Original Research

SAGE Open
July-September 2024: 1–14
Ó The Author(s) 2024
Examining the Influence of DOI: 10.1177/21582440241267373
journals.sagepub.com/home/sgo
Generalized Trust on Life Satisfaction
Across Different Education Levels and
Socioeconomic Conditions Using the
Bayesian Mindsponge Framework

Tam-Tri Le1,2 , Minh-Hoang Nguyen1,2 , Ruining Jin3 , Viet-Phuong La1,2,

Hong-Son Nguyen4, and Quan-Hoang Vuong1

Abstract
Extant literature suggests a positive correlation between social trust (also called generalized trust) and life satisfaction.
However, the psychological pathways underlying this relationship can be complex. Using the Bayesian Mindsponge Framework
(BMF), we examined the influence of social trust in a high-violence environment. Employing Bayesian analysis on a sample of
1,237 adults in Cali, Colombia, we found that in a linear relationship, generalized trust is positively associated with life satisfac-
tion. However, in a model including the interactions between trust and education level as well as between trust and socioeco-
nomic status, generalized trust is found to be negatively associated with life satisfaction. In this non-linear relationship, both
education level and socioeconomic status have moderating effects against the negative association between generalized trust
and life satisfaction. In other words, less educated people living in worse socioeconomic conditions are more likely to have
lower life satisfaction when they have higher levels of social trust. In contrast, highly educated people living in better socioeco-
nomic conditions are more likely to have higher life satisfaction when they have higher levels of social trust. Due to the facili-
tating function of trust in information processing, lowering the rigor of the filtering system in a high-violence social
environment will likely put an individual at risk. Based on our findings, we suggest that policymakers should be more meticu-
lous and consider many socioeconomic factors when advocating for increasing social trust. We also recommend that research-
ers should investigate deeper the complexity of human psychology and the information-processing mechanisms of social trust.

Plain language summary

Employing Bayesian analysis aided by the MCMC technique on 1,237 urban people in Colombia, we found that
generalized trust is positively associated with life satisfaction. However, people with lower educational levels whose
socioeconomic status are lower are more likely to have lower life satisfaction when they have higher levels of social
trust, whereas people with higher educational levels whose socioeconomic status is higher will be more likely to have
higher life satisfaction when they have higher levels of social trust. Our findings also suggest that education level has a

1
Phenikaa University, Ha Dong District, Hanoi, Vietnam
2
A.I. for Social Data Lab (AISDL), Vuong & Associates, Hanoi, Vietnam
3
China University of Political Science and Law, Beijing, China
4
Office of CPV Central Committee, Ba Dinh District, Hanoi, Vietnam

Corresponding Author:
Minh-Hoang Nguyen, Centre for Interdisciplinary Social Research, Phenikaa University, Yen Nghia Ward, Ha Dong District, Hanoi 100803, Vietnam.
Email: [email protected]

Data Availability Statement included at the end of the article

Creative Commons CC BY: This article is distributed under the terms of the Creative Commons Attribution 4.0 License
(https://creativecommons.org/licenses/by/4.0/) which permits any use, reproduction and distribution of
the work without further permission provided the original work is attributed as specified on the SAGE and Open Access pages
(https://us.sagepub.com/en-us/nam/open-access-at-sage).
2 SAGE Open

bigger impact compared to socioeconomic status in moderating the relationship between social trust and life
satisfaction. Based on our findings, promoting social trust in communities with lower education levels, poor living
conditions, and a violent social context should be carefully planned due to the complexity of related information
processes’ interactions.

Keywords
social trust, life satisfaction, education level, socioeconomic status, Bayesian Mindsponge Framework

