Access to this full-text is provided by Springer Nature.
Content available from Scientific Reports
This content is subject to copyright. Terms and conditions apply.
1
Vol.:(0123456789)
Scientic Reports | (2024) 14:19677 | https://doi.org/10.1038/s41598-024-70067-5
www.nature.com/scientificreports
Statistical analysis of the eect
of socio‑political factors
on individual life satisfaction
Alaa Itani
1, Isra Hasan
2, Lolya Younes
2 & Ayman Alzaatreh
3*
Life satisfaction refers to an individual’s cognitive evaluation of the quality of their life. The aim of
the present study is to develop the current understanding of how perceived corruption, attitudes
toward migration, perceived security, and strength of national identity inuence life satisfaction.
Additionally, the study examines how demographic variables of relationship status, social class,
sex, religious aliation, and country impact life satisfaction in the provided cultural context. Ordinal
logistic regression analysis, Conrmatory Factor Analysis, and Structural Equation Modeling are used
to analyze data from the World Values Survey. Findings from the analyses indicate that perceived
corruption, perceived security, and strength of national identity have a signicant impact on life
satisfaction, whereas migration has an indirect eect on life satisfaction through perceived security.
The present research can develop our current understanding of life satisfaction from a socio‑political
perspective.
Keywords Life satisfaction, Socio-political, World values survey, Conrmatory factor analysis, Structural
equation modeling
An individual’s life satisfaction refers to the cognitive and aective evaluation of the quality of their life1.
Researchers oen use life satisfaction as an indicator of subjective well-being, considering life satisfaction as
a cognitive evaluation and happiness as an aective evaluation of positive psychological health2. According
to Diener etal.2, individuals across 41 nations rated both life satisfaction and happiness as highly important,
reecting the value individuals place on life satisfaction. Based on self-report measures and positive outcomes
across psychological, physiological, and organizational domains, research has also demonstrated the importance
of life satisfaction.
Psychologically, lower levels of life satisfaction are associated with outcomes such as distress, depression,
and anxiety. Meanwhile, higher levels of life satisfaction are associated with greater resilience in the face of
adversities3. Physiologically, elevated life satisfaction is linked to reduced sleep complaints and lower rates of
cardiovascular and overall mortality4,5. Organizationally, life satisfaction contributes to lower levels of burnout,
enhanced job performance, and decreased intentions to quit6–8.
Governments are increasingly incorporating happiness and life satisfaction into public policies and national
agendas. Gross National Happiness (GNH) was developed in Bhutan as an indicator of population happiness to
inform policies, develop programs, and track national progress, a move praised by the United Nations9. Similarly,
the European Union’s project, Bringing Alternative Indicators into Policy (BRAINPOoL), explored indicators
that can be used for policy-making, including indicators of happiness and well-being9,10. Currently, the United
Kingdom is emphasizing happiness in policy-making and investigating methods to measure happiness among
its citizens9. Given the trend toward using happiness as an aective measure of personal and national success,
it is critical to consider the factors that inuence life satisfaction, given its relation to happiness and well-being.
Here, we recognize and assess the dimensions associated with the socio-political domain to evaluate the impact
of beliefs and attitudes on life satisfaction.
To explore these dimensions, we turn to the world values survey (WVS), which collects data from over 80
countries about various indicators, including life satisfaction and other socio-political factors relevant to the
current study11. is study uses the WVS to assess the relationship between life satisfaction and several socio-
political factors at a micro-level of analysis, including corruption, attitudes toward migration, security, and
OPEN
1Department of Psychology, American University of Sharjah, Sharjah, UAE. 2Department of Computer Science and
Engineering, American University of Sharjah, Sharjah, UAE. 3Department of Mathematics and Statistics, American
University of Sharjah, Sharjah, UAE. *email: aalzaatreh@aus.edu
Content courtesy of Springer Nature, terms of use apply. Rights reserved
2
Vol:.(1234567890)
Scientic Reports | (2024) 14:19677 | https://doi.org/10.1038/s41598-024-70067-5
www.nature.com/scientificreports/
national identity. Following measurement validation, we argue that these four socio-political factors inuence
life satisfaction and assess how our hypothesized conceptual model ts the collected data.
Given the importance of life satisfaction to people around the globe as well as its various positive outcomes,
it is important to understand the interlinked factors that inuence life satisfaction, especially in socio-political
spheres where there is a divide between micro and macro levels of research12. Specically, the current study aims
to develop an understanding of how socio-political beliefs, values, and perceptions can inuence life satisfaction
in the shiing socio-political landscape shaped by COVID-19 using data collected in 2021. We contextualize
individual satisfaction within the broader socio-political scene from a micro-level, ultimately striving to promote
life satisfaction through socio-political policies and initiatives.
First, we use conrmatory factor analysis (CFA) to determine whether items on the WVS reect the factors
they are intended to measure. Second, we used structural equation modeling (SEM) to assess the nature and
strength of the direct relationships between factors (corruption, migration, security, and national identity) and
life satisfaction. Additionally, we investigate how individual-level variables such as country, sex, relationship
status, socioeconomic status, and religious denomination impact life satisfaction in specic cultural contexts.
We then use SEM controlled by country to investigate whether the path structure would maintain its validity
across countries, using the ndings of the multi-group analysis to develop an updated framework for life satis-
faction and its related factors. e goal is to inform strategies at the macro-level of social institutions, political
systems, and organizational structures to enhance life satisfaction at the micro-level of individual perceptions,
attitudes, and beliefs.
Literature review
Many researchers have investigated individual and relational factors that contribute to life satisfaction. Several
authors, for example, have concluded that individual factors such as self-esteem, optimism, extraversion, and
personal control all contribute to life satisfaction13,14. When it comes to relational factors, relationships with
people (e.g. social support) and relationships with God (e.g. religion/spirituality) signicantly inuence life
satisfaction15,16. According to Bohnke17, employment, job satisfaction, and health are also associated with more
life satisfaction. Given the various factors that inuence life satisfaction as well as the many positive psychologi-
cal, physiological, and organizational outcomes of life satisfaction, it is increasingly important to analyze life
satisfaction in a post-pandemic world where anxiety, burnout, poor sleep quality, depressive symptoms, cognitive
impairments, and poor quality of life have been on the rise18.
Perceived corruption
Corruption involves abusing and exploiting public resources for the purposes of personal interest and private
gain19. Corruption threatens democratic outputs and policies20, reducing economic investments, exacerbating
economic inequality, and undermining the quality of institutions in a state21. Research suggests that individu-
als residing in nations characterized by a high prevalence of corruption tend to experience lower levels of life
satisfaction compared to those residing in countries where corruption is less prevalent22.
Studies indicate that the perception of governmental corruption is associated with a negative impact on sub-
jective well-being23. Perceived corruption erodes the relationship between individuals and the state. Importantly,
individuals who perceive more corruption tend to be less trusting of their governments compared to those who
perceive less corruption, especially when they cannot punish corrupt individuals and hold them accountable21,24.
With a weak relationship to the state, individuals who perceive more corruption report lower levels of satisfaction
with life and subjective well-being than individuals who perceive less corruption25. Given the adverse eects of
individuals’ perceptions of corruption on their subjective well-being, we investigate whether beliefs about cor-
ruption inuence life satisfaction in the current geopolitical context.
H1 Perceived corruption has a signicant impact on individuals’ satisfaction
Attitudes toward migration
With the rise of globalization, the number of migrants has increased, reaching 272 million migrants in 2019
worldwide, with over 750 million individuals expressing a desire to migrate if they could26,27. While there are con-
tested denitions of what makes someone a migrant, a typical migrant is someone who moves to a new country
for a period of time long enough for the new country to become their usual residence28. ere are both economic
and non-economic determinants of attitudes toward migration. Economically, anti-immigration attitudes arise
when the skill set of citizens (natives) is similar to that of immigrants29. When the skill sets are too similar,
competition arises in the labor market, and anti-immigrant attitudes increase and become more salient than
when the skill sets of citizens and immigrants are dierent and no economic competition ensues30. Apart from
economic considerations, racial, ethnic, and cultural considerations are also important31. For example, Dustmann
and Preston32 found that racial and cultural prejudice are important contributors to attitudes toward migration.
Both economic and non-economic factors play a signicant role in attitudes toward migration. Namely,
negative attitudes toward migration are oen shaped by a feeling of threat, such as a threat to one’s employment,
safety and order, or cultural values and practices33,34. Perceived threats, such as ones related to immigrants, can
undermine life satisfaction. For example, perceived nancial threats diminish life satisfaction35, and attitudes
toward migration negatively impact life satisfaction among Europeans36. Here, we use dierent threats to deter-
mine attitudes toward migration, and then test whether endorsed attitudes predict life satisfaction.
H2 Attitudes toward migration have a signicant impact on individuals’ satisfaction
Content courtesy of Springer Nature, terms of use apply. Rights reserved
3
Vol.:(0123456789)
Scientic Reports | (2024) 14:19677 | https://doi.org/10.1038/s41598-024-70067-5
www.nature.com/scientificreports/
Perceived security
Looking at how economic competition and racial intolerance inuence migration attitudes, other factors inu-
ence perceived security, which is another socio-political factor that can develop our understanding of life satisfac-
tion. Human security is characterized by safety and protection from harmful, chronic, or unanticipated threats
(e.g. disease)37. While feelings of safety and protection from threats might not reect reality, they signicantly
contribute to one’s overall sense of perceived security. Notably, perceived security is negatively linked to the fear
of war, fear of crime, and safety anxiety. For example, research indicates that individuals who value security
reported greater fear of war38. Such fears can have a negative impact on both the individual and society. On the
individual, evidence suggests that safety anxiety can restrict freedom of mobility out of fear of being a target of
harm, inducing a state of chronic hypervigilance39.
