A Randomised Enquiry on the Interaction Between Wellbeing and Citizenship

Wellbeing and citizenship have been outlined as important components for personal flourishing and involvement in community life. The Community Engagement Project aimed to identify the strength of the relationship between wellbeing and citizenship. In addition, the study aimed to investigate the capability of changing wellbeing and citizenship levels over time after exposure to a short experiment which also involved the following variables; Social Justice Beliefs, Assertiveness, Justice and Care Values. An intervention research design was implemented with a self-selected sample; 28 participants from the experimental group and 24 participants from the control group completed the final follow-up. Correlational results supported our first hypothesis as a statistically significant positive relationship was found between all the citizenship and wellbeing subscales. Support was found for within group differences for Personal Responsibility and Social Justice Beliefs indicating that this short experiment was able to influence these variables from pre to post. The repeated measures analysis of variance demonstrated that when considering time and condition the subjective perception of greater Legal Rights increased after the experiment only for the intervention group. Professionals could use the findings to provide a platform for vulnerable groups by providing the resources to be more assertive which could improve overall wellbeing and sense of citizenship.


Active Citizenship and Community Engagement
Citizenship has been defined as a measure of the strength of an individuals' connections to their rights, responsibilities, roles, resources and relationships that society offers them through public and social institutions (Rowe 1999;Rowe et al. 2001). Full citizenship refers to the privileges provided by members of a democratic society and to the responsibilities these rights engender. Participation in rights versus passive legal and social rights have been found to motivate individuals to improve the well-being of their communities whilst simultaneously enhancing their own development (Janoski 1998;Zaff et al. 2010).
Historically, the rights of individuals with physical and/or social disadvantages have been neglected. For instance, the rights of people diagnosed with mental illnesses were restricted on the grounds of mental incompetencies due to 'labelling', hospitalisation or other factors which led to them being characterised as 'second class' citizens. Among some members of the latter group, the combination of mental health and extreme social exclusion has led to a further deterioration of rights with many individuals falling into the category of 'noncitizenship'. Homeless people may fall into the 'non-citizenship' category due to their lack of contact with mainstream society and the stigma that is still associated with homelessness (Rowe 1999). Active citizenship, a term that involves a proactive attitude towards the use of rights and the engagement with responsibilities, has also been associated with relationship building and role achievement (Rowe et al. 2009). Reciprocal interpersonal relationships enable the individual to identify with a larger group, such as being part of a community, which in turn provides access to resources. The process of expanding citizenship rights, for example by volunteering, engaging in community groups or being involved in organisations that fight for injustices has many psychological benefits (Flanagan and Levine 2010). This has been linked to positive identity development, fulfillment, self-efficacy, increased social connection and enhanced communication skills (Dolan 2010;Kendrick 1996;Sherrod et al. 2010;Tucker and McCarthy 2001;Youniss et al. 1997). Engagement in community services has been found to initiate an involvement in social and political attitudes which in turn could lead to political development (Fyfe and Milligan 2003;Haste 2004;Sherrod et al. 2002). Referring back to the example of individuals diagnosed with mental disorders, by merely providing them with the resources to participate in elections and allowing them to vote has been found to be a powerful method of enabling them to express their citizenship rights (Chan and Chiu 2007). Additionally fighting for injustices, such as through the use of activism, has not only been linked to improved levels of well-being but has also led to a sense of empowerment (Gilster 2012;Eiroa-Orosa 2017, 2018). Therefore, practices that initiate greater participation in community activities have the ability to empower the individual and lead to better overall wellbeing (Weddington 1995

Social Justice
Social Justice has been defined as the fair and equitable allocation of resources, rights, and treatment of marginalised individuals who do not share equal power in society (Fondacaro and Weinberg 2002;Prilleltensky and Nelson 1997a, b).
Early paradigms of social justice such as the 'Libertarian Justice Model' by John Locke emphasised the connection between merit and liberty, this is based on the idea that the acquisition of resources do not need to be equitable but they should occur fairly and reflect what is deserved or entitled regardless of individual differences (Nozick 1974). Therefore, any multicultural movement that underemphasizes social justice is unlikely to eradicate oppression and will maintain the status quo (Vera and Speight 2003). Individuals who believe that there is social justice, for example they have strong 'Beliefs in a Just World' (BJW) are more likely to support the status quo, that is that if the world is just then the distributions of wealth in a society should and must be maintained (Dittmar and Dickinson 1993). Previous research has demonstrated a strong link between BJW and wellbeing, as the perceptions of fairness and (in)justice have been linked to health and mental health outcomes (i.e. depression) (Furnham 2003;Jiang et al. 2016). Individuals with high BJW are more likely to have greater life satisfaction and use it as a coping mechanism against the stresses of acknowledging injustice (Dalbert 2001;Lipkusa et al. 1996). On the other hand, research has found that BJW provides a rationale for inaction, even among disadvantaged groups, because of the belief that people will eventually receive the rewards they deserve (Stroebe 2013).
How could social justice be promoted? According to the contact hypothesis if individuals identify with particular groups in conflict they have the opportunity to reevalute the relationship they have with that group and eliminate negative stereotypes (Allport 1954). However, the contact hypothesis has been insufficient in explaining the way in which people achieve social justice and change. Direct services such as advocacy, outreach prevention programs, public policy and psychoeducational interventions have been found to be effective in promoting social justice (Bemak and Chung 2005;Ginwright and James 2002;Kiselica and Robinson 2001;Ratts et al. 2016).