Introduction when life satisfaction and social trust in society are both
low, individuals are more likely to suffer from mental ill-
Social Trust and Life Satisfaction ness due to health distress (L. M. Martı́nez et al., 2019).
Trust refers to the trustor’s willingness to accept vulner- Therefore, researchers often suggest policymakers aim
abilities under the assumption that the trustee will act in for increasing social trust.
the trustor’s best interests (Mayer et al., 1995; Schilke However, the relationship between social trust and life
et al., 2021). In other words, trust is often considered a satisfaction is not simple or straightforward. Daukas
second-order emotion that is based on the emotional (2006) asserted that trust towards new knowledge and
state of someone else (Belli & Broncano, 2017). In social information is a complex concept that is shaped by social
science studies, trust is a critical component that would location (including race, gender, religious beliefs, etc.).
facilitate a range of dynamic social processes, including Therefore, in various social contexts, people in diverse
organizational effectiveness, economic development, and social locations would view trust and life satisfaction dif-
so on (Roth, 2022; Rousseau et al., 1998; Tabellini, ferently. For example, Adjaye-Gbewonyo et al. (2018)
2010). Some believe that trust is one of the critical com- found that high individual-level generalized trust was
ponents in corruption reduction and state-building unexpectedly associated with increased depressive symp-
(Rose-Ackerman, 2001). Others suggest that trust is one toms in South Africa. The authors of that study argued
of the most essential components of modern civilizations that in a low-trust society, blindly trusting others may be
and that social conflict and chaos are too commonplace detrimental due to risks of crime or fraud. Hamamura
in the absence of it (Gambetta, 2011; Harraka, 2002; et al. (2017) found that generalized trust was more
Muraskin, 1974; Vuong, 2022). strongly associated with happiness in developed societies
Social trust (also called generalized trust) is trust than in developing societies. Bi et al. (2021)’s study also
toward other people in society in general. It is an impor- found inconsistency in the relationship between social
tant form of trust in modern society that involves a large trust and life satisfaction. Specifically, in probing the
number of social interactions among strangers (Algan & moderating role of social trust on the life satisfaction of
Cahuc, 2013; Intravia et al., 2016). Social trust is the adolescents, Bi et al. (2021) argued that in countries with
bedrock of interactions between individuals in a society, higher levels of social trust, adolescents were reported to
and without trust in the acts of others, individuals may have lower levels of life satisfaction because with a
need to contemplate too many contingencies and uncer- higher level of trust would anticipate more social cooper-
tainties before acting, which would ruin the foundation ation and mutual respect from others, particularly those
of community and civilization (Kwon, 2019). As a result, who are not compelled to assist them. Consequently,
the level of social trust would have an impact on individ- their life satisfaction would be diminished if their expec-
uals’ life satisfaction. tations were not satisfied. To investigate deeper into the
Defined as one’s self-evaluation of one’s own quality psychological pathways in the relationship between
of life based on one’s own rules (Shin & Johnson, 1978), social trust and life satisfaction, the approach of infor-
life satisfaction refers to the cognitive component of sub- mation processing can be advantageous.
jective well-being (Pavot & Diener, 1993). According to The mindsponge mechanism of the human mind’s
Pavot and Diener (1993), life satisfaction is an important information processing was originally conceptualized by
indicator that represents an individual’s life situation as Vuong and Napier (2015) to describe how a person
well as their mental state. Many studies have concluded absorbs and incorporates new values into their mindset.
that increased social trust would lead to better life satis- Throughout the mindsponge process, trust plays a cru-
faction (Graafland & Lous, 2019; Shao et al., 2021; cial role as a facilitator for filtering newly received infor-
Vyrost et al., 2007). Social trust was also found to be mation (Nguyen et al., 2021; Vuong, Le, La, & Nguyen,
positively associated with individual happiness 2022). Trust can influence the cost-benefit judgment by
(Hamamura et al., 2017; Kuroki, 2011). In addition, adding preconceived positive or negative values, which
Le et al. 3

Table 1. Variable Description.

Variable Description Type of variable Value

Satisfaction The degree of respondents’ life satisfaction in general Numeric Ranging from 0 to 10
GenTrust The degree of respondents’ generalized trust Numeric Ranging from 0 to 10
Edulevel The level of respondents’ education Numeric Ranging from 1 to 4
Socioeconomic The respondents’ socioeconomic status Numeric Ranging from 1 to 6

helps accelerate the evaluation process. Normally, trust Columbia. Moreover, it is imperative to probe the infor-
(or distrust) is attached to a source of information or a mation processing mechanism in this relationship, as to
collection of information with similar features (Vuong, how trust/distrust of strangers would affect one’s overall
Le, La, Nguyen, et al., 2022). Thus, the mind can save life satisfaction at a personal-societal level under specific
time and energy by quickly accepting or rejecting infor- conditions. Possible significant moderating factors in this
mation from the same source or group without going process include education level and socioeconomic sta-
through a thorough evaluation process (Le et al., 2022; tus, which determine the immediate infosphere that an
Vuong, 2022). Due to the function of trust in terms of individual interacts with (e.g., living environment and
information processing, the psychological pathways to social circles). Thus, using the Bayesian Mindsponge
life satisfaction may be different in a high-violence socio- Framework (Nguyen et al., 2022b; Vuong, La, &
political context. Nguyen, 2022)—an effective tool for sociopsychological
research—as the basis for theoretical conceptualization
and statistical analysis, our study aims to examine the
The High-Violence Social Infosphere in Columbia following research questions (RQ):
The phrase ‘‘high-violence context’’ in this study refers to
a living condition with aggressions that may cause physi- RQ1: What is the relationship between social trust
cal harm or psychological trauma, such as assault, mur- and life satisfaction in a high-violence context such as
der, drug trading, fraud, etc. (L. M. Martı́nez et al., the case of Columbia?
2019). According to United Nations Office on Drug and RQ2: Do education level and socioeconomic status
Crime (UNODC), Colombia is one of the countries that play a role in the association between social trust and
has the highest homicide rates in the world (United life satisfaction?
Nations Office on Drugs and Crime [UNODC], 2006).
Although the number of intentional homicide cases
dropped in the past 30 years from 73 per 100,000 inhabi- Materials and Methods
tants in 1990 to 23 per 100,000 inhabitants in 2020, the
ratio is still striking when compared to that of the aver-
Materials, Variables, and Models
age countries (6 per 100,000) (World Bank, 2022). In This study employs secondary data, with the dataset
addition, crimes other than homicide also persisted in from the article ‘‘Trust, life satisfaction and health:
Columbia due to the existence of illegal armed organiza- Population data in a mid-size city in the Global South’’
tions and criminal groups’ involvement in drug trading (L. Martı́nez, 2019). Data collection was conducted in
and serious crimes, including kidnapping, money laun- 2017 by the Observatory of Public Policy of ICESI
dering, and running extortion and prostitution rackets. University using face-to-face interviews. Respondents
Consequently, Verisk Maplecroft, a risk analysis firm were explained about the purpose of the study and
based in the United Kingdom, in its global risk report, assured confidentiality. There was a total of 1,237
listed Columbia’s Medellı́n as a nucleus of transnational responses from adults in the city of Cali. The male/
crime networks and tagged Colombia’s capital Bogotá as female ratio is approximately 1:1. The average age of the
one of the top 3 riskiest among the world’s 30 largest cit- respondents is 39.
ies (Parkes & Blanco, 2022). In the context of rampant The outcome variable in the present study is life satis-
violent crimes, exposing one’s vulnerabilities to others faction, and the independent variables are generalized
poses relatively higher risks of physical, mental, or finan- trust, education level, and socioeconomic status (see
cial harm. descriptions in Table 1). Life satisfaction is measured
While many studies have been conducted about social with the following question: ‘‘On a scale of zero to ten
trust and life satisfaction in developed countries, few (zero means you have no satisfaction and ten means you
have been conducted regarding social trust in a highly have complete satisfaction), in general, how satisfied are
violent sociopolitical context, namely, in this case, you with your life?’’ The average score of life satisfaction
4 SAGE Open

