The satisfaction with life scale and the well-being index WHO-5 in young Peruvians: a network analysis

ABSTRACT

The main objective of this study is to calculate the structure of a network of elements of life satisfaction and emotional well-being in young Peruvians, through network analysis, using the Satisfaction with Life Scale (SWLS) and the Well-Being Index WHO-5. A total of 414 university students between the ages of 18 and 27 (M = 22.18; SD = 2.57) participated. The results show that network analysis offers new avenues for exploring the dimensionality of complex constructs, concluding that the SWLS and WHO-5 represent distinct but related dimensions of subjective well-being. In addition, the items linked to the global judgement of satisfaction with one’s own life and with the activation of the arousal level stand out for their relevance in the network. Finally, some differences in network connectivity according to sex are highlighted, which merit further research.

KEYWORDS:

• Subjective well-being
• life satisfaction
• emotional well-being
• Peruvian youth

Introduction

Since 1948, the definition of health has referred to physical, mental and social well-being (World Health Organization, 1958) and not simply to the absence of disease. Since then, interest in the study of well-being has been on the rise and has increased significantly in the last two decades, accumulating evidence of its relevance for its impact on physical health and longevity (Diener & Chan, 2011; Kushlev et al., 2020). In addition, the relevance of well-being has also been manifested in the political sphere, being considered as a social indicator that reflects the quality of life (Campbell, 1976; Diener, 2000) and becoming key for the formulation of public policies (Kusier & Folker, 2021). Indeed, efforts have been made by the governments of different countries to measure and monitor the well-being of populations to promote intervention strategies to improve the mental health of societies (Diener, 2006; Haro et al., 2014; Helliwell et al., 2020).

Well-being is a complex phenomenon and, within the different disciplines that deal with its study, there is great controversy about its nature and theorization (Reppold et al., 2019). In psychology, although there are different positions, the conception of well-being has been approached mainly from two perspectives (Ryan & Deci, 2000). The first approach is the hedonistic one that focuses on the search for pleasure and the reduction of pain, giving rise to subjective well-being (SWB). The second approach is the eudaimonic approach that focuses on full functioning through personal growth and active contribution, leading to psychological well-being that aims to go beyond mere enjoyment and is related to the pursuit of excellence that gives meaning to life (Ryff, 1989).

This study approaches well-being from the hedonist tradition, considering that SWB can be defined by conscious experiences in relation to people’s own emotions or subjective judgements about their lives. Therefore, it refers to a subjective experience that goes beyond the objective conditions in which one lives, which may depend on economic income, marital status or physical health (Angner, 2010). Although SWB was initially understood as a unidimensional phenomenon (Lucas et al., 1996), it later became evident that it is rather a multidimensional concept consisting of a cognitive and an affective component (Diener, 1984).

The cognitive element is represented by life satisfaction, which refers to a person’s objective evaluation of his or her own quality of life in general terms. This component focuses on the self-assessment of the quality of life derived from the comparison between the current state and the standard of what the individual desires (Diener et al., 1985). Life satisfaction is one of the most relevant indicators of psychological functioning and SWB, so its measurements are used in studies of global scope, such as The World Happiness Report (Helliwell et al., 2020). Instruments measuring life satisfaction include items asking people to judge how satisfied they are with their lives, and some authors point out that these questions reflect well-being in a more stable and more comprehensive way than measures associated with emotional experience (Diener & Suh, 1997; Helliwell et al., 2020).

One of the most widely used measures to assess the cognitive component of SWB is the Satisfaction with Life Scale (SWLS; Kusier & Folker, 2021), developed by Diener et al. (1985). It is a brief measure that assesses overall satisfaction with one’s own life experience, under the argument that people choose the domains that are relevant to themselves and that they will consider evaluating their own life, regardless of their emotional state (Pavot & Diener, 1993). This instrument has been used in hundreds of studies in different countries (Pavot & Diener, 2008), has been adapted to numerous languages, and its psychometric properties have been evaluated in different populations, including adolescents, youth, adults, and the elderly, as well as patients with different physical and mental health problems (Emerson et al., 2017).