Assertiveness
Assertiveness has been described as the ability to express oneself without anxiety or aggression in different situations (Bouvard et al. 1999). Behaviours considered low in assertiveness (i.e. avoidance and anxiety) can lead to worse material or instrumental outcomes, as well as a failure to receive resources and support (Ames et al. 2017). Additionally, this has also been linked to neuroticism, depression and psychosomatic complaints (Rushton et al. 1989;Thompson and Berenbaum 2011). Unassertive behaviour may arise from the mistaken belief that one does not have the right to act assertively, or the belief that certain rights are denied due to one's gender or position in society (Lange and Jakubowski 1978).
In contrast, high levels of assertiveness (i.e. competition and aggression) have been linked to compliance and the attainment of one's rights. However, these aggressive responses can result in greater anger and hostility which may only be effective in the short term. This is because they can hinder relationships, lead to stress and diminish health and well-being (Ames and Flynn 2007;Aubé 2008). Therefore, assertiveness is required to achieve equity in human relationships and to exercise personal rights without denying the rights of others (Alberti and Emmons 1995). While it is true that citizens in the world's richest countries are more willing to make demands and engage in the delivery of public services, this is still dominated by the better educated and wealthier middle class. Increased assertiveness and user engagement are key in securing better outcomes, however it is important to consider the barriers that may prevent users from gaining access to the highest quality of service delivery (Griffiths et al. 2009).

3
How can we increase assertiveness for disadvantaged groups? Training individuals to develop their assertiveness skills has led to a number of health benefits such as an increase in self-esteem, life satisfaction and a reduction in interpersonal conflict. Additionally, this has been linked to the obtainment of one's own rights and the advocacy of the rights of others (e.g. Ahmadi et al. 2017;Chan 1993;Rathus 1972).

Justice and Care Values
Justice and Care Values (hereafter referred to as 'values') relates to an individual's moral development which is the replacement of the internalisation of societal standards to cognition and action which have been developed to benefit others (Nelson et al. 2001). As a result, individuals do not only do the right thing for the sake of social conformity or expectations from authority figures, but instead, the individual has developed concerns and an awareness of moral conduct which is implemented in a wide range of situations and in the society they live in (Habib 2011).
Discussions about moral development involve the contributions that have been made by Piaget who focused on the way in which morality developed in children as a means of understanding adult morality. Piaget viewed the development of a child's moral judgement as a shift from the heteronomous stage to the autonomous stage which consists of the emergence of moral concepts such as justice and rights (Ruffy 1981).
Previous work has outlined that values are required in order to build solidarity with disadvantaged groups, as well as promote the well-being of these groups which benefits individuals and communities (Kekes 1993). Justice is a value that guides the fair and equitable allocation of resources and burdens in society. For example, in order to achieve selfdetermination and have a high level of well-being the presence of financial and material resources need to be available, as well as tangible resources such as education, employment, income and housing (Nelson et al. 1998;Prilleltensky and Nelson 1997a, b). Whereas care and compassion provide the motivation to look after someone else's wellbeing in the form of concern for the physical and psychological health of others such as caring for the elderly (Nelson et al. 2001;Swanson 2010).
How can justice and care be encouraged in society? Community psychologists have helped to promote justice and care by forming partnerships with oppressed groups such as families living in poverty. This is due to mainstream organisations holding contrasting values and often blaming the victims for not 'fitting in' (Reinharz 1984;Ryan 1976). These settings are firmly based on the values of care and mutual support which focus on the enhancement of well-being and power sharing. The process of creating alternative settings such as crisis for women and self-help organisations have enabled a greater power distribution and has helped to create a value-based partnership. This could be encouraged further by rotating leadership responsibilities such as having individuals from disadvantaged groups chair meetings and give presentations about the projects that they have been involved in, which has the potential to enhance the participation of citizens and redistribute power (Nelson et al. 2001). On the other hand, justice could be promoted by providing tangible resources which could help to break down barriers for disadvantaged groups and allow them to participate in society. However, in order to achieve justice it is necessary to provide the resources to be an active citizen as opposed to merely redistributing resources. Nelson et al. (1998) discussed the importance of hiring and training people from oppressed groups in order for them to be able to gain access to education, employment, and income.
Active participation of citizens and providing oppressed individuals with a "voice" in society is an essential part of enabling citizen participation for different groups. For example, enabling disadvantaged groups to be involved in decision-making in prevention programs provides them with a meaningful way to be involved as opposed to token participation (Cameron et al. 2009;Church 1995;Valentine and Capponi 1989). This helps to build relationships and trust, establish norms and principles for working together, as well as sharing power and resources which are key processes for value-based partnerships (Nelson et al. 2001). This is important as partnerships with oppressed groups that do not have clear values run the risk of maintaining the status quo (Prilleltensky 1994).