At the societal level, the fear of crime increases economic expenditure, violent attitudes, and crime; fear of
crime also reduces trust in authorities and willingness to help40. Recognizing the adverse eects of diminished
perceived security on both the individual and society, it becomes clear that perceived security is an important
contributor to subjective well-being. Studies have demonstrated that participants who reported feeling unsafe
exhibited lower levels of life satisfaction41. In another study, fear of crime was negatively associated with life
satisfaction42. Seeing how important feeling safe is to one’s well-being, we include perceived security as a factor
in this analysis to explore its relationship and impact on life satisfaction, taking into consideration its interac-
tion with other socio-political factors that constitute threats, such as migration. Although previous research has
studied security, this study reexamines the impact of perceived security on life satisfaction within the distinct
socio-political contexts of selected countries, avoiding the generalization of existing ndings to specic contexts.
H3 Perceived security has a signicant impact on individuals’ satisfaction
National identity
National identity, while abstract and multidimensional, can be an important aspect of one’s social identity43.
National identity encompasses the territorial distinctiveness of cultural groups, shared origin myths and historical
memories, a unied mass culture, territorial division of labor and resource ownership, and a common system
of legal rights and duties44. National identity relates to how individuals associate with characteristics unique to
their nation-state, including religion, history, customs, and social institutions45.
Evidence suggests that a stronger sense of national identity correlates with positive psychological outcomes.
For example, Khan and colleagues46 found that a strong association with one’s nation predicted improved health
outcomes and lower levels of anxiety. Researchers have also found evidence between national identity and self-
esteem, post-traumatic growth, interpersonal trust, and subjective well-being47.
A stronger national identity is also linked to higher levels of life satisfaction48. Across ethnic and cultural
groups, there is a positive relationship between how strongly one identies with their nation and their satisfac-
tion with life48–50. When individuals identify with the nation, they develop meaning systems that stem from the
nation’s values, beliefs, and attitudes43. National identity provides individuals with a sense of belonging, connec-
tion to a greater purpose, and meaning in life, thereby enhancing life satisfaction43.
However, the relationship between life satisfaction and national identity is not culturally robust and is an area
of further research. For example, Jordanov etal.50 found that national identity was associated with life satisfac-
tion among Romanian youth, but this relationship did not extend to Bulgarians. Similarly, national identity did
not signicantly predict subjective well-being among Qataris following a period of national adversity51. Here,
we look at global trends of national identity across 50 countries, providing cross-cultural evidence of the link
between a stronger identication with the nation and an individual’s life satisfaction.
H4 e strength of national identity has a signicant impact on individuals’ satisfaction
Demographic variables
Although socio-political factors are important for understanding life satisfaction, demographic factors can also
shape our experiences and emotions, including life satisfaction. However, research on the impact of demographic
variables on life satisfaction provides mixed results and presents conicting evidence on their role in inuencing
life satisfaction. erefore, it is essential to investigate how demographic variables of sex, relationship status,
social class, religious denomination, and country relate to life satisfaction in the context of the present study.
Indeed, it is important not to make assumptions about generalizability, as boundary conditions such as culture
or environment can vary across ecological contexts.
Relationship status
Even though being in a relationship fullls some of our most essential social needs, the number of single individu-
als has been on the rise as people nd it more dicult to date in the present day52. In 2019, 38% of U.S. adults
reported being single53. Specically, single individuals who are not looking for a relationship report that they
enjoyed being single or had more important life priorities54. However, being single is also associated with worse
educational and economic outcomes than being in a relationship53. Additionally, empirical evidence indicates
that individuals in relationships tend to report higher levels of life satisfaction compared to single individuals. For
example, Adamczyk and Segrin55 found that individuals who were single reported lower levels of life satisfaction,
social support from a signicant other, and loneliness.
However, other studies present ndings that life stressors and changes can impact relationships, as stressors
are associated with reduced desire for emotional and physical closeness56. In addition to the stresses of everyday
Content courtesy of Springer Nature, terms of use apply. Rights reserved
4
Vol:.(1234567890)
Scientic Reports | (2024) 14:19677 | https://doi.org/10.1038/s41598-024-70067-5
www.nature.com/scientificreports/
life, social media and COVID-19 have had signicant impacts on the nature and perceptions of dating, relation-
ships, and social support. For example, social media use is associated with conict in relationships57. Additionally,
while COVID-19 presented new opportunities for couples to spend time together and promote intimacy in the
relationship, it also presented new interpersonal challenges, creating new sources of distress and dysfunction in
relationships58. As a source of both opportunities and challenges, the complexity of relationships is highlighted in
the literature. For example, Adamczyk and Segrin55 found no direct eect of relationship status on life satisfaction.
Instead, the indirect eect of relationship status on life satisfaction was through social support and loneliness.
As societal expectations of marriage and relationships are changing in a world dominated by daily stressors,
dating apps, and social media, alternative paths of social support present themselves, providing alternatives for
meeting social needs and altering the impact of relationship status on individual well-being. Importantly, the
current research looks at whether the eect of relationship status on life satisfaction in the contexts of interest is
consistent with the research, especially given the dynamic and changing nature of the landscape of relationships.
Social class
Social class and socioeconomic status can signicantly shape our lives, both through economic and social means.
At the national level, countries with higher incomes have greater life satisfaction than low-income countries59.
Specically, socioeconomic status can inuence life satisfaction by satisfying the needs of individuals60. It is
unsurprising, then, that there is a positive relationship between socioeconomic status and life satisfaction61.
Research on data from 41 countries found that individuals from more auent families reported higher levels
of life satisfaction compared to those from less auent families62. Here, we dene social class as encompassing
social and economic factors that categorize individuals in a society. For example, social class can be determined
by access to nancial resources, educational opportunities, occupational prestige, and social standing63. Speci-
cally, we categorize social class into upper class, upper middle class, lower middle class, working class, and lower
class based on respondents’ self-categorization64 (see Fig.1).
Importantly, cross-national research on life satisfaction has oen excluded countries investigated in this
research. For example, Armenia is sometimes excluded from analyses in satisfaction research due to outlying
scores62,65. While national-level data is available for countries such as Venezuela, Kenya, Morocco, and the Mal-
dives, their populations remain understudied. In particular, ecological factors in these countries might yield
patterns of results that diverge from the literature and the traditional relation between socioeconomic status and
life satisfaction66. For example, the sudden shi to a market in countries with a history of communism, such as
Armenia, can cause periods of transitional stress, reducing life satisfaction67. Consequently, our research sought
to determine whether the positive relation between socioeconomic status and life satisfaction would extend to
these countries.
Sex
Dierences in biological sex are oen associated with dierences in aect, cognition, and behavior68. For example,
males are less prone to developing psychological problems, more likely to outperform females on motor perfor-
mance and mental rotation tasks, and engage in more risk-taking behaviors than females69,70. Despite research-
ers commonly including sex in studies about life satisfaction, the relationship between life satisfaction and sex
Figure1. Social class pyramid. Upper class, upper middle class, lower middle class, working class, and lower
class descriptions and proportions in a typical capitalistic society, such as the U.S.64.
Content courtesy of Springer Nature, terms of use apply. Rights reserved
5
Vol.:(0123456789)
Scientic Reports | (2024) 14:19677 | https://doi.org/10.1038/s41598-024-70067-5
www.nature.com/scientificreports/
remains unclear71. While some studies report that males tend to be less satised than females, other studies have
reported opposite ndings71. Some research has found no dierences in satisfaction between males and females72.
Many researchers have noted that environmental factors such as culture, gender stereotypes, gender equality,
learning experiences, and biosocial interactions all inuence the current understanding of gender dierences70.
Taking account of such environmental inuences, recent research has challenged some ideas about dierences
between sexes, and recent ndings support that there are little to no sex dierences in most domains70. Given
the lack of clarity and consensus on the current inuence of sex on life satisfaction, we sought to examine the
relationship between the two.
Religious aliation
Religion characterizes both individuals and societies. Okulicz-Kozaryn73 argued that, while social religiosity
encompasses attending religious services, spending time with others at religious events, and belonging to reli-
gious organizations, individual religiosity relates to the belief in a higher being and the importance of religion.
Literature on the relation between religiosity and life satisfaction presents conicting evidence, with 80% of stud-
ies reporting a positive association, 13% of studies reporting no relationship, 7% nding mixed results, and one
study reporting a negative correlation between the two74. Despite the majority of studies supporting the positive
association between religion and life satisfaction, recent research has challenged the inuence of religion on life
satisfaction. For example, Lim and Putnam75 found that social religiosity facilitates social support and develops
social networks, thereby indirectly enhancing life satisfaction. ey found little evidence, however, that aspects
of individual religiosity inuence life satisfaction75.
Such ndings on social religiosity align with other studies comparing religious aliates to atheists and agnos-
tics. For example, Hayward etal.76 found that religious aliates had better psychological well-being, social
support, and health behaviors than atheists and agnostics. Other studies nd no dierences in life satisfaction
between religious aliates and non-aliates77. Not only is the research on the relationship between religion and
life satisfaction unclear, but it is also contextual, with religious people reporting more satisfaction in religious than
non-religious nations73,78. Atheists also had higher levels of well-being in countries with secular populations, indi-
cating that social context is important for life satisfaction with respect to religion77. However, Uğur and Aydın78
found that most non-religious individuals did not feel pressured by societal expectations regarding religiosity.
Overall, the conicting evidence in the literature as well as the context-dependent nature of the relation between
religion and life satisfaction make religious aliation an important variable to examine in the present study.
Country
Across countries, life satisfaction can dier due to a variety of ecological factors. Chapman etal.79 highlighted
the role of country or region in the way participants responded to surveys on life satisfaction. Even when vari-
ables are controlled, individuals report dierent levels of life satisfaction across socio-cultural environments,
suggesting that the relationship between country and life satisfaction is more complex than it appears to be80.
Specically, cross-national research provides evidence that national indicators such as national wealth, environ-
mental conditions, and human development are important when considering life satisfaction in a regional or
national context81. Additionally, being satised with one’s country is positively correlated with their subjective
well-being, suggesting a relation between national-level and individual-level factors82.