The validity of the SWLS has been evidenced through its relationship with measures of mental health, physical health status (Ngamaba et al., 2017) and by its predictive power for behaviours associated with suicidal behaviours (Pavot & Diener, 2008). In addition, evaluations of its psychometric properties have consistently demonstrated adequate reliability and confirmed a unifactorial structure in different populations (Pavot & Diener, 1993, 2008), both in Peru (Calderón De la Cruz et al., 2018) and in samples from other countries (Emerson et al., 2017). In samples of young people, the unidimensional structure of the SWLS was also corroborated in samples from different cultural contexts (Barki et al., 2020; Delgado-Lobete et al., 2020; Joshanloo, 2022; Jurado et al., 2019; Sovet et al., 2016; Vela et al., 2017).

Despite the benefits of the SWLS, it has been recommended that a comprehensive assessment of SWB requires the use of independent measures for the measurement of its cognitive and affective components. Indeed, the SWLS is recommended as a complementary measure to instruments that assess the emotional aspects of well-being or those focused on psychopathology (Pavot & Diener, 1993), considering that these components should be clearly differentiated. Life satisfaction, being an overall judgement of what one expects from life and what one has been able to achieve, is not necessarily linked to recent emotional experience (Kusier & Folker, 2020).

Additionally, the affective component has been approached from different perspectives. In this regard, some authors have pointed out that it is necessary to highlight the differences between the concepts of happiness and satisfaction, considering that happiness is an affective experience, while satisfaction refers to a cognitive experience (Campbell, 1976). Other authors have assumed more complex positions, in which positive affect and the absence of negative affect are part of the affective component of SWB (Andrews & Withey, 1974; Diener, 1984). In any case, measures of the affective element consist of introducing items that ask about people’s emotional states, as, for example, when they are asked to indicate to what extent they have felt cheerful during the course of the last few weeks (Kusier & Folker, 2020).

One of the most widely used scales to measure the affective component of SW is the PANAS (Watson et al., 1988), although its use has been more frequent for the assessment of mood states than in studies linked to health care (McDowell, 2010). It is a two-factor measure of positive and negative affect, although some studies have shown that negative affect is divided into two lower-order factors, distress and fear (Killgore, 2000). Meanwhile, the study of emotions has questioned the relevance of the hedonistic distinction between positive affect and negative affect, and has pointed out that the real relevant differentiation is between liking (pleasurable, happy, pleased) and arousal (active, energetic) (de Boer, 2014); for example, from the circumplex model of affect, the interrelationships of the affective dimensions are represented in a circular space formed by two bipolar dimensions: pleasure-displeasure and arousal-somnolence (Russell et al., 1989).

A circumplex model of affect represents this set of mutual relationships by placing concepts related to feelings in a circular order in a space formed by two bipolar dimensions: pleasure-displeasure and arousal-somnolence.

Another instrument that is widely used in research linked to well-being and mental health is the Well-Being Index WHO-5, designed by Per Bech in collaboration with WHO in 1998 (Topp et al., 2015). It is a brief five-item scale that has been considered a measure of the emotional component of well-being since the hedonistic tradition (Kusier & Folker, 2020) and is often used as a screening test for depression (McDowell, 2010). The WHO-5 has the advantage of addressing both the liking (feeling cheerful) and the arousal (feeling vigorous) dimensions of affectivity and, although its items only address positive affect, it can be argued that the omission of negative affect is offset by its practicality and psychometric parsimony as one of its strengths (Kusier & Folker, 2020).

Since the WHO-5 was introduced, it has been translated into more than 30 languages and has been widely used as a measure of well-being in different countries. Its validity has been demonstrated as a measure for the detection of mental health problems and has been successfully used for the evaluation of the impact of clinical trials (Topp et al., 2015). In addition, its psychometric properties have been tested in different studies with diverse populations, and consistently found adequate reliability and an adequate fit to the unifactorial model (Carranza Esteban et al., 2023; Caycho-Rodríguez et al., 2020; Fung et al., 2022; Lucas-Carrasco, 2012; Perera et al., 2020).