for this sample is 8.4. The measurement of trust followed credible intervals. The current reproducibility crisis in
the OECD guidelines, which measures generalized trust social sciences, especially psychology, is partly attributed
using the following question: ‘‘On a scale from zero to to the over-reliance on p-value as a dichotomous thresh-
ten, where zero is ‘not at all,’ and ten is ‘completely,’ in old for rejecting null hypotheses (Camerer et al., 2018;
general, how much do you trust most people?’’ (OECD, Open Science Collaboration, 2015). Visualizing estimated
2017). The average score of generalized trust for the sam- coefficients in the Bayesian approach is a reliable alterna-
ple is 4.5. For socioeconomic status, the stratification tive for evaluating statistical results instead of p-values
system in Colombia classifies households into six cate- (Halsey et al., 2015).
gories numbered 1 to 6, where 1 means worst conditions Both models were fitted using the following MCMC
and 6 means best conditions. For education level, in this setups: 5,000 iterations, 2,000 warm-up iterations, and 4
study, we use a four-level classification, where 1 is unedu- Markov chains. Due to the exploratory nature of the
cated or primary education, 2 is secondary education, 3 study, we used uninformative priors to provide the least
is undergraduate degrees or equivalent, and 4 is Master amount of prior information possible to the model esti-
or Doctorate. The average education length of respon- mation (Diaconis & Ylvisaker, 1985). To test the models’
dents in the sample is 11.8 years. goodness-of-fit, Pareto smoothed importance-sampling
Following the research questions stated, two models leave-one-out cross-validation (PSIS-LOO) test was
are constructed. In order to increase predictability, we employed (Vehtari et al., 2017). To validate the simulated
adhere to the principle of parsimonious model construc- posteriors, we used diagnostic statistics of effective sample
tion for Bayesian analytics (Nguyen et al., 2022a). size (n_eff) and Gelman-Rubin shrink factor (Rhat) to
Parsimonious models have high predictive power, and check Markov’s chains’ convergence. The model’s
they allow us to focus on examining specific factors and Markov chains are generally considered to be convergent
relationships. While such parsimonious models mean if n_eff is greater than 1,000 and Rhat is equal to 1.
there are many unknown parameters, the properties of Graphical representations such as trace plots, Gelman-
Bayesian inference and validating techniques in analysis Rubin-Brooks plots, and autocorrelation plots were also
help increase the prediction’s accuracy (see the subsec- used for convergence diagnosis. Due to the following
tion below for more detail). Model 1 examines the linear advantages, the present study’s Bayesian analysis was car-
relationship between Satisfaction and GenTrust. In ried out using the bayesvl R package (Vuong et al., 2020).
Model 2, the moderation of Edulevel and Socioeconomic The package’s effective visualization features support the
are added through their interactions with GenTrust. presentation and interpretation of results. The package
also supports replication studies or related follow-up
Satisfaction ; GenTrust ð1Þ research because it is free, open, and simple to use.
Considering the importance of transparency in
Satisfaction ; GenTrust + Edulevel  GenTrust research procedures and cost management (Vuong, 2018,
ð2Þ 2020), the study’s data files and code snippets were
+ Socioeconomic  GenTrust
deposited at the Open Science Framework (OSF) server
(DOI: 10.17605/OSF.IO/H8GQ7).
Statistical Analysis
The Bayesian analysis method was used because it is Results
compatible with mindsponge-based reasoning, among
other advantages (Nguyen et al., 2022b). While parsimo- Model 1
nious models have high predictability, there are also PSIS diagnostic for Model 1 shows that the Pareto k-val-
many unknown parameters. The advantage of Bayesian ues are less than 0.5 (see Figure 1), indicating that the
inference for this problem is that it treats all parameters model fits the actual data well. In other words, the model
probabilistically, including unknown ones (Gill, 2015). has an acceptable goodness-of-fit.
Moreover, prediction accuracy in fitting complex models In Table 2, we see that the n_eff values are greater
containing interaction terms is increased thanks to the than 1,000, and Rhat values are equal to 1, generally con-
Markov Chain Monte Carlo (MCMC) algorithms, which sidered to be signals of good convergence of the Markov
iteratively generate a large number of samples from the chains. The trace plots for Model 1 (see Figure 2) show
joint posterior distribution of the parameters (Cowles, fluctuations around a central equilibrium, meaning that
2013; Dunson, 2001; Wagenmakers et al., 2018). Another there are no divergent chains after warmup iterations.
major advantage of using Bayesian analysis in psycholo- This indicates that the Markov chains converge to the
gical research is that it does not make binary judgments same posteriors and the results are reliable.
based on p-values to evaluate statistical results but rather The Gelman plots (see Figure 3) support the healthy
makes interpretations with estimated and visualized convergence by showing that the shrink factors reduce to
Le et al. 5

GenTrust at 95% Highest Posterior Distribution

Intervals (HPDI) lies completely on the positive side
(greater than 0.0195), which indicates that the result has
high reliability (see Figure 5).