At the national level, economic factors such as labor market policies, energy aordability, and economic
growth have been shown to inuence life satisfaction79,83. However, national conditions related to security, health,
equality, and politics are also associated with life satisfaction, suggesting that looking at the eects of regional
context is complex and not adequately accounted for using economic indicators79,84. While economic factors
might not completely account for life satisfaction, they still play an important role at the national level. For
example, post-material values such as personal autonomy are more important for life satisfaction than material
values such as income in auent countries85. In the current research, we investigate data collected in ve coun-
tries, including Armenia, Kenya, Morocco, Maldives, and Venezuela. In 2021, the gross domestic product (GDP)
varied across these countries, with US$9808 per capita in the Maldives, US$4522 per capita in Armenia, US$3291
per capita in Morocco, US$2070 per capita in Venezuela, and US$1705 per capita in Kenya86,87. Research on the
relationship between GDP and life satisfaction varies. While some evidence supports the positive relationship
between GDP and life satisfaction88, Proto and Rustichini89 depict a more complex relationship between the two
variables. Since life satisfaction signicantly varies at the national level, we sought to determine the impact of
one’s country on life satisfaction in the ve aforementioned countries81.
In summary, while evidence indicates that socio-political factors are important for life satisfaction, there is
little research providing information on how to cluster such factors that most inuence life satisfaction, leaving
a gap in the literature that we intend to address. Here, we examine the relationship between life satisfaction and
perceived corruption, attitudes toward migration, perceived security, and strength of national identity. Addition-
ally, because they are impacted by contextual factors, mixed results on the impact of individual-level variables
such as sex, relationship satisfaction, social class, and religious denomination on life satisfaction leave another
gap in the literature. Here, we investigate the role of these demographic variables on life satisfaction within the
relevant cultural context. Our research, therefore, seeks to address these two gaps and promote life satisfaction.
Methodology
In this study, the methodology involved several key steps:
We used data from the seventh wave of the World Values Survey (WVS) conducted in 2021. Data collec-
tion of the seventh wave, the most recent wave, began in 2017 and concluded in 2021 in collaboration with
academic, research, and non-governmental organizations11. e survey in the seventh wave includes questions
Content courtesy of Springer Nature, terms of use apply. Rights reserved
6
Vol:.(1234567890)
Scientic Reports | (2024) 14:19677 | https://doi.org/10.1038/s41598-024-70067-5
www.nature.com/scientificreports/
about self-reported social beliefs, attitudes, and values, including happiness and well-being, corruption, migra-
tion, religious values, and others. For its survey, the WVS gathered data through interviews with individuals aged
18 and above11. e data are freely available and can be accessed at https:// www. world value ssurv ey. org/ wvs. jsp.
We began with cleaning the WVS dataset to include associated dimensions and their corresponding items,
based on the categorization in the WVS and related literature11,90. We then used measurement validation to select
items that measure each dimension. Missing data were removed, and full information maximum likelihood was
used in the Structural Equation Modeling (SEM). Although the primary language of the master questionnaire
is English, the language of instruction was translated based on the linguistic context of each region to enable
speakers of dierent languages to respond.
Secondly, we conducted a conrmatory factor analysis (CFA) to measure the four socio-political dimensions
(perceived corruption, attitudes toward migration, perceived security, and strength of national identity) using
the validated items from the WVS and model validation using goodness of t tests. Finally, we conducted SEM
to evaluate the impact of the exogenous dimensions (perceived corruption, attitudes toward migration, perceived
security, and national identity) on the endogenous dimension (life satisfaction).
Survey design
Based on the previous literature review, we used questions from the WVS addressing demographic information,
life satisfaction, and four factors related to socio-political beliefs, including perceived corruption, attitudes toward
migration, perceived security, and strength of national identity. Items for each factor were selected based on the
pre-existing classication by the World Values Survey (WVS), related literature, reliability analyses, and CFA
analysis (refer to Sect. "Data analysis" for more details)11,90. In this section, we outline each measure, including
its items and the scale it uses. e following section explains how items were selected for each factor. Details
of selected items from the WVS, corresponding coded items, and a complete list of questions, are provided in
TableS1.
Demographics
While the WVS includes many demographic variables, we focused on key demographic questions about sex, age,
ethnicity, religious denomination, and socioeconomic status to contextualize our sample and understand the
background of our responses. Additionally, we examined variables such as relationship status, employment status,
and country. Data collected in 2021 overlapped with the COVID-19 pandemic, which greatly contributed to
reshaping some demographic variables. Economically, the pandemic caused extensive unemployment, economic
instability, and income uctuations, leading to a signicant gap between dierent social classes. Additionally,
social distancing presented psychological challenges globally, signicantly impacting emotional connections
and mental health, as well as potentially straining relationships91.
e data primarily included responses from ve countries: Armenia, Kenya, Morocco, Maldives, and Ven-
ezuela (See Fig.2). e nal cleaned data consists of 3334 participants (1628 Males, 1706 Females) with a mean
age of 36.67 (SD = 14.17). Participants were asked about their socioeconomic status (Upper Class = 1.7%, Upper
Middle Class = 17.7%, Lower Middle Class = 42.7%, Working Class = 24.7%, Lower Class = 13.2%). For their
ethnic groups, 35.14% of the participants were White, 11.84% were Black, 13.48% were South Asian Indian or
Pakistani, 2.86% were East Asian Chinese or Japanese, 33.23% were Arab or Central Asian, and 3.45% reported
that they identied with “other” ethnic groups. For their religious denomination, 27.9% were Roman Catholic,
Figure2. Bar chart showing the percentage of responses from each Country (2021). Data was collected from
Armenia (ARM), Kenya (KEN), Morocco (MOR), Venezuela (VEN), and Maldives (MAD).
Content courtesy of Springer Nature, terms of use apply. Rights reserved
7
Vol.:(0123456789)
Scientic Reports | (2024) 14:19677 | https://doi.org/10.1038/s41598-024-70067-5
www.nature.com/scientificreports/
17.27% were Protestant, 7.6% were Orthodox, 0.1% were Jew, 37.1% were Muslim, and 1.6% reported that they
belonged to “other” religious denominations.
Life satisfaction
We used a single item as a measure of life satisfaction (See TableS1 for a full list of items). e question meas-
ured the extent to which participants were satised with their lives as a whole in the present. e life satisfaction
question asks: All things considered, how satised are you with your life as a whole these days? e item uses
a 10-point Likert-type scale ranging from “Completely dissatised” (1) to “Completely satised” (10). Higher
scores on the ten items indicate higher levels of life satisfaction.
Corruption
Four items were selected C1, C2, C3, and C4, as a measure of perceived corruption (See TableS1 for a full list of
items). ese questions measured the extent to which participants perceived corruption among dierent social
groups, including state authorities, business executives, local authorities, and civil service providers. e items
use a 4-point Likert-type scale ranging from “none of them” (1) to “all of them” (4). Higher scores on the four
items indicate higher levels of perceived corruption.
Migration
Four items were selected (M1, M2, M3, and M4) as a measure of attitudes toward migration (See TableS1 for a
full list of items). ese questions measured participants’ self-reported beliefs about the eects of immigration
on the development of their country, including the eect of immigration on crime rates, risk of terrorism, unem-
ployment, and social conict. e items use a 3-point Likert-type scale ranging from “Agree” (2) to “Disagree”
(0). Higher scores on the three items indicate more negative attitudes toward migration.
Security
ree items were selected (S1, S2, and S3) as a measure of perceived security (See TableS1 for a full list of items).
ese questions measured participants’ self-reported attitudes about the extent to which they were worried about
their security from dierent threats, including a war involving their country, a terrorist attack, and a civil war.
e items use a 4-point Likert-type scale ranging from “Very much” (1) to “Not at all” (4). Higher scores on the
four items indicate less worry about these threats, so they indicate higher levels of perceived security.
National identity
Four items were selected (N1, N2, N3, and N4) as a measure of the strength of national identity (See TableS1).
ese questions measured participants’ self-reported pride in their country and closeness to dierent bodies,
including their town, city, and country. e items use a 4-point Likert-type scale ranging from “Very proud” (1)
to “Not proud at all” (4) and “Very proud”, and “Very close” (1) to “Not close at all” (4). Higher scores on the
four items indicate a weaker sense of national identity; items were reverse scored for higher scores to reect a
stronger identity.
Data analysis
Measurement validation
As mentioned above, we selected a specic set of questions to measure each factor. is selection was based on
the classication from the World Values Survey and related literature, the average variance extracted, Cronbach’s
α, the estimate of the standardized path list, and the p-value of the unstandardized path list11,90. Cronbach’s α is
a measure of internal consistency used to assess how well each item relates to the construct. First, we conducted
a reliability analysis for each construct, removing unreliable indicators during this process. en, we performed
path analysis using conrmatory factor analysis (CFA). Insignicant indicators were iteratively removed to rene
the model. We aimed to maintain an acceptable level of Cronbach’s alpha for each construct and acceptable
standardized path coecient (0.6 and above), and a signicant p-value (less than 5%) for each indicator. e nal
list of indicators includes questions C1 to C4 for corruption, M1 to M4 for Migration, S1 to S3 for Security, and
N1 to N4 for National Identity. e nal conceptual model is illustrated in Fig.3. Table1 presents the reliability
measures, which include the standardized correlation coecient with total observed indicators within the same
construct and Cronbach’s α. From Table1, Cronbach’s alpha values are higher or close to the acceptable reli-
ability value of 0.7, and the standardized correlation coecient with the total shows a high level of association.
e model above has several hypotheses:
H1: Perceived corruption has a signicant impact on individuals’ satisfaction.
H2: Attitudes toward migration have a signicant impact on individuals’ satisfaction.
H3: Perceived security has a signicant impact on individuals’ satisfaction.
H4: e strength of national identity has a signicant impact on individuals’ satisfaction.
Average variance extracted and conrmatory factor analysis
First convergent validity assessment using the average variance extracted (AVE) values was conducted. Table1
presents the AVE values, all of which exceed the required minimum threshold of 0.5092. Next, we tested discrimi-
nant validity, by comparing all the diagonal values, which are the square root of the AVEs, with the correlations
between the constructs in the o-diagonal position as shown in Table2. Overall, the results showed that the
square roots of the AVEs for the constructs Corruption, Migration, Security, and National Identity are higher
than the correlations of these constructs with other latent variables in the path model, which indicates that the
Content courtesy of Springer Nature, terms of use apply. Rights reserved
8
Vol:.(1234567890)
Scientic Reports | (2024) 14:19677 | https://doi.org/10.1038/s41598-024-70067-5
www.nature.com/scientificreports/
construct measures empirically demonstrate discriminant validity93. Additionally, we performed a conrmatory
factor analysis (CFA) to ensure that no further indicators needed to be removed. e results of the CFA, includ-
ing the parameter estimates, p-values, and standardized path estimates, are reported in Table1. All p-values are
signicant at the 5% error level. In addition, the standardized parameter estimate for each indicator is over 0.6.