As described above, well-being is a complex concept that has been defined from multiple perspectives, so its measurement must consider the theoretical conceptualization from which the instrument used is derived in order to arrive at valid and pertinent interpretations. In this regard, the results of the application of instruments such as the SWLS and the WHO-5 should not be understood as replaceable, as they examine different aspects of the SWB. The scores obtained from the SWLS would represent global judgements of life satisfaction that do not necessarily depend on conjunctural aspects; while the scores obtained from the WHO-5 would represent reports on the frequency of emotional states in a given period that may respond to specific situational conditions.

In this context, network analysis, as a data-driven approach (Epskamp et al., 2018), is considered to be a suitable analysis technique to assess the two-dimensional model of well-being represented in a weighted network between observed variables, through the use of the SWLS and WHO-5 scales. Therefore, this study provides a novel methodological perspective for the assessment of the dimensionality of well-being, which allows a better understanding of the interactions between the cognitive and affective components of SWB. The study in question focuses on calculating the structure of a network of elements of life satisfaction and emotional well-being in young Peruvians. Specifically, the purpose is to discover how these elements are interconnected, determine which is the most central, and explore how these dynamics vary by gender.

SWB research in university students is relevant as it has been shown to be associated with better perceived health and increased physical activity (Zhang et al., 2021), greater sense of peace and hope (Alorani & Alradaydeh, 2018), and less procrastination (Berber Çelik & Odaci, 2022). All these aspects should be considered when dealing with SWB in young university students, as they are a vulnerable group as they are in the transition from adolescence to youth in a context of multiple demands related to their social relationships, their academic and professional future. They are therefore at higher risk for mental health problems compared to other young people of the same age group in the general population (Farrer et al., 2016). This reality is not foreign to the Peruvian context; indeed, a recent study reported that Peruvian university students presented moderate levels of stress; in addition, those aged between 16 and 25 years showed more symptoms of anxiety and depression, compared to those of older age (Farfán-Latorre et al., 2023). Another study conducted during the pandemic by COVID-19 found that 47.2% of students in four regions of Peru presented symptoms of anxiety or depression, 12.7% presented moderate or severe symptoms of depression, and 33.3% reported suffering symptoms of psychological distress (Figueroa-Quiñones et al., 2022). Therefore, this study is justified by the need for empirical evidence that contributes not only to theoretical reflection on SWB, but also to decision-making related to its evaluation and intervention in young Peruvian university students.

Method

Participants

Information was collected from 458 university students from two regions in urban areas of Peru, although 44 cases were excluded in the data cleaning process because they were students older than 29 years of age, from international exchange programmes or from graduate programmes. Finally, there were 414 participants aged 18–27 (Mean = 22.18; SD = 2.57); 209 males (50.50%) and 205 females (49.50%). The individuals who made up the sample for this study were chosen according to specific inclusion criteria, among which the following stand out: Be between 18 and 29 years of age and be enrolled in a university in an undergraduate programme. Considering the global context of the pandemic, a non-probabilistic convenience sampling was used (Maxwell, 2012). The sample size was obtained through a simulation procedure, using netSimulator from the bootnet library. In that regard, it was observed that a sample size of 250 data is achieved with a correlation, sensitivity, and specificity greater than 0.80 between the true and estimated network (Epskamp & Fried, 2018). Therefore, this study exceeded the minimum necessary to implement the network analysis. Among other characteristics of the study, there is a larger number of people who only study (55.10%) and a smaller number of people who study and work (44.90%). In addition, the majority of students pursue degrees in areas related to health (23.2%), engineering (17.6%) and business (15%).

Measurement instrument

Sociodemographic data, in order to obtain a detailed overview of key variables such as semester, occupation and career, an ad hoc sociodemographic file was prepared.

Life Satisfaction Scale (SWLS; Diener et al., 1985). This is a unidimensional measure of life satisfaction, composed of five items. The version adapted to the Peruvian context by Caycho-Rodríguez et al. (2018) was used. The items are scored on a five-point Likert-type scale, from strongly disagree (1 point) to strongly agree (5 points).