Model 2
The relationship tree for the parameters in Model 2 is
shown in Figure 6. GenTrust affects Satisfaction in a lin-
ear relationship and through two interacting pathways
with Edulevel and Socioeconomic.
The k values of the PSIS diagnostic plot for Model 2
are less than 0.5 (see Figure 7), indicating that the model
is well-specified.
Figure 1. Model 1’s PSIS diagnostic plot. Similar to the explanation of statistical validation in
Model 1, Model 2 also shows good convergence of the
Markov chains, as seen through the n_eff and Rhat val-
Table 2. Model 1’s Simulated Posterior Coefficients. ues (see Table 3) as well as the trace plots (Figure 8), the
Gelman plots (Figure 9), and the autocorrelation plots
Parameters Mean SD n_eff Rhat (Figure 10).
When considering the interactions with education level
Constant 7.50 0.18 4,899 1
GenTrust 0.09 0.03 5,005 1
and socioeconomic status in the model, generalized trust is
negatively associated with life satisfaction (mGenTrust = 20.34
and sGenTrust = 0.07). However, both Edulevel*Gentrust and
one after warmup iterations, meaning that there is almost Socioeconomic*GenTrust moderate the relationship between
no difference between variance between chains and var- GenTrust and Satisfaction in the opposite direction
iance within chains. (mEdulevelGenTrust = 0.16, sEdulevelGenTrust = 0.03, mSocioeconomic
The autocorrelation plots (see Figure 4) also indicate GenTrust = 0.02, and sSoioeconomicGenTrust = 0.01). In model 2,
that the Markov property is held by showing autocorre- the coefficient for GenTrust has a negative sign (compared
lations being eliminated rapidly, meaning that MCMC- to Model 1’s coefficient being positive). This is because, in
simulated samples are memoryless during the stochastic Model 2, the influence of GenTrust on Satisfaction is con-
simulation process. sidered together with its interactions with Edulevel and
Analysis results show that generalized trust is posi- Socioeconomic simultaneously (see the Satisfaction value
tively associated with life satisfaction (mGenTrust = 0.09 estimation across different scenarios below). It is also worth
and sGenTrust = 0.03). The posterior distribution of noting that although the coefficients are rather small, they

Figure 2. Model 1’s trace plots.

6 SAGE Open

Figure 3. Model 1’s Gelman plots.

Figure 4. Model 1’s autocorrelation plots.

Figure 5. Model 1’s posterior distribution plots with HPDI at 95%.

Le et al. 7

Table 3. Model 2’s Simulated Posterior Coefficients.

Parameters Mean SD n_eff Rhat

Constant 7.53 0.18 9,176 1

GenTrust 20.34 0.07 7,201 1
Gentrust*Edulevel 0.16 0.03 7,231 1
GenTrust*Socioeconomic 0.02 0.01 9,351 1

Similarly, the estimated Satisfaction values for Edulevel

of 1 (uneducated or primary education) and Edulevel of
2 (secondary education) are visualized in Figures 13 and
14, respectively. The x-axis represents the degree of gen-
eralized trust, the y-axis represents the degree of life
satisfaction, and the line color represents socioeconomic
Figure 6. Relationship tree of Model 2. status. Figure 13 shows a downward trend as the degree
of trust increases, whereas Figure 14 shows an upward
trend instead. Edulevel of 3 and 4 continue further in this
upward tendency.

Discussion
Employing Bayesian analysis aided by the MCMC tech-
nique on 1,237 urban people in Colombia, we found that
generalized trust is positively associated with life satisfac-
tion. This result aligns with the findings in other studies
(Graafland & Lous, 2019; Kuroki, 2011; Shao et al.,
2021; Vyrost et al., 2007). However, in a model including
the interactions between trust and education levels as
well as between trust and socioeconomic status, general-
ized trust is found to be negatively associated with life
satisfaction. In this non-linear relationship, both educa-
Figure 7. Model 2’s PSIS diagnostic plot.
tion levels and socioeconomic status have moderating
effects against the negative association between general-
are significant. Regarding the meaning of life satisfaction ized trust and life satisfaction. In other words, people
for a human, the degrees of changes here are considerable. with lower educational levels whose socioeconomic sta-
The visualization of the posterior distributions are tus are lower are more likely to have lower life satisfac-
presented in the interval plots (see Figure 11) and the tion when they have higher levels of social trust, whereas
two-dimensional density plots (see Figure 12). The distri- people with higher educational levels whose socioeco-
butions of GenTrust lie completely on the negative; the nomic status is higher will be more likely to have higher
distributions of Edulevel*GenTrust lie completely on the life satisfaction when they have higher levels of social
positive; the distributions of Socioeconomic*GenTrust lie trust. Our findings also suggest that education level has a
mostly on the positive. bigger impact compared to socioeconomic status in mod-
To aid in interpreting the results of Model 2, the esti- erating the relationship between social trust and life satis-
mation of life satisfaction values based on the posterior faction. People with lower educational levels (primary
coefficients is presented below. The mean values are used education or uneducated) tend to have lower life satisfac-
because they have the highest probabilities of occurrence. tion as social trust increases, regardless of socioeconomic
For example, the estimated Satisfaction value of a person status (poor people have a stronger negative association).
having a GenTrust value of 5, Edulevel value of 3 (under- For those with secondary education levels or above, life
graduate degrees or equivalent), and Socioeconomic value satisfaction tends to increase as social trust increases,
of 4 (slightly above average) is as follows: regardless of socioeconomic status (wealthy people have
a stronger positive association). These findings suggest
ySatisfaction = 7:53  0:34  5 + 0:16  3  5 + 0:02  4  5 = 8:63 that the relationship between generalized trust and life
8 SAGE Open