Table3 summarizes the goodness-of-t statistics for the CFA model. e SRMR is below 0.05, while both the
GFI and BCFI are above 0.9, indicating an acceptable t.
Figure3. Conceptual model.
Table 1. Survey measurement validation.
Factor Correlation with total Cronbach’s, α AVE Estimate (p-value) Standardized estimate
Corruption 0.833287 0.665
C1 0.764506 0.65355 (< 0.0001) 0.81774
C2 0.820995 0.50760 (< 0.0001) 0.64783
C3 0.763213 0.65192 (< 0.0001) 0.82444
C4 0.805368 0.57019 (< 0.0001) 0.69648
Migration 0.738901 0.523
M1 0.667917 0.58361 (< 0.0001) 0.67450
M2 0.686849 0.55185 (< 0.0001) 0.63638
M3 0.683055 0.53038 (< 0.0001) 0.62911
M4 0.680846 0.54506 (< 0.0001) 0.63505
Security 0.884911 0.799
S1 0.839939 0.88703(< 0.0001) 0.84189
S2 0.804425 0.96761 (< 0.0001) 0.90644
S3 0.865450 0.88800 (< 0.0001) 0.79950
National identity 0.731415 0.540
N1 0.324266 0.33539(< 0.0001) 0.38248
N2 0.573686 0.54303 (< 0.0001) 0.72970
N3 0.646252 0.74480 (< 0.0001) 0.75584
N4 0.568497 0.50793(< 0.0001) 0.63707
Table 2. Fornell–Larcker criterion table.
Factor Corruption Migration Security National Identity
Corruption 0.815 – – –
Migration 0.017 0.723 – –
Security − 0.055 − 0.172 0.894 –
National identity − 0.008 0.034 − 0.071 0.735
Content courtesy of Springer Nature, terms of use apply. Rights reserved
9
Vol.:(0123456789)
Scientic Reports | (2024) 14:19677 | https://doi.org/10.1038/s41598-024-70067-5
www.nature.com/scientificreports/
Structural equation modeling
To understand the impact of the exogenous factors (corruption, migration, security, and national identity) on the
endogenous construct (satisfaction), we used structural equation modeling (SEM). e polychoric correlation
was used in the SEM analysis to address the ordinal nature of the outcome variable. e analysis was performed
using PROC CALIS in SAS soware. e model’s goodness-of-t results, shown in Table4, indicate an accept-
able t. e path estimates, and their p-values are summarized in Table5. According to Table5, all indicators are
signicant (p-value < 0.05). For the structural relationships, satisfaction is negatively impacted by corruption and
national identity, and positively impacted by security. On the other hand, migration does not have a signicant
impact on satisfaction with a p-value of 0.1491.
Multiple‑group models
Conducting a multi-group analysis is used to test the path signicance across dierent countries. For instance,
countries may vary in their support for migration and levels of polarization94. Additionally, national identity
varies in countries due to factors such as historical legacies, state-building strategies, and social policies95. ere-
fore, employing multi-group analysis is essential to comprehend these variations. To determine if the same path
structure was valid, based on the country demographic, fully constrained SEM models were created for the
following countries: Kenya, Armenia, Maldives, Morocco, and Venezuela. All parameters were held identically
across all models. e results indicated that the path structures were consistent. e goodness-of-t statistics
Table 3. Goodness of t for CFA.
Criteria Valu e
Standardized root mean square residual (SRMR) 0.0260
Goodness of t index (GFI) 0.9808
Bentler comparative t index (BCFI) 0.9788
Table 4. Measures of t for SEM modeling.
Criteria Valu e
Standardized root mean square residual 0.0552
Goodness of t index 0.9401
Bentler comparative t index (incremental index) 0.9443
Table 5. SEM path list.
Path Standardized estimate P-value for the unstandardized estimate
Corruption → C1 0.86264 < 0.0001
Corruption → C2 0.81180 < 0.0001
Corruption → C3 0.82486 < 0.0001
Corruption → C4 0.72372 < 0.0001
Migration → M1 0.72966 < 0.0001
Migration → M2 0.58491 < 0.0001
Migration → M3 0.63178 < 0.0001
Migration → M4 0.64293 < 0.0001
Security → S1 0.83386 < 0.0001
Security → S2 0.98681 < 0.0001
Security → S3 0.77055 < 0.0001
National → N1 0.37929 < 0.0001
National → N2 0.79235 < 0.0001
National → N3 0.99082 < 0.0001
National → N4 0.81842 < 0.0001
Satisfaction → Q49 1.01156 < 0.0001
Corruption → Satisfaction − 0.16594 < 0.0001
Security → Satisfaction 0.09954 0.0011
Migration → Satisfaction − 0.04919 0.1491
National → Satisfaction 0.08116 0.0050
Content courtesy of Springer Nature, terms of use apply. Rights reserved
10
Vol:.(1234567890)
Scientic Reports | (2024) 14:19677 | https://doi.org/10.1038/s41598-024-70067-5
www.nature.com/scientificreports/
for the multi-group SEM models demonstrate a good t, as illustrated in Table6. Moreover, the parameters for
security and migration were found to be insignicant, with p-values over 0.05.
Although both security and migration are individually insignicant, there is a signicant correlation between
these constructs across all countries, as evidenced by the covariance matrix with a p-value of (< 0.0001). Upon
further investigation, including a review of the literature, we found a notable relationship in which negative atti-
tudes toward migration inuence perceived security. e association between migration attitudes and perceived
security aligns with realistic group conict theory (RGCT), as competition for resources primes negative attitudes
toward out-groups96,97. Accordingly, negative attitudes toward out-group members, such as immigrants, lead to
heightened feelings of threat and a lack of perceived security, including feelings of fear, anxiety, and stress97,98.
Consequently, we modied the framework by adding a path from migration to security, leading to a direct
relationship between migration and security, as well as an indirect eect of migration on life satisfaction through
security. TablesS2 and S3 show the goodness-of-t and the SEM path list respectively. e modied framework,
illustrated in Fig.4, yielded the signicance of all factors across all countries in 2021. e goodness-of-t statistics
for the multi-group modied SEM models, as shown in Table7, demonstrate a good model t.
Validation. To capture potential variations in the model across years, SEM was analyzed, but we included
data from multiple years (2017 to 2021), rather than limiting it to 2021. is broader scope provides a more
comprehensive understanding of the relationships between Migration, Corruption, Security, National Identity,
and Satisfaction. e model’s goodness-of-t results are shown in Table8, indicating an acceptable t across all
years as GFI and BCFI are both higher than 0.9. e path estimates, standard errors, and their p-values are sum-
marized in Table8. According to Table8, all indicators are signicant (p-value < 0.05) in each year analyzed. For
the structural relationships, satisfaction is negatively impacted by corruption in all years, while national identity
generally has a positive impact. Security has positive impact on satisfaction while Migration consistently shows
a negative impact on security across all years studied.
e countries participating each year varied. In 2017, the respondents were from Greece (7%), Serbia (14%),
Argentina (5%), Bolivia (19%), Russia (10%), and the USA (45%). In 2018, the respondents came from Andorra
Table 6. Measure of t for multi-group models—original framework.
Criteria Valu e
Standardized root mean square residual 0.1832
Chi-square 2171.9193
Chi-square DF 545
Goodness of t index 0.9320
Bentler comparative t index (incremental index) 0.9151
Figure4. Conceptual model of the modied framework.
Table 7. Measure of t for multi-group models—modied framework.
Criteria Valu e
Standardized root mean square residual 0.0855
Chi-square 1892.1533
Chi-square DF 440
Goodness of t index 0.9338
Bentler comparative t index (incremental index) 0.9231
Content courtesy of Springer Nature, terms of use apply. Rights reserved
11
Vol.:(0123456789)
Scientic Reports | (2024) 14:19677 | https://doi.org/10.1038/s41598-024-70067-5
www.nature.com/scientificreports/
(4%), Australia (6%), Bangladesh (2%), Brazil (2%), Chile (1%), Colombia (7%), Germany (4%), Ecuador (4%),
Hong Kong (8%), Indonesia (10%), Iraq (3%), Kazakhstan (2%), South Korea (7%), Lebanon (3%), Mexico (6%),
Malaysia (7%), Nigeria (4%), Pakistan (4%), Peru (4%), Puerto Rico (4%), Romania (1%), ailand (4%), and
Turkey (3%). In 2019, the respondents were from Cyprus (4%), Japan (3%), Macao (18%), Philippines (32%),
Tunisia (16%), and Taiwan (27%). In 2020, the respondents included Canada (42%), Ethiopia (4%), Guatemala
(9%), Iran (4%), Kyrgyzstan (4%), Mongolia (17%), Nicaragua (8%), New Zealand (2%), Ukraine (1%), and
Zimbabwe (9%). Finally, in 2021, the respondents were from Armenia (8%), Kenya (20%), Morocco (31%),
Maldives (10%), and Venezuela (31%).
Ordinal logistic regression analysis (OLR)
To determine how dierent demographics predict life satisfaction, we developed an ordinal logistic regression
model, focusing on 2021 data. In this model, we treat the satisfaction variable from the survey data as ordinal
rather than continuous due to its representation of ordered categories. e scale from 1 to 10 reects dier-
ent levels of satisfaction, but the numerical dierences between the ratings may not consistently denote equal
increments in satisfaction, making it more appropriate to consider it as an ordinal scale for statistical analysis.
e model included several demographic factors, such as country, gender, relationship status, social class,
and religion. e countries involved in the study were Armenia, Kenya, Venezuela, Morocco, and the Maldives.