General Well-Being Index (WHO-5 WBI). It is a unidimensional scale consisting of five items that inquire about the presence of aspects related to emotional well-being. The version adapted to the Peruvian context by Caycho-Rodríguez et al. (2020) was used. The instructions ask participants to respond to the items in relation to how they have felt in the last two weeks. The items are scored using a Likert-type scale with four response options, ranging from never (0 points) to always (3 points).

Procedures

For data collection, a virtual questionnaire was developed through the Google Forms platform. The link was sent through the social networks Facebook and Instagram. The form consisted of a first section containing the informed consent form, in which the objective of the study and the anonymous and voluntary nature of participation in the study were made known. The following sections present the sociodemographic data sheet and the measurement instruments. Data collection lasted approximately four weeks. The study was approved by the Teaching Review Board corresponding to the Psychology Program Department of the Universidad San Ignacio de Loyola based on ethical guidelines within the institutional regulations, the Helsinki Convention, and the Code Ethics of Association of Psychologists of Colegio de Psicólogos del Perú (2017).

Data analysis

During the development of this research, the analyses are supported by the R programming language and RStudio (RStudio Team, 2022) is mainly used to perform the necessary analyses. To design network graphs, the tools ‘qgraph’ version 1.9 (Epskamp et al., 2012) and ‘bootnet’ version 1.5 (Epskamp, 2020) are incorporated. For the proper interpretation of various data sets, the guidelines proposed by (Burger et al., 2020), which address issues such as estimation, stability, inference, and network comparison, were followed.

An independent evaluation of the networks was carried out using the ggmModSelect method, in combination with a Spearman correlation matrix, and this combination has proven to be the most useful for examining items in Likert format (Isvoranu & Epskamp, 2023). Centrality was analysed through Strength, since it is the most stable centrality index within the field of psychology (Hallquist et al., 2021). Along the same lines, Bridge Strength was used as a bridge index to probe centrality between communities (Jones et al., 2021).

The robustness of the network was evaluated in detail using the bootnet package version 1.5 (Epskamp, 2020). For this purpose, a nonparametric resampling technique, focused on the case bootstrap method, was used, with a total of 1000 iterations. The correlation stability coefficient (CS) was adopted, which indicates the maximum number of cases that can be omitted, ensuring a minimum correlation of 0.70 between the original measurement and the measurement obtained with a smaller data set. It is crucial that the CS value does not fall below 0.25, ideally above 0.50 (Epskamp & Fried, 2018).

The accuracy of the network was evaluated through repeated estimation of the model based on sampled data. Thus, connections in the network can be assessed. This facilitates the creation of resampled edges and 95% bootstrap confidence intervals, the length of which indicates the accuracy of the connections (Epskamp et al., 2018).

A comparison between two networks (male and female) was conducted making use of the ‘NetworkComparisonTest’ library (Van Borkulo, 2015). The NCT employs a two-tailed permutation test where the discrepancies between groups are determined by 1000 iterations by randomly assigning each individual. The null hypothesis holds that both groups are equivalent at a significance level of 0.05. Finally, as a way of exploring the effect size of the matrices, Spearman’s correlation was calculated.

The R codes and database are available at OSF: https://osf.io/fntvz/?view_only=036ab1c05d8344668d8c92dd7e10d2eb [Temporary link for peer review purposes]

Results

Network estimation and centrality

Figure 1 shows associations between life satisfaction (SWLS) and emotional well-being (WHO-5) items. Among the SWLS items, the relationships of swls01 with swls3 (r = .27) and swls3 with swls5 (r = .32) stand out. In the emotional well-being items, who2 and who3 have a relationship of .34, while who4 and who5 have a value of .22. When crossing both groups, swls5 and who1 present an association of .19, and swls3 with who5 of .15, indicating relevant connections between life satisfaction and emotional well-being. In addition, 19 out of 45 edges were other than zero (42.22% of the density).

Figure 1. Network estimation and centrality indices.