Study Aims
We aim to investigate the relationship between wellbeing and sense of citizenship. We also aim to analyse if this relation is influenced by variables such as 'Social Justice Beliefs', 'Assertiveness' and 'Values'. Furthermore, we wanted to measure all of these variables and test the temporal stability and susceptibility to change after taking part in an experiment which tried to increase social awareness.

Research Questions and Hypotheses
According to the aforementioned aims of the study, we had the following research questions and hypotheses.
Our first research question focused on the relationship between wellbeing and citizenship. We aimed to measure the correlations and identify possible modificators of these variables. Our hypothesis was that wellbeing and sense of citizenship are correlated (H1a) and their relation is possibly altered by 'Social Justice Beliefs', 'Assertiveness' and 'Values' (H1b). An intervention conducted on individuals receiving public mental health services found a relationship between citizenship, wellbeing and assertiveness. This citizenship intervention involved a group component which focused on assertiveness training, problem solving, relationship building and valued roles. The findings demonstrated that there was an enhancement in the amount of satisfaction related to social activity, finances and work. Additionally, an improvement was found in the individual's quality of life (Clayton et al. 2013;Rowe and Pelletier 2012). Support has also been found for a link between beliefs in a just world and higher levels of wellbeing. BJW has been found to contribute to subjective well-being as it increases one's sense of control and enables individuals to cope more adaptively with negative life events, BJW has also been linked to psychological wellbeing and greater life satisfaction (Lipkusa et al. 1996). Research has found a relationship between wellbeing and values as subjective wellbeing may be associated with emphasizing particular values such as compassion, as opposed to others such as security (Sagiv and Schwartz 2000). Additionally values that emphasize personal autonomy, freedom of expression and equality of opportunities should potentially bring more satisfaction (Welzel and Inglehart 2010). These variables will be measured within an experiment which is outlined below.
Our second research question is related to the stability of wellbeing and citizenship over time. We wanted to find out if doing a short experiment designed by our study group could change the subjective view of wellbeing, citizenship and their relation. Our second hypothesis is that the levels of citizenship and wellbeing could be altered (H2) by implementing a short experiment focusing on the awareness of the aforementioned variables, i.e. 'Social Justice Beliefs', 'Assertiveness' and 'Values' (whose evolution will also be tested). This will be tested within and between groups. Within groups interactions (expecting change after the experiment, H2a) will be tested taking both groups together and testing pre-post differences. Between groups interactions (expecting differences among groups, with higher levels for the experimental group, H2b) will be tested after using a wait-list controlled randomised design, in which 50% of people participating will be randomised to an experimental group and the rest to a wait-list group. The latter group will then be offered the opportunity to go through the same experiment after one week. Both groups will be followed-up one week after the experiment. Interactions between time and experimental group (expecting time by group interactions, H2c) will also be tested.

Participants
The study aimed to recruit participants using a self-selected sampling method. For this reason participants were recruited using social network sites and forums over a period of four months. The participants were informed that the study aimed to deepen their knowledge of the influence of citizenship on wellbeing. In order to take part in the study, the participants had to be 16 or over and have a high comprehension of English.
Sample size was calculated using GRANMO (https ://www.imim.cat/ofert adese rveis / softw are-publi c/granm o). The calculation was done taking into account hypothesis H2a. Accepting an alpha risk of 0.05 and a beta risk of 0.2 in a two-sided test, 63 subjects would be necessary in each group to recognize a statistically significant difference greater than or equal to 0.5 standard deviations.
Although a total of 242 people were enrolled in the study, 60 participants did not finish the baseline and therefore could not be included in any analysis (see Fig. 1). Of the remaining 182 participants who had completed stage one, 175 continued to randomisation. From the latter, 84 were allocated to the experimental condition and 91 were allocated to the control condition. Due to attrition, the number completing the follow-up from the experimental group reduced to 28 and for the control group this reduced to 24.