To simplify the analysis, relationship status and religion were recoded as binary variables, indicating whether the
respondent was in a relationship or not, and whether they identied as religious or not. Table9 below presents
the Type 3 analysis of eects results of the ordinal regression, which explains the statistical signicance of each
demographic while controlling for the eect of the others. Interestingly, gender was not statistically signicant
(p-values > 0.05), while country, relationship status, social class, and religion were signicant (p-values < 0.05)
predictors of life satisfaction.
e ordinal logistic regression analysis provides insights into the associations between the response variable
and various predictor variables using the Odds Ratios. Table10 presents the odds ratios for the signicant vari-
ables. As shown in Table10 below, all countries seem to have a higher satisfaction compared to Kenya. When
Table 8. SEM path list—modied framework.
Year Path list Standardized estimate P-value for the unstandardized
estimate Goodness of t index Bentler comparative
t index
2017
Migration → Secur ity − 0.1718 < 0.0001
0.9494 0.9305
Corruption → Satisfaction − 0.0460 < 0.0001
Security → Satisfaction 0.01910 < 0.0001
National Identity → Satisfaction 0.1325 < 0.0001
2018
Migration → Secur ity − 0.1641 < 0.0001
0.9549 0.9459
Corruption → Satisfaction − 0.0480 < 0.0001
Security → Satisfaction 0.0208 0.0002
National Identity → Satisfaction 0.1380 < 0.0001
2019
Migration → Secur ity − 0.1625 < 0.0001
0.9467 0.9413
Corruption → Satisfaction − 0.0467 < 0.0001
Security → Satisfaction 0.0205 0.0002
National Identity → Satisfaction 0.1345 < 0.0001
2020
Migration → Secur ity − 0.1664 < 0.0001
0.9617 0.9644
Corruption → Satisfaction − 0.0417 < 0.0001
Security → Satisfaction 0.0179 0.0002
National Identity → Satisfaction 0.1199 < 0.0001
2021
Migration → Secur ity − 0.2031 < 0.0001
0.9750 0.9668
Corruption → Satisfaction − 0.0562 < 0.0001
Security → Satisfaction 0.0197 0.0002
National Identity → Satisfaction 0.1618 < 0.0001
Table 9. OLR type 3 analysis of eects.
Eect P-value
Country < 0.0001
Gender 0.7790
Relationship 0.0009
Social class 0.0123
Religion 0.0034
Content courtesy of Springer Nature, terms of use apply. Rights reserved
12
Vol:.(1234567890)
Scientic Reports | (2024) 14:19677 | https://doi.org/10.1038/s41598-024-70067-5
www.nature.com/scientificreports/
comparing individuals in a relationship to those who are not, the odds of satisfaction are 22.7% higher for indi-
viduals in a relationship. As for socioeconomic class, the odds of satisfaction for the lower middle class are 32.8%
higher compared to individuals in the lower class. In addition, individuals in the upper class and upper middle
class have odds that are 46.4% and 45.5% greater, respectively, than those in the lower class. Moreover, individuals
who identied as religious have 42.8% higher odds of satisfaction than individuals who identied as not religious.
Discussion and conclusion
In this paper, we explored how dierent socio-political factors contribute to individual life satisfaction. We
developed a conceptual model, as shown in Fig.3, illustrating how corruption, migration, security, and national
identity inuence life satisfaction. rough reliability analysis, AVE, and CFA, we selected relevant items in the
WVS to measure each factor. en, using these items, we constructed an SEM to evaluate the eect of each socio-
political dimension on life satisfaction. We found that perceived corruption, perceived security, and national
identity were signicant dimensions that impact life satisfaction.
Consistent with previous research, our ndings supported hypothesis H1. e negative relationship between
perceived corruption and life satisfaction aligns with the work of Helliwell25, who found that individuals per-
ceiving more corruption report lower levels of life satisfaction and subjective well-being compared to those
perceiving less corruption. Since perceived corruption relies on the regular interactions between individuals
and their surrounding establishments, it is unsurprising that the many negative outcomes of corruption extend
to life satisfaction99.
Furthermore, results from our SEM model support H3, indicating that perceived security has a positive
impact on life satisfaction. However, this nding did not align with Sortheix and Lönnqvist100, who found that
high security values were associated with lower life satisfaction across 25 European nations. One possible expla-
nation for such results is that high security values reect a greater need to protect oneself and are more likely
to be connected to feelings of threat than security101. Our ndings also support H4, as the strength of national
identity had a signicant impact on individuals’ life satisfaction. e positive association in our analyses aligns
with the literature on national identity, and its enhancement of satisfaction with life through providing a sense
of connection and belonging43,48.
e current ndings of our SEM analysis did not support H2, implying that negative attitudes toward migra-
tion have no direct eect on an individual’s satisfaction. Later multi-group analyses examining the impact of
country on the validity of our conceptual model indicated an indirect eect of attitudes toward migration on
life satisfaction through perceived security. e association between negative attitudes toward migration and
perceived security, which was negative, aligns with realistic group conict theory (RGCT), in which negative
attitudes toward outgroups such as immigrants cause heightened perceptions of threat and a lack of security,
including feelings of fear and intergroup anxiety96,98.
We extended our modied framework, including the indirect eect of attitudes toward migration on life
satisfaction, to other years in Wave 7 of the WVS, testing our framework across 42 countries. Findings of the
modied framework indicate signicance across all four socio-political factors, supporting the conceptual model
in Fig.4. We conclude that perceived corruption, perceived security, and national identity have a direct signicant
impact on life satisfaction, while attitudes toward migration have an indirect negative eect on life satisfaction
through perceived security.
Furthermore, we conducted OLR to better understand the relationship between demographic factors and
life satisfaction. As seen in Table9, we found that gender was not a signicant factor, which supports the recent
research challenging gender dierences in outcomes such as life satisfaction70. Given the lack of consensus about
sex dierences in life satisfaction in the literature, our ndings contribute to the literature by determining that
no sex dierences in life satisfaction were found in the examined ecological context.
Our ndings also revealed that relationship status was a signicant predictor of life satisfaction, aligning with
the literature review, which indicates that being in a relationship is signicantly associated with higher levels
of life satisfaction. Single individuals, as shown in the study by Adamczyk & Segrin55, tend to have lower levels
of life satisfaction and social support, and higher levels of loneliness compared to individuals in relationships.
Table 10. Odds ratios for the signicant variables.
Eect Point estimate
Armenia vs Kenya 2.215
Morocco vs Kenya 1.721
Maldives vs Kenya 2.245
Venezuela vs Kenya 2.452
In relationship vs no relationship 1.227
Lower middle vs lower 1.328
Upper vs lower 1.464
Upper middle vs lower 1.455
Working vs lower 1.215
Religious vs not religious 1.428
Content courtesy of Springer Nature, terms of use apply. Rights reserved
13
Vol.:(0123456789)
Scientic Reports | (2024) 14:19677 | https://doi.org/10.1038/s41598-024-70067-5
www.nature.com/scientificreports/
Furthermore, OLR analysis indicates that individuals in a relationship have higher odds of satisfaction than
those not in a relationship.
e analysis also revealed the signicance of social class, which is not surprising given the literature that
links socioeconomic status with life satisfaction by meeting people’s needs, as noted by Gitmez and Morcöl60.
Our results further support the relationship between socioeconomic status and life satisfaction, showing that
individuals in the Lower Middle class, Upper class, and Upper Middle class had higher levels of satisfaction com-
pared to those in the Lower class. is aligns with previous research indicating a positive relationship between
socioeconomic status and life satisfaction. For example, studies have shown that individuals from more auent
families report higher levels of life satisfaction compared to those from less auent families62.
Meanwhile, the OLR model revealed that religious aliation was signicant, which can be supported by
research, like Hayward etal.’s76, which stated that individuals who have religious aliations tend to have enhanced
psychological well-being, social support, and health behaviors compared to those who identify as atheists or
agnostics. Our ndings support this association further since they indicate that religious people have greater odds
of life satisfaction than non-religious people. is is consistent with earlier studies that showed a link between
religion and life satisfaction. For instance, research has shown that social religiosity encourages the growth of
social networks and social support, which in turn indirectly increases life happiness75. erefore, current ndings
oer additional evidence in favor of the positive impacts of religious connection on life happiness.
Lastly, our analysis revealed that country was signicant, which aligns with cross-national research providing
evidence that national indicators such as national wealth, environmental conditions, and human development
are important when considering life satisfaction in a regional or national context81. Kenya’s reported level of life
satisfaction in our analysis, which is the lowest among the studied countries, supports Degutis etal.’s88 research
on the positive relationship between GDP and life satisfaction, as Kenya has the lowest GDP per capita among
countries in 2021.
Overall, our ndings make important contributions to the existing literature by presenting a conceptual model
that examines the relationship between life satisfaction and multiple socio-political factors of perceived corrup-
tion, perceived security, attitudes toward migration, and national identity. Additionally, ndings develop the
current understanding of the impact of demographic variables that are subjects of debate in the literature, such
as sex, relationship status, social class, and religious denomination on life satisfaction. Specically, we determine
how these individual-level variables can shape life satisfaction within the ecological context of the present study.
is study can advance the current understanding of life satisfaction from a socio-political perspective and
can potentially be used to enhance life satisfaction in a population through implementing policies that consider
satisfaction and promoting a local culture of positivity and happiness among community members102. Enhanc-
ing life satisfaction at the micro level using the presented conceptual model can be of particular importance to
governing authorities such as the Ministry of Happiness in the United Arab Emirates, whose main objectives
include promoting happiness, well-being, and development102. Importantly, while national initiatives worldwide
focus on using happiness to inform policies, integrating a cognitive model of life satisfaction can be equally
important for evaluating overall success and well-being103.
Limitations of the current study include the inuence of demographic factors, other than country, on shap-
ing life satisfaction and moderating the eects observed in the modied framework. However, our multi-group
models were limited to country, as existing research supported the relevant impact of country on the studied
factors94,95. As a recommendation for future research, it is important to consider how culture interacts with socio-
political factors, inuencing social and political spheres. Future research could investigate whether countries can
be clustered based on their life satisfaction and their position on the collectivism-individualism scale.
Data availability
e data set used in this research is available at https:// www. world value ssurv ey. org/ WVSDo cumen tatio nWV7.
jsp.
