from this that unemployment has a negative effect on coping resources.
In contrast to expectation, high identifiers with the unemployed community experienced greater perceived stress than those who did not identify highly. Cassidy (2001) stipulates that because individuals perceived social support is weakened through unemployment, the acceptance of a self categorization as unemployed is more likely. This finding is further supported by McFayden (1997) who notes that the adoption if an unemployed identity will influence an individual's coping ability. Furthermore, individuals who adopt an unemployed identity experience greater psychological distress than those who adopt an occupational identity (Cassidy, 2001). Those who did not identify highly with unemployment choose not to accept an unemployed self categorization. Their lower levels of overall stress might be explained by Cassidys' (2001) findings.
The source of the message did not impact on reported stress of the participants, however together with the message content it did have an effect. The sense of social identity with the unemployed community participants felt did spark their internalised sense of membership in a particular group. In this case it was a negative one as they only responded to the stressful message presented by an unemployed person. This study extends on pervious findings by demonstrating that the appraisal of stress is not only affected by group membership of the provider but also by high levels of identification and the content of stress related feedback. The negative stereotype attributed to the unemployed community does impact on the social identity of the group which is shown by participants stress responses after exposure to a stressful message from an ingroup source (Kelvin & Jarret, 1985). Why does this negative identity persevere even when a member of the unemployed community tells the participant that the message content is challenging? Sitting an exam is a stressful event for most people, when faced with a high stress situation many people would comprehend it as stressful regardless of what people tell them. In the experiment the exam was described quite vividly to ensure people would be able to imagine themselves in the situation. Previous experiences of exams may have influenced people's responses to the stress measures, along with the fact that unemployment is typically a stressful time in a person's life leading to numerous ailments (see Eisenberg & Lazarsfeld, 1938; Hanisch, 1999).
This study was an elaboration of Haslam et al., (2004) research which demonstrated that the appraisal of stress is affected by the group membership of the source providing the stress related feedback. This effect was understood through the lens of the self categorization model of stress. Informational support from ingroup members had more impact compared to an outgroup source. Although the present studies results were in line with this, the effect was only significant if the advice being received was stressful in nature. The social identity approach as a whole struggles to explain this finding. Furthermore, identification with the unemployed community produced negative effects as it seemed to increase reported stress. This is in direct contrast to the self categorization model of stress which postulates that identification with a negatively stigmatized group produces an enhancement of a sense of belonging (Schmitt & Branscombe, 2002). Cassidy (2001) has found that unemployed people with poorer coping styles tent to self categorize as unemployed. By accepting an unemployed identity, people demonstrate this poor coping mechanism. One theory which has been overlooked and may prove beneficial in this area is Uncertainty Identity Theory. This theory is based on the relation between uncertainty about and reflecting on self and group identification (Hogg, 2009). Self categorization is seen as the most effective way to reduce self uncertainty. Literature has suggested that effects of unemployment can lower moral and lead to increased personal instability (Eisenberg & Lazarsfeld, 2007). Individuals experiencing unemployment may be uncertain about the future as they have lost their positive social identity gained through working. The depersonalisation of self is associated with a sense of identification and belonging to a group, thus self- categorization reduces uncertainty about who one is (Hogg, 2009). Individuals in unemployment are experiencing an uncertain time in their lives; by depersonalising themselves and self categorizing as unemployed they may feel a sense of belonging to that group. However, this self categorisation may also lead to conformity of group norms thus explaining why high identifiers experienced more stress in general than low identifiers.
This research has several notable limitations which should be taken into consideration for future research. The design of the questionnaire required participants to 'imagine' themselves in the exam scenario. This was possibly difficult to do while queuing outside the welfare office. As well as this, the content of the hypothetical exam included both english and mathematics were included, there is a possibility that some participants may have appraised one more stressful than the other. Furthermore, the content of the stressful message was extremely persuasive, which may imply that people believed it more or could relate to it more than the content of the challenging message. External validity must also be addressed as the non-experimental setting may have allowed for extraneous variables to affect responses. There were a large number of people in the queue, thus issues of social desirability may have come into play. The setting of the research may have influenced people to take on a stronger social identity; in that people may have identified more strongly as unemployed in the company of so many other unemployed people or they may have tried to reject this to preserve their own identity.
Gore (1978) found that the rural unemployed people showed a significantly higher level of social support than did the urban unemployed. As all data was collected in Limerick city, the majority of the sample was possibly urban unemployed. The measures used in this study were validated by similar previous research (Haslam et al., 2004). One of the measures of stress used, the Beck Anxiety Inventory typically measures the physiological symptoms of stress. As participants did not have to physically partake in an exam, responses to this scale may be questioned as it is difficult to imagine physical outcomes of stress without actually completing a stressful experience.
Although the present results strongly support the primacy of cognitive appraisal processes in threat responses, they do not necessarily generalise to all forms of affective emotional experience. The timing of the research may have resulted in increased stress levels as a member of the unemployed community. The Irish government introduced budget cuts to social welfare allowance in January 2011, which may have increased monetary stresses for participants as data collection took place in February 2011.
Despite these limitations, this study has provided some interesting data on the role of self categorization in coping with stress. This exploratory study raises questions about the suitability of a self categorization model of stress in explaining the role that social influence as unemployed plays in the appraisal of a potentially stressful situation. Future researchers should take into consideration the impact of various participant groups on self categorization, especially negatively stigmatised groups. It is an area that could be fruitfully pursued in future study regarding the unemployed community. Results verified that high identifiers with the unemployed community experience higher levels of stress when compared with low identifiers. On a practical note, it is imperative that mental health services are made readily available to unemployed members of the community.