Procedure
The participants used the online Qualtrics platform (https ://www.qualt rics.com/) to complete the 'Community Engagement Project', which involved providing them with information and a consent form to participate in the study.
All consenting participants were directed to the baseline survey which involved answering the wellbeing and citizenship subscales and sociodemographic questions (i.e. age, gender, place of origin, etc.). At the end of the baseline, with the option of refusing to do so, participants provided their email address. This was essential in order to be able to access the next sections of the study, as an embedded link was included in the email reminders which were sent automatically by the Qualtrics platform. The participants were then allocated at random to either the wait list-control or the experimental condition. At each stage after completing the baseline measures, both participant groups were provided with their overall wellbeing and citizenship score in order to inform the participant of their progress. A CONSORT flow chart including the number of participants allocated to each condition, followed up and analysed can be seen in Fig. 1. Figure 2 shows a schema of the experimental allocation and flowchart. After taking the baseline survey the participants in the experimental condition were directed immediately to the short experiment which involved watching a series of videos and answering questions related to 'Social Justice Beliefs', 'Assertiveness' and 'Values'. At time 2, which was approximately one week after the baseline, the experimental group took part in the follow-up which involved reassessing their wellbeing and citizenship levels using the same subscales presented during the baseline assessment. Additionally, at time 2 the participants reanswered questions related to 'Social Justice Beliefs', 'Assertiveness' and 'Values' which had originally been presented during the experiment. This was in order to identify whether their wellbeing and citizenship levels had changed after taking part in the short experiment.
In contrast, the control group were only given the baseline subscales and were then informed that they had to wait for approximately a week to receive the second part of the study. Email reminders were sent out in order to prompt participants to complete the short experiment. At time 2, which was approximately one week later, they were re-evaluated on the citizenship and wellbeing subscales before being diverted to the short experiment involving the same videos that had been presented to the experimental group (i.e. related to 'Social Justice Beliefs', 'Assertiveness' and 'Values'). Lastly at time 3, which was approximately one week after the experiment, the control group received the follow-up which involved reassessing their opinions related to wellbeing, citizenship, 'Social Justice Beliefs', 'Assertiveness' and 'Values'. After the participants in each condition completed the Community Engagement Project they were thanked for their time and participation.
The online questionnaire was structured in a way that facilitated participation in our cross-sectional inquiry on the relationship between wellbeing and citizenship and their possible modificators as these themes were embedded in the experiment with the aim of raising awareness on all these issues. All the measures and activities used can be found in the measures and experimental procedures sections, mirroring the structure in which they were presented to participants.
The study was approved by the Bioethics Committee at the University of Barcelona (Institutional Review Board: IRB 00003099).

Active Citizenship Experiment
This involved a video outlining the importance of citizen participation and overall wellbeing. The video promoted the five R's; Rights, Responsibilities, Roles, Resources and Relationships (Rowe and Pelletier 2012). Promoting the five R's is important as it gives meaning to our lives and helps to build identities.

Social Justice Experiment
Immediately after completing the Social Justice Beliefs questionnaire, a video was presented which described the lives of two individuals; Richard and Paula. This provided the story of how it is not so easy to be successful if you do not have the right opportunities and you do not come from a privileged background. After watching the video, the participants were presented with open questions which were 'In your opinion would you say that everyone has the same opportunities regardless of their social class? If you work hard can you still achieve the same education and job opportunities?'.

Assertiveness Experiment
After completing the questionnaire, the participant was presented with a video related to assertiveness (https ://www.youtu be.com/watch ?v=5tqu6 r4Bmj 8) which describes the benefits of being assertive as it has been associated with more positive relationships, group harmony and greater wellbeing.

Justice and Care Values Experiment
The last section of the experiment was divided into two different types of values; justice and care. The participant was shown a moral judgement vignette (Habib 2011) which was based on a moral situation. This described the story of a young boy from a disadvantaged background who was contemplating shoplifting in order to help his sick brother.

I Coppe Scale (Prilleltensky et al. 2015)
The I Coppe Scale is a self-anchoring method using a likert scale that measures subjective life experiences with a range of 0 "worst" to 10 "best". This refers to the respondent's levels of satisfaction based on present, past and future experiences (Cantril 1965). Strong psychometric evidence of validity and reliability have been documented for this self-anchoring method by wellbeing researchers (Prilleltensky et al. 2015). Overall the cronbach alpha in our study was high for this questionnaire (α = .947). The I Coppe scale consists of 21 items which assess the underlying constructs of wellbeing; Overall Life, Interpersonal, Community, Occupational, Physical health, Psychological and Economic factors. The term overall wellbeing refers to the subjective perception of the individual's state of affairs, for example "When it comes to the best possible life for you, on which number do you stand now?". Interpersonal wellbeing refers to the quality of relationships that an individual has with friends, family and colleagues, such as "When it comes to relationships with important people in your life, on which number did you stand a year ago?". Community wellbeing refers to satisfaction with one's community "When it comes to the community where you live, on which number do you think you will stand a year from now". Occupational wellbeing refers to satisfaction with one's job or vocation. Physical wellbeing refers to satisfaction with overall health and wellness. Psychological wellbeing refers to satisfaction with one's emotional life and Economic wellbeing relates to satisfaction with one's financial situation.