Received: 24 January 2024; Accepted: 12 August 2024
References
1. Veenhoven, R. Questions on happiness: Classical topics, modern answers, blind spots. In Subjective Well‑Being: An Interdisci‑
plinary Perspective (eds Strack, F. et al.) 7–26 (Pergamon Press, 1991).
2. Diener, E., Sapyta, J. J. & Suh, E. Subjective well-being is essential to well-being. Psychol. Inq. 9(1), 33–37. https:// doi. org/ 10.
1207/ s1532 7965p li0901_3 (1998).
3. Beutel, M. E., Glaesmer, H., Wiltink, J., Marian, H. & Brähler, E. Life satisfaction, anxiety, depression and resilience across the
life span of men. Aging Male 13, 32–39. https:// doi. org/ 10. 3109/ 13685 53090 32966 98 (2010).
4. Brand, S. et al. Associations between satisfaction with life, burnout-related emotional and physical exhaustion, and sleep com-
plaints. World J. Biol. Psychiatry 11, 744–754. https:// doi. org/ 10. 3109/ 15622 97100 36242 05 (2010).
5. Chida, Y. & Steptoe, A. Positive psychological well-being and mortality: A quantitative review of prospective observational
studies. Psychosom. Med. 70(7), 741–756. https:// doi. org/ 10. 1097/ PSY. 0b013 e3181 8105ba (2008).
6. Chiron, B., Michinov, E., Olivier-Chiron, E., Laon, M. & Rusch, E. Job satisfaction, life satisfaction and burnout in French
anaesthetists. J. Health Psychol. 15(6), 948–958. https:// doi. org/ 10. 1177/ 13591 05309 360072 (2010).
7. Jones, M. D. Which is a better predictor of job performance: Job satisfaction or life satisfaction. J. Behav. Appl. Manag. 8(1),
20–42 (2006).
8. Rode, J. C., Rehg, M. T., Near, J. P. & Underhill, J. R. e eect of work/family conict on intention to quit: e mediating roles
of job and life satisfaction. Appl. Res. Qual. Life 2(2), 65–82. https:// doi. org/ 10. 1007/ s11482- 007- 9030-6 (2007).
9. Musikanski, L. Happiness in public policy. J. Soc. Change 6(1), 55–85. https:// doi. org/ 10. 5590/ JOSC. 2014. 06.1. 06 (2014).
10. Freeze, R. A. Policy, politics and happiness. Soc. Indic. Res. Ser. https:// doi. org/ 10. 1007/ 978-3- 031- 10913-3_ 12 (2022).
Content courtesy of Springer Nature, terms of use apply. Rights reserved
14
Vol:.(1234567890)
Scientic Reports | (2024) 14:19677 | https://doi.org/10.1038/s41598-024-70067-5
www.nature.com/scientificreports/
11. Haerpfer, C., Inglehart, R., Moreno, A., Welzel, C., Kizilova, K., Diez-Medrano, J., Lagos, M., Norris, P., Ponarin, E. & Puranen
B. World Values Survey Wave 7 (2017–2022) Cross-National Data-Set. Version: 4.0.0. World Values Survey Association. https://
doi. org/ 10. 14281/ 18241. 18 (2022).
12. Goldspink, C. & Kay, R. Bridging the micro–macro divide: A new basis for social science. Hum. Relat. 57(5), 597–618. https://
doi. org/ 10. 1177/ 00187 26704 044311 (2004).
13. Dember, W. N. & Brooks, J. A new instrument for measuring optimism and pessimism: Test-retest reliability and relations with
happiness and religious commitment. Bull. Psychon. Soc. 27(4), 365–366. https:// doi. org/ 10. 3758/ BF033 34629 (1989).
14. Headey, B. & Wearing, A. Understanding Happiness: A eory of Well‑Being (Longman Cheshire, 1992).
15. Mahanta, D. & Aggarwal, M. Eect of perceived social support on life satisfaction of university students. Eur. Acad. Res. 1(6),
1083–1094 (2013).
16. Uğur, Z. B. Does Ramadan aect happiness? Evidence from Turkey. Arch. Psychol. Relig. 40(2–3), 163–175. https:// doi. org/ 10.
1163/ 15736 121- 12341 358 (2018).
17. Bohnke, P. First European Quality of Life Survey: Life Satisfaction, Happiness and Sense of Belonging. European Foundation for
the Improvement of Living and Working Conditions (Oce for Ocial Publications of the European Communities, 2005).
18. Fernández-de-las-Peñas, C. et al. Trajectory curves of post-COVID anxiety/depressive symptoms and sleep quality in previously
hospitalized COVID-19 survivors: e long-COVID-EXP-CM multicenter study. Psychol. Med. https:// doi. org/ 10. 1017/ s0033
29172 20000 6x (2022).
19. Canache, D. & Allison, M. E. Perceptions of political corruption in Latin American democracies. Latin Am. Polit. Soc. 47(3),
91–111. https:// doi. org/ 10. 1111/j. 1548- 2456. 2005. tb003 20.x (2005).
20. Schedler, A., Diamond, L. & Plattner, M. F. e Self‑Restraining State: Power and Accountability in New Democracies (Lynne
Rienner, 1999).
21. Ciziceno, M. & Travaglino, G. A. Perceived corruption and individuals’ life satisfaction: e mediating role of institutional trust.
Soc. Indic. Res. 141(2), 685–701. https:// doi. org/ 10. 1007/ s11205- 018- 1850-2 (2018).
22. Helliwell, J. F. How’s life? Combining individual and national variables to explain subjective well-being. Econ. Model. 20(2),
331–360. https:// doi. org/ 10. 1016/ s0264- 9993(02) 00057-3 (2003).
23. Ma, J., Guo, B. & Yu, Y. Perception of ocial corruption, satisfaction with government performance, and subjective wellbeing:
An empirical study from China. Front. Psychol. https:// doi. org/ 10. 3389/ fpsyg. 2022. 748704 (2022).
24. O’Donnell, G. Horizontal accountability in new democracies. J. Democr. 9(3), 112–126. https:// doi. org/ 10. 1353/ jod. 1998. 0051
(1998).
2 5. Helliwell, J. F. Well-being, social capital and public policy: What’s new?. Econ. J. 116(510), 34–45. https:// doi. org/ 10. 3386/ w11807
(2006).
26. Chamie, J. International migration amid a world in crisis. J. Migr. Hum. Secur. 8(3), 230–245. https:// doi. org/ 10. 1177/ 23315
02420 948796 (2020).
27. Esipova, N., Pugliese A., & Ray, J. More than 750 million worldwide would migrate if they could. Gallup News. https:// news.
gallup. com/ poll/ 245255/ 750- milli on- world wide- migra te. aspx (2018).
28. Anderson, B. Towards a new politics of migration?. Ethnic Rac. Stud. 40(9), 1527–1537. https:// doi. org/ 10. 1080/ 01419 870. 2017.
13002 97 (2017).
29 . Scheve, K. & Slaughter, M. Labor-market competition and individual preferences over immigration policy. Rev. Econ. Stat. 83(1),
133–145. https:// doi. org/ 10. 1162/ 00346 53017 50160 108 (2001).
30. Mayda, A. Who is against immigration? A cross-country investigation of individual attitudes toward immigrants. Rev. Econ.
Stat. 88(3), 510–530. https:// doi. org/ 10. 1162/ rest. 88.3. 510 (2006).
31. Dustmann, C. & Preston, I. Attitudes to ethnic minorities, ethnic context and location decisions. Econ. J. 111(470), 353–373.
https:// doi. org/ 10. 1111/ 1468- 0297. 00611 (2001).
32. Dustmann, C. & Preston, I. P. Racial and economic factors in attitudes to immigration. B.E. J. Econ. Anal. Policy 7(1), 1–41
(2007).
3 3. Berg, J. A. Explaining attitudes toward immigrants and immigration policy: A review of the theoretical literature. Sociol. Compass
9(1), 23–34. https:// doi. org/ 10. 1111/ soc4. 12235 (2015).
34. Hellwig, T. & Sinno, A. Dierent groups, dierent threats: Public attitudes towards immigrants‡. J. Ethnic Migr. Stud. 43(3),
339–358. https:// doi. org/ 10. 1080/ 13691 83X. 2016. 12027 49 (2017).
35. Danish, R. Q., Shahid, R. & Ali, H. F. Factors aecting life satisfaction of employees under nancial threat. SEISENSE J. Manag.
2(1), 85–98. https:// doi. org/ 10. 33215/ sjom. v2i1. 82 (2019).
36. Bazán-Monasterio, V., Gil-Lacruz, A. I. & Gil-Lacruz, M. Life satisfaction in relation to attitudes towards immigrants among
Europeans by generational cohorts. Int. J. Intercult. Relat. 80, 121–133. https:// doi. org/ 10. 1016/j. ijint rel. 2020. 10. 005 (2021).
37. United Nations Development Program. Human Development Report (Oxford University Press, 1994).
38. Boehnke, K. & Schwartz, S. H. Fear of war: Relations to values, gender, and mental health in Germany and Israel. Peace Con.
J. Peace Psychol. 3(2), 149–165. https:// doi. org/ 10. 1207/ s1532 7949p ac0302_3 (1997).
39. Calogero, R. M., Tylka, T. L., Siegel, J. A., Pina, A. & Roberts, T.-A. Smile pretty and watch your back: Personal safety anxiety
and vigilance in objectication theory. J. Pers. Soc. Psychol. 121(6), 1195–1222. https:// doi. org/ 10. 1037/ pspi0 000344 (2021).
4 0. Wilk, L. & Fibinger, B. Social Fear of crime and its consequences. ASEJ Sci. J. Bielsko‑Biala Sch. Finance Law 24(1), 54–58. https://
doi. org/ 10. 5604/ 01. 3001. 0014. 1353 (2020).
4 1. Ambrey, C.L., Fleming, C.M. & Manning M. Greenspace and life satisfaction: e moderating role of fear of crime in the neighbour‑
hood. In: Opportunities for the critical decade: Enhancing wellbeing within planetary boundaries. Presented at the Australia New
Zealand Society for Ecological Economics 2013 Conference, e University of Canberra and Australia New Zealand Society for
Ecological Economics, Canberra, Australia (2014).