Figure 8. Model 2’s trace plots.

Figure 9. Model 2’s Gelman plots.

Le et al. 9

Figure 10. Model 2’s autocorrelation plots.

Figure 11. Posterior distributions of Model 2’s parameters.

satisfaction can have different and complex information- et al., 2022; Vuong, Le, La, & Nguyen, 2022). The more
processing pathways underneath its linear association people trust a certain piece of information (meaning that
From the mindsponge standpoint, trust acts as the its carried value was integrated into the mindset), the
‘‘gatekeeper’’ of an information absorption channel (Le more likely and quickly they will accept related
10 SAGE Open

Figure 12. Model 2’s two-dimensional density plot of Edulevel*GenTrust and Socioeconomic*GenTrust.

7.5 Socioeconomic
level
7 1
Satisfaction

2
6.5
3
6 4
5
5.5
6

5
0 1 2 3 4 5 6 7 8 9 10
Social Trust

Figure 13. Estimated life satisfaction in those with no educated or primary education level.

information carrying similar values. Likewise, when a themselves with ‘‘bad’’ social circles where deception and
source of information (including other people—as in hostility are more prevalent, which will create more risks
interpersonal trust) is trusted, new information coming if they lower the rigor of the information filtering system.
from that source is filtered more quickly and favorably. Intuitively, it is commonly considered an unwise idea to
On the other hand, under the influence of distrust, trust strangers in a ‘‘bad neighborhood.’’ This result
related information carrying similar values will likely be aligns with Adjaye-Gbewonyo et al. (2018) findings,
rejected quickly. In a high-violence social environment which suggest that having a high level of trust in a low-
(treated as an infosphere), a high level of generalized trust society may increase the likelihood of depression
trust means that one is more open and susceptible to due to risks of crime or fraud. This result is in alignment
information coming from other people in society, includ- with the claim made by Daukas (2006) that epistemic
ing strangers. Less educated and poor people surround trust is shaped by social location.
Le et al. 11

8.8
8.6
Socioeconomic
8.4 level
Satisfaction 8.2 1
8 2
7.8 3
7.6 4
7.4 5
7.2 6

7
0 1 2 3 4 5 6 7 8 9 10
Social trust

Figure 14. Estimated life satisfaction in those with secondary education level.

A higher level of generalized trust does not mean naive- advocating social trust enhancement as an attempt to
ness. The impact of the received information is largely enhance life satisfaction. Regarding related research
determined by the processing capacity of a system. This efforts, human psychology is multiplex and often con-
also involves the quality of stored information in memory text-dependent, so researchers should pay more attention
used as references in information filtering. In other words, to context-specified psychological processes during their
knowledge and thinking skills are very important for analysis and reasoning to provide more accurate sugges-
effectively utilizing the function of trust in an information tions for policymaking. Regarding future research direc-
process (to speed up evaluation). Information processing tion, further qualitative studies on the relationship
capacity is enhanced by education (Clouston et al., 2012; between social trust and life satisfaction may help shed
Mather, 2020), and our results show that those with more light on possible psychological pathways in specific
higher education levels likely are not negatively affected social contexts.
by high levels of social trust. In fact, from secondary to
undergraduate and graduate levels, the higher the educa-
tion levels are, the stronger the positive correlation Limitations
between generalized trust and life satisfaction is. Overall, First of all, the interplay between social trust and the
the relationship between social trust and life satisfaction high-violence infosphere in a given social context is
has complex underlying psychological pathways. Social highly dynamic, but the study could not go into detail
trust heavily depends on social contexts (Boyadjieva & and examine the complex real-time interactions due to
Ilieva-Trichkova, 2015; Vuong et al., 2021), and complex data limitations. Therefore, further studies are needed to
information processes can sometimes produce a negative explore such specific information processes more clearly.
correlation in specific cases (Bi et al., 2021). Secondly, the study’s sample size of 1,237 participants in
Based on our findings, there are some implications for Cali may not sufficiently reflect the overall situation in
policymakers and researchers on the issue of social trust. Columbia or other developing countries having similar
While it is beneficial to advocate for increasing social issues. Follow-up studies using Bayesian inference can
trust in communities with high education levels and rela- update the existing beliefs on the factors and interac-
tively good socioeconomic conditions, promoting the tions. Thirdly, culture (as a collective mindset from an
same thing in communities with lower education levels, information processing perspective) may potentially be a
poor living conditions, and a violent social context significant moderator in the relationship between social
should be carefully planned due to the complexity of trust and life satisfaction; thus, future studies may want
related information processes’ interactions. So, policy- to focus on exploring this aspect.
makers in countries and regions plagued by uneven edu-
cation resource distribution, underdeveloped economies,
and violent crimes should take extra steps to monitor Acknowledgments
and assess the social conditions of their regions before None.
12 SAGE Open