Citizenship Measure (Rowe et al. 2012)
The Citizenship measure, originally proposed by , consists of 45 questions based on the five R's; Rights, Responsibilities, Roles, Resources and Relationships. These items were further divided into seven domains; Personal Responsibility, Government and Infrastructure, Caring for self and others, Civil Rights, Legal Rights, Choice and World Stewardship. An example item includes, 'Thinking about your life in general right now, please read each statement and rate on a scale of 1 to 5 how much you feel that the statement applies to you… Your personal decisions and choices are respected' which in this case relates to the choice domain. The instrument was able to capture subjective information regarding the individual's experiences using a likert scale ranging from 1 "Not at all/Never" to 5 "A lot/Very Often". Analyses of the citizenship measure have illustrated that the measure is psychometrically sound (O'Connell et al. 2016). Overall in this study the Cronbach's alpha for this scale indicated that it had good internal consistency (α = .946).

Demographics
A questionnaire consisting of ten items related to gender, age, background, country of origin, country of residence, relationship status, educational qualifications, occupation and social class. Additionally two ideology questions on materialism/postmaterialism and traditionalism-secularism were included; 'Government should ensure that everyone is provided for' and 'A child needs a home with both a father and a mother in order to grow up happily' (Inglehart et al. 2000). These questions were operationalised using a likert scale ranging from 1 (strongly agree) to 6 (strongly disagree).

Social Justice Beliefs Measure (Dalbert 2000)
Social Justice was measured using the 'Belief in a Just World' questionnaire. This is a sixitem questionnaire related to the concept of justice 'I believe that, by and large, people get what they deserve' and injustices 'I am convinced that in the long run people will be compensated for injustices'. Participants had to rate themselves on a Likert scale ranging from 1 (strongly disagree) to 6 (strongly agree). The reliability and construct validity of the questionnaire has been approved in different studies (Ambrosio and Sheehan 1990;Caputi 1994;Couch 1998;Mohr and Luscri 1995;O'Quin and Vogler 1990;Whatley 1993). The reliability of the questionnaire has been reported to be 0.672 and 0.678 for belief in a just and unjust world (Roshani et al. 2013). In the current study, a good Cronbach alpha was found for this questionnaire (α = .846).

Assertiveness Scale (Thompson and Berenbaum 2011)
Assertiveness was tested using the 'Adaptive and Aggressive Assertiveness Scale' (AAA-S) which involved 19 hypothetical situations which reflected adaptive and aggressive assertiveness. Participants had to indicate the extent to which they would react to a given situation using a scale of 1 (never) to 5 (always). An example of a hypothetical question related to aggressive assertiveness was "When someone close to me unjustly criticizes my behaviour, I…b. React angrily and tell the person that she/he shouldn't be throwing stones", whereas an example of an assertiveness question "It has been over a year since I received a promotion, I…a. Ask my boss about getting a promotion". This assertiveness scale was chosen due to displaying good internal consistency, split-half reliability, when employing the Spearman-Brown correction, and a 2-week test-retest reliability (Thompson and Berenbaum 2011). As reiterated previously, this questionnaire demonstrated a good Cronbach alpha (α = .866).

Justice and Care Values Scale (Gump et al. 2000)
The participant was first presented with a few open questions such as 'Do you think doing the right thing is always easy? Are there times when it is acceptable not to do the right thing?'. After answering these questions and seeing the vignette the participant completed the 'Moral Justification Scale' (Gump et al. 2000) which consisted of a Likert scale ranging from 1 (not at all important) to 10 (very important). These statements related to 'Justice' values "The amount his parents had been overcharged was nearly the same as the price of the coat. Tony felt it was only fair that he take the coat since the store really owed it to his family anyway" and 'Care' values "Tony could not bear seeing his fragile-looking younger brother suffer, for the little boy was already sick with the flu and could easily catch pneumonia". The 'Justice and Care Values' scale has displayed good internal consistency as when a test-retest method was used to measure the reliability of the scale, this was found to be 0.99. Content validity of the scale indicates the appropriateness of each dilemma (Habib 2011). There is a good Cronbach alpha for this questionnaire (α = .723).

Data Analysis
T tests and Chi squared tests with odds ratios were used to compare socio-demographic characteristics between the experimental and control group and to compare baseline scores of participants completing the first follow-up with those who did not. The normality of variance was checked using skewness and kurtosis parameters. Pearson correlation matrices were conducted in order to test the relationship between wellbeing, citizenship and their possible modificators; 'Social Justice Beliefs', 'Assertiveness' and 'Values'. A paired samples t test was conducted to test within group differences pre and post experiment. Independent samples t tests were used to test group differences between the conditions; control and experimental group. A repeated measures factorial analysis of variance (RM-ANOVA) was used in order to test within subject changes from the baseline to the follow-up, taking into account time and condition. Additionally, to take into account possible bias as a result of attrition, multilevel linear models were performed using measures of wellbeing, citizenship and their possible modificators as within-subject variables. Statistical analyses were conducted using SPSS version 22.