The z-score reflects the relationship of each item to the centrality metric: positive values (right) indicate greater interconnectedness of the item in the network, whereas, negative values (left) indicate less interconnectedness. Strength is the sum in absolute value of all the relationship weights. The bridge strength is the sum in absolute value of all the relationship weights considering their community of origin.

Within the centrality indicators, item swls3 stands out for having the highest value in Strength with 1.94, while item swls4 presents the lowest value with −1.55. In relation to Bridge Strength, item swls1 has the lowest z-score with −1.86 and, surprisingly, item who1 shows the highest one in the list. These z-scores are important because they reflect the influence or connectedness of a specific item in the network. A high z-score suggests greater influence or connectedness, as opposed to low values that indicate less influence.

Stability and accuracy of the network

Figure 2A shows a series of mean values for Bridge Strength. These values, ordered from highest to lowest, decrease from .92 to .24. This decrease in mean correlation values indicates instability in Bridge Strength resampling. The mean values of the Strength index according to the number of people are shown. Starting with a high value of .99 and gradually decreasing to .75. Despite that, the Strength values are observed to be more robust as opposed to the Bridge Strength.

Figure 2. Precision and stability of the network.

A: This is the stability graph of the correlations for ‘strength’ and ‘bridgeStrength’ with a reference sample when varying the sample size. Strength’ shows a stable correlation despite reducing the number of cases sampled, while ‘bridgeStrength’ indicates a higher sensitivity to sample size. The shaded areas represent the margin of error associated with each correlation. B: This is the accuracy graph, showing a comparison between the bootstrap means (black line) and the actual data (red line) of a sample, applied to the connections of a network. The closeness between the black and red lines indicates precision of the estimates. The shaded area represents the confidence interval or variability of the bootstrap estimates.

Figure 2B shows the differences between original and bootstrapping-resampled association values. These differences range from minimal (0.00) to somewhat higher, but still low (0.08). These small values indicate precision in the edges. The largest difference occurs in “swls2-swls5 although it is still considerably small. Overall, the resampled data are consistent and accurate with the original data.

Comparisons

Overall, the comparison networks prove to be invariant (M = 0.29; p = .44) and exhibit uniform connectivity (S = .11; p = .52). Additionally, Spearman’s correlation between the edges reveals that the matrices are different (r = .66).

Figure 3 presents the comparison according to sex. The women’s graph shows the relationships between Satisfaction and Well-being, the most outstanding correlation was between swls3-who5 (r = .18) and swls5-who1 (r = .14). Furthermore, in the male network, when evaluating the correlations between Satisfaction and Emotional Well-being, the highest correlation was found to be between swls2- who2 (r = .29), and the smallest one was found to be between swls4-who3 (r = .14). Finally, when comparing the matrices by sex, it is observed that, in women, the most robust correlation between Satisfaction and Well-being was swls3 and who5 (r = .18), while swls2-who2 predominated in males (r = .29). In addition, correlations within Satisfaction are generally more pronounced in females, whereas, in Emotional well-being, males show more prominent correlations. Finally, the most influential SWB in males is swls4 and in females swls3.

Figure 3. Comparison of networks according to gender.

The z-score reflects the relationship of each item to the centrality metric: positive values (top) indicate greater interconnectedness of the item in the network, whereas, negative values (bottom) indicate less interconnectedness.

Discussion

The study of SWB has gained special relevance in recent decades, accumulating great evidence of its relationship with the quality of life, physical health, and mental health of people. In addition, the discussion regarding its theorization and the validity of the instruments for its measurement is still ongoing. In this context, the purpose of this research is to calculate the structure of a network of elements of life satisfaction and emotional well-being in young Peruvians. Specifically, we sought to discover how these elements are interconnected, to determine which is the most central and to explore how these dynamics vary according to gender.

In relation to the first specific objective, based on a visual inspection of the estimated network, it was shown that the items of the SWLS and the WHO-5 appear clearly grouped in two communities of nodes. Therefore, it can be stated that the measurements with the SWLS and the WHO-5 are not replaceable, and it is possible to measure separately the cognitive and affective components of well-being, since they constitute different sources of information. On the one hand, SWLS is considered as a measure of cognitive judgement about one’s own life, in general terms and not limited to a time period (McDowell, 2010; Pavot & Diener, 2008); on the other hand, WHO-5 is the measure of well-being, linked to affectivity in terms of positive emotions of liking and activation states (Kusier & Folker, 2020).