42. Adams, R. E. & Serpe, R. T. Social integration, fear of crime, and life satisfaction. Sociol. Perspect. 43(4), 605–629. https:// doi.
org/ 10. 2307/ 13895 50 (2000).
43 . Grozdanovska, E. e relationship between national identity, subjective well-being and meaning in life. Suvremena Psihol. 19(1),
91–99. https:// doi. org/ 10. 21465/ 2016- sp- 191- 08 (2016).
44. Smith, A. D. National identity and the idea of European unity. Int. A. 68(1), 55–76. https:// doi. org/ 10. 2307/ 26204 61 (1992).
45. Huntington, S. P. e clash of civilizations?. Foreign A. 72(3), 22. https:// doi. org/ 10. 2307/ 20045 621 (1993).
46. Khan, S. S., Garnett, N., Khazaie, D. H., Liu, J. H. & De Zúñiga, H. G. Opium of the people? National identication predicts
well-being over time. Br. J. Psychol. 111(2), 200–214. https:// doi. org/ 10. 1111/ bjop. 12398 (2019).
47. Ellena, A. M., Aresi, G., Marta, E. & Pozzi, M. Post-traumatic growth dimensions dierently mediate the relationship between
national identity and interpersonal trust among young adults: A study on COVID-19 crisis in Italy. Front. Psychol. https:// doi.
org/ 10. 3389/ fpsyg. 2020. 576610 (2021).
4 8. Jaspal, R., Da Silva Lopes, B. C. & Breakwell, G. M. British national identity and life satisfaction in ethnic minorities in the United
Kingdom. Natl. Identities 23(5), 455–472. https:// doi. org/ 10. 1080/ 14608 944. 2020. 18227 93 (2020).
49. Caseres, G. A. (2024). How helping my nation protects my well-being: National identity as a predictor of well-being and the
mediating role of volunteering behavior among the youth in the post-pandemic. MA esis, De La Salle University. https://
animo repos itory. dlsu. edu. ph/ etdm_ psych/ 71
Content courtesy of Springer Nature, terms of use apply. Rights reserved
15
Vol.:(0123456789)
Scientic Reports | (2024) 14:19677 | https://doi.org/10.1038/s41598-024-70067-5
www.nature.com/scientificreports/
50. Jordanov, V., Buzea, C., Ljujic, V. & Dimitrova, R. e inuence of nationalism and national identity on well-being of Bulgarian
and Romanian youth. Studia Univ. Babes‑Bolyai Sociol. 1, 69–86 (2013).
51. Amin, A., McCashin, D., Abdelrahman, M., Al-Adwan, D., & Hasan, H. e psychological eects of perceived threat, national
identity and self-esteem on the well-being of Qatari youth during the blockade. Research Square, Version 1. https:// doi. org/ 10.
21203/ rs.3. rs- 305119/ v1 (2020).
52. Bucher, A., Neubauer, A. B., Voss, A. & Oetzbach, C. Together is better: Higher committed relationships increase life satisfaction
and reduce loneliness. J. Happiness Stud. 20(8), 2445–2469. https:// doi. org/ 10. 1007/ s10902- 018- 0057-1 (2019).
53. Fry, R., & Parker, K. Rising share of U.S. adults are living without a spouse or partner. Pew Research Center. https:// www. pewre
search. org/ social- trends/ 2021/ 10/ 05/ rising- share- of-u- s- adults- are- living- witho ut-a- spouse- or- partn er/ (2021).
54. Brown, A. A prole of single Americans. Pew Research Center. https:// www. pewre search. org/ social- trends/ 2020/ 08/ 20/a- pro
le- of- single- ameri cans/ (2020).
55. Adamczyk, K. & Segrin, C. Direct and indirect eects of young adults’ relationship status on life satisfaction through loneliness
and perceived social support. Psychol. Belgica 55(4), 196–211. https:// doi. org/ 10. 5334/ pb. bn (2015).
56. Estlein, R. & Lavee, Y. Eect of daily stress on desire for physical proximity and emotional closeness. J. Fam. Issues 43(4),
1039–1067. https:// doi. org/ 10. 1177/ 01925 13x21 10075 28 (2021).
57 . Arikewuyo, A. O., Lasisi, T. T., Abdulbaqi, S. S., Omoloso, A. & Arikewuyo, H. O. Evaluating the use of social media in escalating
conicts in romantic relationships. J. Public A. https:// doi. org/ 10. 1002/ pa. 2331 (2020).
58 . Estlein, R., Gewirtz-Meydan, A. & Opuda, E. Love in the time of COVID-19: A systematic mapping review of empirical research
on romantic relationships one year into the COVID-19 pandemic. Fam. Process 61(3), 1208–1228. https:// doi . o r g/ 10. 1111/ famp.
12775 (2022).
59. Deaton, A. What do self-reports of wellbeing say about life-cycle theory and policy?. J. Public Econ. 162, 18–25. https:// doi. org/
10. 1016/j. jpube co. 2018. 02. 014 (2018).
60. Gitmez, A. I. & Morçöl, G. Socio-economic status and life satisfaction in Turkey. Soc. Indic. Res. 31(1), 77–98. https:// doi. org/
10. 1007/ bf010 86515 (1994).
61. Daraei, M. & Mohajery, A. e impact of socioeconomic status on life satisfaction. Soc. Indic. Res. 112(1), 69–81. https:// doi.
org/ 10. 1007/ s11205- 012- 0040-x (2013).
62. Zaborskis, A. et al. Social inequality in adolescent life satisfaction: Comparison of measure approaches and correlation with
macro-level indices in 41 countries. Soc. Indic. Res. 141(3), 1055–1079. https:// doi. org/ 10. 1007/ s11205- 018- 1860-0 (2019).
63. Adler, N. E., Epel, E. S., Castellazzo, G. & Ickovics, J. R. Relationship of subjective and objective social status with psychological
and physiological functioning: Preliminary data in healthy, white women. Health Psychol. 19(6), 586–592. https:// doi. org/ 10.
1037/ 0278- 6133. 19.6. 586 (2000).
64. Alexsyed. Social class pyramid. Slideshare. https:// www. slide share. net/ alexs yed/ social- class- pyram id (2011).
65. de Looze, M. E., Huijts, T., Stevens, G. W., Torsheim, T. & Vollebergh, W. A. M. e happiest kids on earth. Gender equality and
adolescent life satisfaction in Europe and North America. J. Youth Adolesc. 47(5), 1073–1085. https:// doi. org/ 10. 1007/ s10964-
017- 0756-7 (2018).
6 6. Abbott, P., Wallace, C., Lin, K. & Haerpfer, C. e quality of society and life satisfaction in China. Soc. Indic. Res. 127(2), 653–670.
https:// doi. org/ 10. 1007/ s11205- 015- 0989-3 (2016).
67. Haerpfer, C. W., Wallace, C. & Abbott, P. A sociological explanation for the health consequences of the transition from com-
munism in the former Soviet Union. Perspect. Eur. Polit. Soc. 14(4), 460–479 (2013).
68. Szadvári, I., Ostatníková, O. & Durdiaková, J. B. Sex dierences matter: Males and females are equal but not the same. Physiol.
Behav. 259, 114038. https:// doi. org/ 10. 1016/j. physb eh. 2022. 114038 (2023).
69. Brand, J. A. et al. Sex dierences in the predictability of risk-taking behavior. Behav. Ecol. 34(1), 108–116. https:// doi. org/ 10.
1093/ beheco/ arac1 05 (2023).
70. Hyde, J. S. e gender similarities hypothesis. Am. Psychol. 60(6), 581–592. https:// doi. org/ 10. 1037/ 0003- 066x. 60.6. 581 (2005).
7 1. Diener, E., Suh, E. M., Lucas, R. E. & Smith, H. L. Subjective well-being: ree decades of progress. Psychol. Bull. 125(2), 276–302.
https:// doi. org/ 10. 1037/ 0033- 2909. 125.2. 276 (1999).
72. Liang, J. Sex dierences in life satisfaction among the elderly. J. Gerontol. 37(1), 100–108. https:// doi. org/ 10. 1093/ geronj/ 37.1.
100 (1982).
73. Okulicz-Kozaryn, A. Religiosity and life satisfaction across nations. Ment. Health Relig. Cult. 13(2), 155–169. https:// doi. org/
10. 1080/ 13674 67090 32738 01 (2009).
74. Koenig, H. G., McCullough, M. E. & Larson, D. B. Handbook of Religion and Health (Oxford University Press, 2001).
75. Lim, C. & Putnam, R. D. Religion, social networks, and life satisfaction. Am. Sociol. Rev. 75(6), 914–933. https:// doi. org/ 10.
1177/ 00031 22410 386686 (2010).
76. Hayward, R. D., Krause, N., Ironson, G., Hill, P. S. & Emmons, R. A. Health and well-being among the non-religious: Atheists,
agnostics, and no preference compared with religious group members. J. Relig. Health 55(3), 1024–1037. https:// doi. org/ 10.
1007/ s10943- 015- 0179-2 (2016).
77. Pöhls, K., Schlösser, T. & Fetchenhauer, D. Non-religious identities and life satisfaction: Questioning the universality of a linear
link between religiosity and Well-Being. J. Happiness Stud. 21(7), 2327–2353. https:// doi. org/ 10. 1007/ s10902- 019- 00175-x (2020).
78. Uğur, Z. B. & Aydın, F. Are religious people happy or non-religious people unhappy in religious contexts?. Soc. Psychol. Pers.
Sci. 14(2), 156–172. https:// doi. org/ 10. 1177/ 19485 50622 10823 34 (2023).
79. Chapman, A., Fujii, H. & Managi, S. Multinational life satisfaction, perceived inequality and energy aordability. Nat. Sustain.
2(6), 508–514. https:// doi. org/ 10. 1038/ s41893- 019- 0303-5 (2019).
80. Gaymu, J. & Springer, S. Living conditions and life satisfaction of older Europeans living alone: A gender and cross-country
analysis. Ageing Soc. 30(7), 1153–1175. https:// doi. org/ 10. 1017/ s0144 686x1 00002 31 (2010).