Sociodemographic Characteristics and Missing Analysis
All kurtosis and skewness parameters were within the -1.96 and 1.96 range recommended by Doane and Seward (2011), thus we decided to use parametric statistics. We only found that 'Occupational wellbeing' and 'Care' had a skewness and kurtosis outside the range of − 1 and +1. 'World Stewardship' had a skewness value below 1, whereas 'Economic wellbeing' and 'Aggressive assertiveness' had a kurtosis value greater than 1. However, all the variables were within the − 1.96 and 1.96 range. Sociodemographic characteristics and the ideology variables have been grouped and are shown in Table 1. There were no statistically significant differences between groups. The sample of randomised participants 74% females (n = 128), represented 13 countries (including Australia, Bahrain, Belgium, Canada, China, France, Germany, Spain, Switzerland. The Netherlands, United Arab Emirates, The United Kingdom and United States of America). The participants were predominantly Caucasian 74.7% (n = 130) with a small minority of participants representing other ethnicities such as Asian 10.9% (n = 19), Hispanic 2.9% (n = 5), and African 4% (n = 7). The majority of the respondents indicated that they were in employment 59.8% (n = 104) had an undergraduate degree 50.6% (n = 88) and were currently in a relationship 67.8% (n = 118).

3
When we carried out the mean comparisons of baseline scores of participants completing the first follow-up with those who did not, we found statistically significant differences for the 'Choice' subscale of the citizenship questionnaire (t = 1.976, p = .05). Table 2 shows that there were statistically significant positive correlations between all the citizenship and wellbeing subscales. In general the results suggest a positive relationship, with greater citizenship levels consequently leading to higher levels of wellbeing. The strongest correlation was found between 'Overall Life Wellbeing' and the citizenship subscales related to 'Personal Responsibility', 'Government and Infrastructure', 'Caring for others', 'Legal rights' and 'Choice'. We found a strong positive correlation between 'Civil Rights' and 'Economic Wellbeing' (r (182) = .54, p = < .001). The citizenship subscale 'World Stewardship' indicates a statistically significant positive relationship with 'Psychological Wellbeing' (r (182) = .46, p = < .001). Table 3 illustrates the relationship between the wellbeing and citizenship baseline totals and the modificator variables: 'Social Justice Beliefs', 'Assertiveness' and 'Values'. The correlation matrix confirms previous results as a positive relationship was found between the wellbeing and citizenship totals. There appears to be a statistically significant relationship between the wellbeing and citizenship baseline totals (r (182) = .62, p < .001). However, there are no statistically significant correlations taking into account wellbeing and citizenship with the aforementioned variables. Within these variables, the matrix shows a low positive correlation between the 'Justice Value' subscale and the 'Aggressive Assertiveness' subscale (r (79) = .32, p < .05).

Table 2
Correlation matrix for all subscales of wellbeing and citizenship  Table 4 demonstrates the relationship between the modificator variables and the wellbeing and citizenship subscales. Statistically significant correlations, always with low effect sizes, were found for the following variable diads:  Table 5 shows within (paired samples t test) and between (GLM, time by condition) group differences for wellbeing, citizenship and the modificator variables.

Within groups prediction of group differences
The results from the paired samples t test indicate that no statistically significant differences were found for any of the wellbeing subscales. Regarding the citizenship variable, statistically significant differences were found for the subscale 'Personal Responsibility'

Between Groups Prediction of Group Differences
An independent samples t test was conducted in order to compare wellbeing and citizenship levels between the control (within their intervention) and experimental (within their follow up) groups. The results demonstrated that there were no statistically significant differences for wellbeing t (51) Table 4 Correlation matrix of the subscales and the modificators Note: (1) Economic wellbeing; (2)Occupational wellbeing; (3) Psychological wellbeing; (4) Personal responsibility; (5) Government and infrastructure; (6) Caring for others; (7) Civil Rights; (8)Beliefs in a just world baseline total; (9) Aggressive assertiveness baseline total; (10) Adaptive assertiveness baseline total; (11) Justice baseline total; (12) Care baseline total  * For these calculations we included the control group before they completed the intervention (second measurement) and compared this with the follow-up of the experimental group ** These calculations were performed comparing baseline with follow up (after having participated in the intervention) for both groups 1 3

Time × Condition Interaction in Predicting Within Group Differences
The two-way factorial repeated measures analysis of variance (RM-ANOVA) shown in Table 5 found a statistically significant effect of time and condition on the citizenship subscale 'Legal Rights' (F (1,51) = 5.72, p = .03, ηp 2 = .10). The multilevel linear models showed no statistical significance for time by group interaction for any of the wellbeing, citizenship or modificator variables. However, when adding 'Care' to a model and using citizenship as the dependent variable, we found statistical significance for time by 'Care' interaction (F (1,121.34) = 6.7, p = .01). This effect was found for all the citizenship subscales except for 'Caring for Others', which remained in a statistical tendency (F (1,99.384) = 2.8, p = .096).