In addition, the results show that there is an interconnection between the cognitive and affective components of the SWB, highlighting the relationships between item 5 of the SWLS (’The circumstances of my life are good.“) and item 1 of the WHO-5 (”I have felt cheerful and in good mood.’), demonstrating the association between about one’s own life and mood states (Collier et al., 2020; Pavot & Diener, 2008), although this relationship is not moderate or strong. Also, a relationship has been found between item 3 of the SWLS (‘I am satisfied with my life’) and item 5 of the WHO-5 (’My daily life has had interesting things for me.’). This can be explained if we consider that this last item, unlike the other items of the WHO-5, refers to a judgement about how interesting one’s own life is and not to aspects linked to positive emotions or states of arousal, so it would be a reagent closer to the cognitive component of well-being.

Regarding the contribution and specific relevance of each network element, the results showed that item 3 of the SWLS (‘I am satisfied with my life’) stands out for having the highest value in Strength; that is, this item is strongly connected to the other nodes, reflecting its stability and central relevance within the well-being assessment; which may indicate that it represents the most general and relevant level as an indicator of SWB. Similar results were found by Giuntoli and Vidotto (2021), who further noted that it is likely that this item incorporates cognitive functioning and affective experience as reference points for the construction of the global evaluation of one’s own well-being. Item 4 of the SWLS (‘If I could live my life over again, I would repeat it as it has been’) presents the lowest Strength value, which means that it contains the least relevant indicator of the SWB. This finding is consistent with previous psychometric studies that, through factor analysis techniques, examined the internal structure of the SWLS and found that this item has the lowest factor loadings (Calderón De la Cruz et al., 2018; Caycho-Rodríguez et al., 2018; Delgado-Lobete et al., 2020; DiFabio & Gori, 2016; Reppold et al., 2019). This can be explained by the temporal orientation towards the past in the wording of this item compared to the other items of the SWLS and WHO-5 which are posed from the present perspective; this could indicate that the SWB assessment is more focused with the assessment of the present.

In relation to Bridge Strength, item 1 (‘I have been feeling cheerful and in good mood’) of the WHO-5 demonstrated differential behaviours according to the samples, suggesting its relevance in the connection between the nodes of the SWB network. This variability can be interpreted as an indication that good mood and happiness significantly influence network activation, and it is possible that this item functions as a bridge between the different elements of the SWB, enhancing network activation. Although more studies are needed to understand why aspects such as happiness and good mood are relevant for the activation of the SWB network in the group of Peruvian young people, there is evidence indicating that countries in emerging economies, such as Peru, present better indicators of happiness (e.g. lower suicide rates, lower rates of anxiety and depression) compared to developed countries (World Health Organization, 2017, 2019).

From the zoo well-being syndrome hypothesis, this can be explained because the security provided by developed societies is opposed to the challenges that involved human survival and that have allowed the evolution of the hedonic system (Yamamoto et al., 2022). Hence, the importance that the Peruvian community gives to positive emotions, as a means of coping to ensure balance and well-being, can be reflected in the findings of this study.

Therefore, considering that, according to network theory, changes in bridge nodes can stimulate changes in the rest of the network (Jones et al., 2021), it is reasonable to focus on the intervention and promotion of positive affect and thereby achieve better SWB. This is related to the results of previous studies that have shown that the training of positive emotions -such as joy- or of biological reactions associated with them -such as laughter-, generate a reduction of stress and anxiety symptoms, while improving psychological well-being and perceived health (Papousek & Schulter, 2008; Weinberg et al., 2014).