81. Bonini, A. Cross-National variation in individual life satisfaction: Eects of national wealth, human development, and environ-
mental conditions. Soc. Indic. Res. 87(2), 223–236. https:// doi. org/ 10. 1007/ s11205- 007- 9167-6 (2008).
82. Morrison, M., Tay, L. & Diener, E. Subjective well-being and national satisfaction. Psychol. Sci. 22(2), 166–171. https:// doi. org/
10. 1177/ 09567 97610 396224 (2011).
83. Carr, E. & Chung, H. Employment insecurity and life satisfaction: e moderating inuence of labour market policies across
Europe. J. Eur. Soc. Policy 24(4), 383–399. https:// doi. org/ 10. 1177/ 09589 28714 538219 (2014).
84. Argan, M., Argan, M. T. & Dursun, M. T. Examining relationships among well-being, leisure satisfaction, life satisfaction, and
happiness. Int. J. Med. Res. Health Sci. 7(4), 2319–5886 (2018).
8 5. Delhey, J. From materialist to post-materialist happiness? National auence and determinants of life satisfaction in cross-national
perspective. Soc. Indic. Res. 97(1), 65–84. https:// doi. org/ 10. 1007/ s11205- 009- 9558-y (2010).
86. Trading Economics. (2021, December). GDP per capita: Countries list world. From https:// tradi ngeco nomics. com/ count ry- list/
gdp- per- capita? conti nent= world (Acessed 23 May 2023).
87. World Economic Outlook. (2023, April). GDP per capita: Current prices. International Monetary Fund. From https:// www. imf.
org/ exter nal/ datam apper/ NGDPD PC@ WEO/ OEMDC/ ADVEC/ WEOWO RLD (Accessed 23 May 2023).
88. Degutis, M., Urbonavičius, S. & Gaižutis, A. Relation between life satisfaction and GDP in the European union. Ekonomika
89(1), 9–21. https:// doi. org/ 10. 15388/ ekon. 2010.0. 997 (2010).
Content courtesy of Springer Nature, terms of use apply. Rights reserved
16
Vol:.(1234567890)
Scientic Reports | (2024) 14:19677 | https://doi.org/10.1038/s41598-024-70067-5
www.nature.com/scientificreports/
89. Proto, E. & Rustichini, A. A reassessment of the relationship between GDP and life satisfaction. PLoS One 8(11), e79358. https://
doi. org/ 10. 1371/ journ al. pone. 00793 58 (2013).
90. Van Bavel, J. et al. National identity predicts public health support during a global pandemic. Nat. Commun. https:// doi. org/ 10.
1038/ s41467- 021- 27668-9 (2022).
91. Kontis, V. et al. Magnitude, demographics and dynamics of the eect of the rst wave of the COVID-19 pandemic on all-cause
mortality in 21 industrialized countries. Nat. Med. https:// doi. org/ 10. 1038/ s41591- 020- 1112-0 (2020).
92. Hair, J. F., Risher, J. J., Sarstedt, M. & Ringle, C. M. When to use and how to report the results of PLS-SEM. Eur. Bus. Rev. 31(1),
2–24 (2019).
93. Hair Jr., J. F., Hult, G. T. M., Ringle, C. M., Sarstedt, M., Danks, N. P., & Ray, S. Partial least squares structural equation modeling
(PLS-SEM) using R: A workbook. https:// libra ry. oapen. org/ handle/ 20. 500. 12657/ 51463 (2021).
94. Verkuyten, M. Public attitudes towards migrants: Understanding cross-national and individual dierences. World Psychiatry
O. J. World Psychiatr. Assoc. (WPA) 20(1), 132–133. https:// doi. org/ 10. 1002/ wps. 20819 (2021).
95. Wimmer, A. Nation building: Why some countries come together while others fall apart. Survival 60(4), 151–164. https:// doi.
org/ 10. 1080/ 00396 338. 2018. 14954 42 (2018).
96. Berkowitz, L. & Sherif, M. In common predicament: Social psychology of intergroup conict and cooperation. Am. Sociol. Rev.
32(2), 333. https:// doi. org/ 10. 2307/ 20918 43 (1967).
97. Böhm, R., Rusch, H. & Baron, J. e psychology of intergroup conict: A review of theories and measures. J. Econ. Behav. Org.
178, 947–962. https:// doi. org/ 10. 1016/j. jebo. 2018. 01. 020 (2020).
98. Pinillos-Franco, S. & Kawachi, I. Hostile attitudes toward immigrants and refugees are associated with poor self-rated health.
Analysis of 21 European countries. Soc. Sci. Med. 301, 114969. https:// doi. org/ 10. 1016/j. socsc imed. 2022. 114969 (2022).
99. Achim, M. V., Văidean, V. L. & Borlea, S. N. Corruption and health outcomes within an economic and cultural framework. Eur.
J. Health Econ. 21(2), 195–207. https:// doi. org/ 10. 1007/ s10198- 019- 01120-8 (2019).
100. Sortheix, F. M. & Lönnqvist, J.-E. Personal value priorities and life satisfaction in Europe: e moderating role of socioeconomic
development. J. Cross‑Cult. Psychol. 45(2), 282–299. https:// doi. org/ 10. 1177/ 00220 22113 504621 (2014).
101. Sortheix, F. M. & Schwartz, S. H. Values that underlie and undermine well–being: Variability across countries. Eur. J. Pers. 31(2),
187–201. https:// doi. org/ 10. 1002/ per. 2096 (2017).
102. Ribeiro, D., Costa, A. P. & Remondes, J. Government communication—e Dubai and United Arab Emirates Ministry of Hap-
piness. Adv. Intell. Syst. Comput. https:// doi. org/ 10. 1007/ 978-3- 030- 31787-4_ 19 (2019).
103. Musikanski, L. & Polley, C. Life, liberty, and the pursuit of happiness: Measuring what matters. J. Soc. Change https:// doi. org/
10. 5590/ josc. 2016. 08.1. 05 (2016).
Acknowledgements
e authors are grateful for the comments and suggestions by the referees and the handling Editor. eir com-
ments and suggestions have greatly improved the paper. e authors also gratefully acknowledge that the work
in this paper was supported, in part, by the Open Access Program from the American University of Sharjah.
Author contributions
All authors have contributed to all sections including the methodology, the data analysis, and the conclusion.
All authors read and approved the nal manuscript.
Competing interests
e authors declare no competing interests.
Additional information
Supplementary Information e online version contains supplementary material available at https:// doi. org/
10. 1038/ s41598- 024- 70067-5.
Correspondence and requests for materials should be addressed to A.A.
Reprints and permissions information is available at www.nature.com/reprints.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and
institutional aliations.
Open Access is article is licensed under a Creative Commons Attribution 4.0 International License, which
permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give
appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and
indicate if changes were made. e images or other third party material in this article are included in the article’s
Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included
in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or
exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy
of this licence, visit http://creativecommons.org/licenses/by/4.0/.
© e Author(s) 2024
Content courtesy of Springer Nature, terms of use apply. Rights reserved
1.
2.
3.
4.
5.
6.
Terms and Conditions
Springer Nature journal content, brought to you courtesy of Springer Nature Customer Service Center GmbH (“Springer Nature”).
Springer Nature supports a reasonable amount of sharing of research papers by authors, subscribers and authorised users (“Users”), for small-
scale personal, non-commercial use provided that all copyright, trade and service marks and other proprietary notices are maintained. By
accessing, sharing, receiving or otherwise using the Springer Nature journal content you agree to these terms of use (“Terms”). For these
purposes, Springer Nature considers academic use (by researchers and students) to be non-commercial.
These Terms are supplementary and will apply in addition to any applicable website terms and conditions, a relevant site licence or a personal
subscription. These Terms will prevail over any conflict or ambiguity with regards to the relevant terms, a site licence or a personal subscription
(to the extent of the conflict or ambiguity only). For Creative Commons-licensed articles, the terms of the Creative Commons license used will
apply.
We collect and use personal data to provide access to the Springer Nature journal content. We may also use these personal data internally within
ResearchGate and Springer Nature and as agreed share it, in an anonymised way, for purposes of tracking, analysis and reporting. We will not
otherwise disclose your personal data outside the ResearchGate or the Springer Nature group of companies unless we have your permission as
detailed in the Privacy Policy.
While Users may use the Springer Nature journal content for small scale, personal non-commercial use, it is important to note that Users may
not:
use such content for the purpose of providing other users with access on a regular or large scale basis or as a means to circumvent access
control;
use such content where to do so would be considered a criminal or statutory offence in any jurisdiction, or gives rise to civil liability, or is
otherwise unlawful;
falsely or misleadingly imply or suggest endorsement, approval , sponsorship, or association unless explicitly agreed to by Springer Nature in
writing;
use bots or other automated methods to access the content or redirect messages
override any security feature or exclusionary protocol; or
share the content in order to create substitute for Springer Nature products or services or a systematic database of Springer Nature journal
content.
In line with the restriction against commercial use, Springer Nature does not permit the creation of a product or service that creates revenue,
royalties, rent or income from our content or its inclusion as part of a paid for service or for other commercial gain. Springer Nature journal
content cannot be used for inter-library loans and librarians may not upload Springer Nature journal content on a large scale into their, or any
other, institutional repository.
These terms of use are reviewed regularly and may be amended at any time. Springer Nature is not obligated to publish any information or
content on this website and may remove it or features or functionality at our sole discretion, at any time with or without notice. Springer Nature
may revoke this licence to you at any time and remove access to any copies of the Springer Nature journal content which have been saved.
To the fullest extent permitted by law, Springer Nature makes no warranties, representations or guarantees to Users, either express or implied
with respect to the Springer nature journal content and all parties disclaim and waive any implied warranties or warranties imposed by law,
including merchantability or fitness for any particular purpose.
Please note that these rights do not automatically extend to content, data or other material published by Springer Nature that may be licensed
from third parties.
If you would like to use or distribute our Springer Nature journal content to a wider audience or on a regular basis or in any other manner not
expressly permitted by these Terms, please contact Springer Nature at
onlineservice@springernature.com