Discussion
The present study investigated the relationship between wellbeing, citizenship and the variables that may modify this relationship 'Social Justice Beliefs', 'Assertiveness' and 'Values'. For instance individuals who have high levels of beliefs in a just world and a low adaptive assertive style are less likely to fight for their citizenship rights and the rights of others, and will be less concerned with promoting well-being (Ames et al. 2017;Ames and Flynn 2007;Dalbert 2001;Lange and Jakubowski 1978;Lipkusa et al. 1996;Stroebe 2013;Thompson and Berenbaum 2011).
The study also tested the stability of the wellbeing and citizenship measures over time after taking part in a short experiment. This was related to the modificators with the view that after completing the questionnaires and visualising the videos related to active citizenship and community engagement, 'Social Justice Beliefs', 'Assertiveness' and 'Values', participant's overall wellbeing and sense of citizenship would change.
The findings supported the first hypothesis as a strong positive relationship was found between wellbeing and citizenship (H1a). These results reiterate previous research which found that subjective wellbeing was positively linked to citizenship dimensions (Zalewska and Zawadzka 2016). As previously found 'giving back' to the community and being a valued citizen, such as being involved in volunteering, led to increases in overall quality of life and has been linked to a number of health benefits (Clayton et al. 2013;Hunter and Linn 1981).
Despite the strong relationship between wellbeing and citizenship, the second part of the first hypothesis was only partially met (H1b). This is because a statistically significant relationship was not found between all the wellbeing and citizenship subscales when taking into account the variables; 'Social Justice Beliefs', 'Assertiveness' and 'Values'. The results indicated a negative relationship between 'Economic Wellbeing' and 'Beliefs in a Just World'. Research has suggested that after exposure to inequality this threatens 'Economic Wellbeing' with an increased acknowledgement of income inequality in society (Furnham and Gunter 1984;Jost et al. 2003). In the case of our study, the video which illustrated priviledged and underpriviledged individuals may have led to a greater awareness of inequality, which is still apparent between lower and higher social class groups. A negative relationship has also been attributed between 'Economic Wellbeing', 'Social Justice' and political beliefs. This is because previous research has shown that left-wing individuals believe that there is a negative relationship between economic performance and escaping from poverty. This could in part explain the results of the current research as a high percentage of participants (89.7%) could be considered as having 'left-wing' views.
Additionally, in line with our findings, a negative correlation was found between the citizenship subscale 'Government and Infrastructure' and 'Beliefs in a Just World'. This may be linked to previous research indicating that left-wing individuals tend to believe that the government is doing too little to help the poor (Di Tella and MacCulloch 2014). This view is further divided among countries, with 69.7% of Europeans, which made up 83.2% of our sample, reporting that they are receiving too little help from the government. Alesina and Angeletos (2002) reported that 60% of Americans-yet only 26% of Europeans-believe the poor are lazy. It has been argued that the countries where few people hold this belief also have more government intervention, this indicates that aid from the government does not necessarily influence the perception of a just world.
A positive relationship was found between 'Adaptive Assertiveness' and 'Occupational Wellbeing' which reiterates previous findings which have implemented assertiveness programmes. Occupational therapists have administered 'Assertive Community Treatment' (ACT) in order to assist vulnerable individuals so that they have the capacity to experience rich occupational lives which is essential to overall health and well-being (Krupa et al. 2002). 'Adaptive Assertiveness' was also correlated to 'Psychological Wellbeing' which has previously been linked to lower levels of anxiety and depression, as well as improvements in life satisfaction and self-esteem (Ahmadi et al. 2017;Chan 1993;Rathus 1972). 'Adaptive Assertiveness' has previously been found to be strongly correlated to the citizenship subscale 'Personal Responsibility', which was also the case in the current study. Previous research has found that assertive citizens are more engaged in wider society due to discussing and negotiating the kinds of benefits they wish to receive from public services. For example, parents who take an active interest in their child's education are more likely to see them perform better at school. Similarly patients who are informed and willing to take a role in securing their own health tend to recover more quickly and stay healthier for longer. Under this view, service users and providers share the responsibility for outcomes which helps to shift power and responsibility towards the citizen (Griffiths et al. 2009).
'Justice Values' were positively correlated to the citizenship subscale 'Caring for others'. Previous work has found a close link between 'Justice Values' and 'Care' with the belief that they work together in achieving social change. For instance justice has been linked to ensuring that individuals are provided for by distributing the necessary resources in society, but more importantly that they are allocated equally (Nelson et al. 1998a, b;Prilleltensky and Nelson 1997a, b). The concept of equal allocation is governed by care and compassion which is administered in the form of physical and psychological concern for others in the community (Nelson et al. 2001;Swanson 2010).
The within group hypothesis was partly supported as although no statistically significant alterations were found in overall wellbeing, some statistically significant changes were found in the citizenship subscales (H2). Changes were seen, after completing the short experiment, for the citizenship subscale 'Personal Responsibility' and 'Beliefs in a Just World' (H2a). The variable 'Personal Responsibility' has previously been seen as a moderator for the effects of just-world beliefs. More specifically when evidence provides the opportunity to blame the victim, individuals who score high in just world beliefs are more likely to seize this opportunity. Whereas, in other cases where the individual could not do anything to avoid their fate, high just world perceivers actually demonstrate greater sympathy towards the victim (Braman and Lambert 2001). In this experiment we saw higher rates of 'Personal Responsibility' at the follow-up but lower rates of 'Beliefs in a Just World'. This may suggest that individuals felt more responsible for their own actions and for the wellbeing of others, but due to the lower levels of beliefs in a just world they did not appear to attribute the blame to the individual presented in the video.
The between groups hypothesis was not supported as there were no statistically significant differences between groups (H2b). The experimental group did not show statistically significant differences at follow-up compared to the control group. This may suggest that condition was not a statistically significant contributor in achieving changes in overall wellbeing and sense of citizenship.
The last hypothesis was partly supported as there was a statistically significant effect for the citizenship subscale 'Legal Rights' (H2c). Research has proposed that a person who is aware that he possesses rights will utilise the legal system as a resource and a means of validating his rights (Lister 2004), whereas a lack of support for legal rights perpetuates and maintains inequality (Watts 2017).
Additionally a statistically significant interaction was found between 'Care' and time for the citizenship subscales. This is not surprising due to the close link that has been found between 'Care Values' and 'Legal Rights'. Past research has found a relationship between the existing laws and policies and the diminishing value of care (McClain 2000). This is because much of the common law is based on the assumption that we are independent from others and therefore are entitled to have our rights of self determination and autonomy fiercely protected (Lloyd 2004).
Despite seeing increases in the citizenship subscales after taking part in the short experiment, there were no statistically significant differences for the wellbeing measures. These results could be due to the nature of this study which in turn influenced outcomes. For example, this was a web-based experiment which measured participants longitudinally after only a one or two week gap. Studies have indicated improvements in 'Interpersonal', 'Community', 'Psychological' and 'Economic' wellbeing after 30 and 60 days post-baseline. For wellbeing subscales related to 'Physical', 'Emotional' and 'Overall Life' even greater time is needed in order to see improvements. Therefore, it may be the case that more time was needed between the baseline and follow-up in order to see statistically significant changes in the different wellbeing subscales (Myers et al. 2017).
Limitations of the current study should be noted, firstly due to the voluntary nature of the experiment there was a high attrition rate. The study's advertisement on social media initially attracted 242 participants who followed the link and enrolled in the study. However, this reduced to 182 as not all of the participants finished the wellbeing and citizenship measures, indicating an initial attrition rate of 24.79%. This reduction in numbers may have been due to it being an online voluntary study, the length of each section and even participant variables such as not having the time to complete the follow-ups. Additionally, it is also important to take into account that due to the nature and structure of the study (i.e. it was a questionnaire-based study and that a wellbeing and citizenship score was provided at each stage) this may have led to more favourable responses by the participants. Therefore, it may not be the case that the responses were completely representative of the respondent's subjective wellbeing and citizenship levels.
A further limitation was that this was an exploratory study and therefore its results should be replicated in larger trials in order to establish more clearly the link between the variables tested in the current experiment. Lastly, the participants were predominantly representative of white well-educated females from the United Kingdom. However, the study achieved a range of participants worldwide from 13 different countries ranging from 16 to 76 years of age.