Author Contributions Bi, S., Stevens, G. W. J. M., Maes, M., Boer, M., Delaruelle,
Conceptualization: Tam-Tri Le, Minh-Hoang Nguyen; K., Eriksson, C., Brooks, F. M., Tesler, R., van der Schuur,
Methodology: Minh-Hoang Nguyen, Viet-Phuong La, Quan- W. A., & Finkenauer, C. (2021). Perceived social support
Hoang Vuong, Tam-Tri Le; Formal analysis and investigation: from different sources and adolescent life satisfaction across
Tam-Tri Le, Viet-Phuong La; Writing—original draft prepara- 42 countries/regions: The moderating role of national-level
tion: Tam-Tri Le, Ruining Jin, Hong-Son Nguyen; Writing— generalized trust. Journal of Youth and Adolescence, 50(7),
review and editing: Tam-Tri Le, Minh-Hoang Nguyen; 1384–1409. https://doi.org/10.1007/s10964-021-01441-z
Validation: Minh-Hoang Nguyen, Quan-Hoang Vuong; Boyadjieva, P. A., & Ilieva-Trichkova, P. I. (2015). Higher edu-
Resources: Viet-Phuong La, Tam-Tri Le; Supervision: Quan- cation and social trust: A European comparative perspec-
Hoang Vuong. tive. In Wiseman, A. W. & Popov N. (eds.), Comparative
sciences: Interdisciplinary approaches (Vol. 26, pp. 153–187).
Declaration of Conflicting Interests Emerald Group Publishing Limited. https://doi.org/10.1108/
S1479-367920140000026007
The author(s) declared no potential conflicts of interest with Camerer, C. F., Dreber, A., Holzmeister, F., Ho, T.-H., Huber,
respect to the research, authorship, and/or publication of this J., Johannesson, M., Kirchler, M., Nave, G., Nosek, B. A.,
article. Pfeiffer, T., Altmejd, A., Buttrick, N., Chan, T., Chen, Y.,
Forsell, E., Gampa, A., Heikensten, E., Hummer, L., Imai,
Funding T., & . Wu, H. (2018). Evaluating the replicability of social
The author(s) disclosed receipt of the following financial sup- science experiments in nature and science between 2010 and
port for the research, authorship, and/or publication of this 2015. Nature Human Behaviour, 2(9), 637–644. https://doi.
article: This work did not receive any financial support. org/10.1038/s41562-018-0399-z
Clouston, S. A., Kuh, D., Herd, P., Elliott, J., Richards, M., &
Hofer, S. M. (2012). Benefits of educational attainment on
Ethical Approval adult fluid cognition: International evidence from three
The data collection followed the local and international rules birth cohorts. International Journal of Epidemiology, 41(6),
for empirical research and was approved by the Institutional 1729–1736. https://doi.org/10.1093/ije/dys148
Review Board of Universidad Icesi. Cowles, M. K. (2013). Applied Bayesian statistics: With R and
OpenBUGS examples. Springer.
Daukas, N. (2006). Epistemic trust and social location. Epis-
Informed Consent teme, 3(1–2), 109–124.
All the respondents’ consents were obtained before survey Diaconis, P., & Ylvisaker, D. (1985). Quantifying prior opinion.
commencement. In J. M. Bernardo, M. H. DeGroot, D. V. Lindley, & A. F.
M. Smith (Eds.), Bayesian statistics (Vol. 2, pp. 133–156).
North Holland Press.
ORCID iDs Dunson, D. B. (2001). Commentary: Practical advantages of
Tam-Tri Le https://orcid.org/0000-0003-3384-4827 Bayesian analysis of epidemiologic data. American Journal
Minh-Hoang Nguyen https://orcid.org/0000-0002-7520- of Epidemiology, 153(12), 1222–1226. https://doi.org/10.
3844 1093/aje/153.12.1222
Ruining Jin https://orcid.org/0000-0002-8542-7614 Gambetta, D. (2011). ‘‘The Sicilian Mafia’’. Twenty years after
publication. Sociologica, 2, 1–11. https://doi.org/10.2383/
35869
Data Availability Statement Gill, J. (2015). Bayesian methods: A social and behavioral
The data and code that support the findings of this study are sciences approach (3rd ed.). CRC Press.
available on The Open Science Framework for later replications Graafland, J., & Lous, B. (2019). Income inequality, life satis-
(https://osf.io/h8gq7/). faction inequality and trust: A cross country panel analysis.
Journal of Happiness Studies, 20, 1717–1737. https://doi.org/
10.1007/s10902-018-0021-0
References
Halsey, L. G., Curran-Everett, D., Vowler, S. L., & Drum-
Adjaye-Gbewonyo, K., Kawachi, I., Subramanian, S. V., & mond, G. B. (2015). The fickle P value generates irreprodu-
Avendano, M. (2018). High social trust associated with cible results. Nature Methods, 12(3), 179–185. https://doi.
increased depressive symptoms in a longitudinal South Afri- org/10.1038/nmeth.3288
can sample. Social Science & Medicine, 197, 127–135. Hamamura, T., Li, L. M. W., & Chan, D. (2017). The associa-
https://doi.org/10.1016/j.socscimed.2017.12.003 tion between generalized trust and physical and psychologi-
Algan, Y., & Cahuc, P. (2013). Trust and growth. Annual cal health across societies. Social Indicators Research, 134(1),
Review of Economics, 5, 521–549. https://doi.org/10.1146/ 277–286. https://doi.org/10.1007/s11205-016-1428-9
annurev-economics-081412-102108 Harraka, M. (2002). Bowling alone: The collapse and revival of
Belli, S., & Broncano, F. (2017). Narratives of trust: Sharing American community, by Robert D. Putnam. Journal of
knowledge as a second-order emotion. Human Affairs, 27(3), Catholic Education, 6(2), 266–268. https://doi.org/10.15365/
241–251. https://doi.org/10.1515/humaff-2017-0021 joce.0602122013
Le et al. 13