A third objective was to compare the networks by sex. The results indicate that there are no overall differences between the networks, although the adjacent matrices are different. Thus, it is observed that in women the most robust correlation between SWLS and WHO-5 was between item 3 of SWLS (’I am satisfied with my life.“) and item 5 of WHO-5 (”My daily life has had interesting things for me.“). This association was also found between the total sample and its significance in the group of women, which denotes that the perception that our own life contains attractive elements is an important factor affecting the cognitive attitude of being satisfied with our life as a whole. Whereas, in males, the correlation that stood out was between item 2 of SWLS (”So far I have gotten out of life the things I consider important.“) and item 2 of WHO-5 (”I have felt calm and relaxed.’). This result indicates that in the male group the positive evaluation of the achievement of personal goals and objectives is associated with the presence of pleasant emotions linked to calmness and tranquillity. Indeed, it has previously been shown that factors associated with employment and education have greater relevance in determining life satisfaction in the male group, as opposed to the female group, who value more aspects related to marital status and interpersonal relationships (Joshanloo, 2018). Moreover, item 3 (‘I am satisfied with my life’) of the SWLS was the item with the highest Bridge Strength value and, therefore, the most influential item of the network in the female group; whereas in the male group the highest Bridge Strength was item 4 (‘If I could live my life over again, I would repeat it as it has been’) of the SWLS. These results may indicate the difference in the temporal orientation of the SWB; while women have a temporal orientation towards the present for the evaluation of their life, men have a predominantly positive orientation towards the past. Previous research examining gender differences in temporal orientation has obtained mixed results (Usunier & Valette-Florence, 2007), so their study still needs to be deepened by future research.

The findings of this study have relevant implications for public health, insofar as they provide scientific evidence that contributes to the description of possible pathways for SWB assessment and intervention. Regarding the assessment of SWB, it has been shown that measurements with the SWLS and WHO-5 are not equivalent and, in any case, can provide complementary information that helps to build a comprehensive perspective of SWB. Regarding the SWB intervention, this study identifies the core items of the network, which are related to the global perception of life satisfaction (swls3: ’I am satisfied with my life.“) and positive moods (who1: ”I have been feeling cheerful and in good mood.’). Therefore, it may be useful to consider both aspects as priorities in the intervention for SWB improvement and prevention of mental health problems in students. This is especially relevant in the group of young Peruvian university students, who present moderate levels of stress and are considered a vulnerable population to suffer different mental health problems (Farfán-Latorre et al., 2023).

It is important to identify the limitations of the study. First, the non-probabilistic nature of the sampling limits the generalizability of the results. Moreover, only university students from two regions of Peru were considered. Therefore, it is recommended that future studies consider heterogeneous samples that represent the ethnic and cultural diversity of the Peruvian population, considering relevant variables such as socioeconomic level. Additionally, it is recommended that future research replicate the research considering other population groups (e.g. adolescents, adults, or seniors) in order to evaluate the consistency of the findings. Second, the cross-sectional nature of the research design does not provide for causal-level conclusions on the interactions of the SWB dimensions, nor does it allow for an assessment of the stability of the relationships between the cognitive and affective components of the SWB. Therefore, longitudinal studies should be developed to identify causal relationships of SWB components and the stability of their connections over time. Third, differences in network connectivity have been explored, only considering gender, with the exploration of differences with respect to other relevant variables, such as age or socioeconomic level, still pending.

Finally, it has been shown that the network analysis model offers new avenues for exploring the dimensionality of complex constructs, such as SWB. Thus, it is concluded that the SWLS and the WHO-5 represent different but related dimensions of the SWB. In addition, the items linked to the global judgement of satisfaction with one’s own life and to the arousal level activation are highlighted. Both aspects offer insights to optimize SWB assessment and intervention. Finally, we highlight some differences in network connectivity by gender that merit further research.

Author contributions

All the authors participated in the preparation of the paper. Specifically, SKLH and JVL participated in the initial drafting of the manuscript and its conceptualization; authors CTS and CCC also coordinated and executed the data collection procedures; author JVL was responsible for data analysis; and author SKLH was responsible for editing the final version. All authors have approved the final version of the paper and agree to be responsible for all aspects of their work.

Disclosure statement

No potential conflict of interest was reported by the authors.

Data availability statement

The raw data of this research will be made available by email request.