Practical Implications
The findings from the brief experiment have helped to contextualise citizenship in the context of wellbeing. The concern for individuals in their social context and for promoting wellness in communities creates a space to understand and support other members of the community in becoming a full citizen (Ponce and Rowe 2018). This is based on the premise that communities and society as a whole have an obligation to support the citizenship of marginalised groups by supporting their access to the five R's; Rights, Responsibilities, Roles, Resources and Relationships (Ponce and Rowe 2018). The current experiment can help to change the way in which individuals in the general public consider the concept of citizenship, moving away from it being the 'person's responsibility' to one that places the responsibility on society as a whole (Rowe and Davidson 2016). We aimed to change views by educating the public about the needs of marginalised groups and the contributions they can make in their home communities.
In conclusion, the brief experiment has provided an insight into the complexity and multitude of factors that contribute to both wellbeing and citizenship. As we have seen from the results, the variables have been influential in changing some of the citizenship subscales and may have the potential to contribute to wellbeing. This is encouraging as it has enabled us to understand that these variables, when studied over a longer period of time and applied to real life settings, could give us a better indication of the interplay between wellbeing and citizenship.
Funding This project has received funding from the European Union's Framework Programme for Research and Innovation Horizon 2020 (2014-2020) under the Marie Sklodowska-Curie Grant Agreement No 654808.