Intravia, J., Stewart, E. A., Warren, P. Y., & Wolff, K. T. Open Science Collaboration. (2015). Estimating the reproduci-
(2016). Neighborhood disorder and generalized trust: A bility of psychological science. Science, 349(6251),
multilevel mediation examination of social mechanisms. aac4716aac4716. https://doi.org/10.1126/science.aac4716
Journal of Criminal Justice, 46, 148–158. https://doi.org/10. Parkes, J., & Blanco, J. (2022). Latin America hosts over 60%
1016/j.jcrimjus.2016.05.003 of world’s most crime-stricken cities—Global ranking. Verisk
Kuroki, M. (2011). Does social trust increase individual Maplecroft. https://www.maplecroft.com/insights/analysis/
happiness in Japan? The Japanese Economic Review, latin-america-hosts-over-60-of-worlds-most-crime-stricken-
62(4), 444–459. https://doi.org/10.1111/j.1468-5876. cities–global-ranking/
2011.00533.x Pavot, W., & Diener, E. (1993). The affective and cognitive
Kwon, O. Y. (2019). Interpersonal trust and its socioeconomic context of self-reported measures of subjective well-being.
effects in Korea. In Kwon, O. Y. (ed.), Social trust and eco- Social Indicators Research, 28, 1–20. https://doi.org/10.1007/
nomic development (pp. 177–204). Edward Elgar Publishing. BF01086714
https://doi.org/10.4337/9781784719609.00014 Rose-Ackerman, S. (2001). Trust, honesty and corruption:
Le, T.-T., Nguyen, M.-H., & Vuong, Q.-H. (2022). Chapter 4: Reflection on the state-building process. European Journal
Trust in mindsponge: A new perspective on information of Sociology/Archives Européennes de Sociologie, 42(3),
reliability. In Q.-H. Vuong, V.-P. La, & M.-H. Nguyen 526–570. https://doi.org/10.1017/S0003975601001084
(Eds.), The mindsponge and BMF analytics for innovative Roth, F. (2022). The effect of the financial crisis on systemic
thinking in social sciences and humanities (pp. 67–86). De trust. In F. Roth (Ed.), Public support for the euro (pp. 219–
Gruyter. https://doi.org/10.2478/9788367405119-009 229). Springer International Publishing. https://doi.org/10.
Martı́nez, L. (2019). Trust, life satisfaction and health: Popula- 1007/978-3-030-86024-0_11
tion data in mid-size city in the Global South. Data in Brief, Rousseau, D. M., Sitkin, S. B., Burt, R. S., & Camerer, C.
27, 104639. https://doi.org/10.1016/j.dib.2019.104639 (1998). Not so different after all: A cross-discipline view of
Martı́nez, L. M., Estrada, D., & Prada, S. I. (2019). Mental trust. Academy of Management Review, 23(3), 393–404.
health, interpersonal trust and subjective well-being in a https://doi.org/10.5465/amr.1998.926617
high violence context. SSM—Population Health, 8, 100423. Schilke, O., Reimann, M., & Cook, K. S. (2021). Trust in social
https://doi.org/10.1016/j.ssmph.2019.100423 relations. Annual Review of Sociology, 47(1), 239–259.
Mather, M. (2020). How do cognitively stimulating activities https://doi.org/10.1146/annurev-soc-082120-082850
affect cognition and the brain throughout life? Psychological Shao, L., Dong, Y., & Zhang, D. (2021). Effects of security on
Science in the Public Interest, 21(1), 1–5. https://doi.org/10. social trust among Chinese adults: Roles of life satisfaction
1177/1529100620941808 and ostracism. The Journal of Social Psychology, 161(5),
Mayer, R. C., Davis, J. H., & Schoorman, F. D. (1995). An 560–569. https://doi.org/10.1080/00224545.2020.1871312
integrative model of organizational trust. The Academy of Shin, D. C., & Johnson, D. M. (1978). Avowed happiness as an
Management Review, 20(3), 709–734. https://doi.org/10. overall assessment of the quality of life. Social Indicators
2307/258792 Research, 5, 475–492. https://doi.org/10.1007/BF00352944
Muraskin, W. (1974). The moral basis of a backward sociolo- Tabellini, G. (2010). Culture and institutions: Economic devel-
gist: Edward Banfield, the Italians, and the Italian-Ameri- opment in the regions of Europe. Journal of the European
cans. American Journal of Sociology, 79(6), 1484–1496. Economic Association, 8(4), 677–716. https://doi.org/10.
https://doi.org/10.1086/225716 1111/j.1542-4774.2010.tb00537.x
Nguyen, M.-H., La, V.-P., Le, T.-T., & Vuong, Q.-H. United Nations Office on Drugs and Crime (UNODC). (2006).
(2022a). Chapter 5: Bayesian Mindsponge Framework Violence, crime and illegal arms trafficking in Colombia.
analytics: A novel methodological approach for social Vehtari, A., Gelman, A., & Gabry, J. (2017). Practical Bayesian
sciences and humanities. In Q.-H. Vuong, V.-P. La, & M.- model evaluation using leave-one-out cross-validation and
H. Nguyen (Eds.), The mindsponge and BMF analytics for WAIC. Statistics and Computing, 27(5), 1413–1432. https://
innovative thinking in social sciences and humanities doi.org/10.1007/s11222-016-9696-4
(pp. 87–116). De Gruyter. https://doi.org/10.2478/978836 Vuong, Q.-H. (2018). The (ir)rational consideration of the cost
7405119-010 of science in transition economies. Nature Human Behaviour,
Nguyen, M.-H., La, V.-P., Le, T.-T., & Vuong, Q.-H. (2022b). 2(1), 5. https://doi.org/10.1038/s41562-017-0281-4
Introduction to Bayesian Mindsponge Framework analytics: Vuong, Q.-H. (2020). Reform retractions to make them more
An innovative method for social and psychological research. transparent. Nature, 582(7811), 149. https://doi.org/10.1038/
MethodsX, 9, 101808. https://doi.org/10.1016/j.mex.2022. d41586-020-01694-x
101808 Vuong, Q.-H. (2022). The Kingfisher story collection (Nguyen,
Nguyen, M.-H., Le, T.-T., Nguyen, H.-K. T., Ho, M.-T., H.-K. T., Trans.). AISDL.
Nguyen, H. T. T., & Vuong, Q.-H. (2021). Alice in Suicide- Vuong, Q.-H. (2023). Mindsponge theory. De Gruyter.
land: Exploring the suicidal ideation mechanism through Vuong, Q.-H., La, V.-P., & Nguyen, M.-H. (Eds.). (2022). The
the sense of connectedness and help-seeking behaviors. mindsponge and BMF analytics for innovative thinking in
International Journal of Environmental Research and Public social sciences and humanities. De Gruyter. https://doi.org/
Health, 18(7), 3681. https://doi.org/10.3390/ijerph18073681 10.2478/9788367405119
OECD. (2017). OECD guidelines on measuring trust. https://doi. Vuong, Q.-H., La, V.-P., Nguyen, M.-H., Ho, M.-T., Tran, T.,
org/10.1787/9789264278219-en & Ho, M.-T. (2020). Bayesian analysis for social data: A
14 SAGE Open

step-by-step protocol and interpretation. MethodsX, 7, Vuong, Q.-H., Nguyen, M.-H., & Le, T.-T. (2021). A Mind-
100924. https://doi.org/10.1016/j.mex.2020.100924 sponge-based investigation into the psycho-religious mechan-
Vuong, Q.-H., Le, T.-T., La, V.-P., Nguyen, H. T. T., Ho, M.-T., ism behind suicide attacks. De Gruyter.
Van Khuc, Q., & Nguyen, M.-H. (2022). Covid-19 vaccines Vyrost, J., Kentos, M., & Fedakova, D. (2007). Value orienta-
production and societal immunization under the serendipity- tions as factors of social trust, political participation, and life
mindsponge-3D knowledge management theory and concep- satisfaction. Studia Psychologica, 49(4), 283–294.
tual framework. Humanities and Social Sciences Communica- Wagenmakers, E.-J., Marsman, M., Jamil, T., Ly, A., Verha-
tions, 9, 22. https://doi.org/10.1057/s41599-022-01034-6 gen, J., Love, J., Selker, R., Gronau, Q. F., Šmı́ra, M., Eps-
Vuong, Q.-H., Le, T.-T., La, V.-P., & Nguyen, M.-H. (2022). kamp, S., Matzke, D., Rouder, J. N., & Morey, R. D.
The psychological mechanism of internet information pro- (2018). Bayesian inference for psychology. Part I: Theoreti-
cessing for post-treatment evaluation. Heliyon, 8(5), e09351. cal advantages and practical ramifications. Psychonomic
https://doi.org/10.1016/j.heliyon.2022.e09351 Bulletin & Review, 25(1), 35–57. https://doi.org/10.3758/
Vuong, Q.-H., & Napier, N. K. (2015). Acculturation and glo- s13423-017-1343-3
bal mindsponge: An emerging market perspective. Interna- World Bank. (2022). Intentional homicides (per 100,000 peo-
tional Journal of Intercultural Relations, 49, 354–367. https:// ple)—Colombia. https://data.worldbank.org/indicator/VC.
doi.org/10.1016/j.ijintrel.2015.06.003 IHR.PSRC.P5?locations=CO