The information was gathered after receiving authorization from the institution's authorities, the parents or guardians' informed consent, and the children's assent. the aforementioned measures were distributed online using Google Form during the second period of confinement (May to September 2021); the link to the questionnaires was sent to the parents' e-mail, and by clicking on it, the students could log in to answer it. The first display consisted of the informed consent of the parent/guardian and an instruction to explicitly ask for the assent of the child Finally, the answers were automatically sent to the researchers who oversaw their analysis.

This research is a risk-free study under the considerations of the ethical standards indicated by the American Psychological Association (APA, 2020) and the participation was voluntary, anonymous, and confidential. This project is registered (6337/2021SF) and approved by the Ethics Committee (2021/P05).

Cronbach's alpha and McDonald's omega were used to measure internal consistency as evidence of precision. A latent profile analysis (Berlin et al., 2014, p.174) was performed using the maximum likelihood estimation with the following parameters: 1) Akaike information criterion (AIC) and sample-size-adjusted Bayesian Information Criterion (SABIC), being the best model the one with the lowest AIC and SABIC values; 2) Lo-Mendell-Rubin likelihood ratio test (LMRT) and the bootstrap likelihood ratio test (BLRT); 3) Entropy, which evaluates the quality of individuals' classification into groups (Araújo et al., 2018), it ranges from 0 to 1, the value closer to 1 indicates a better classification; 4) Group size, which is recommended to be no less than 5% of the sample (Del Valle et al., 2019, p. 140-141).

3.1 Transparency and openness

Information was provided on how sample size, data exclusions, and study measurements were determined. The results were obtained using SPSS version 23 (IBM, 2016) and Mplus for LPA version 8.9 (Muthén & Muthén, 2017) were used.

As a person-centered approach, the LPA emphasizes grouping individuals with similar characteristics rather than describing relationships among variables; LPA classifies individuals based on the likelihood of everyone being clustered in a particular profile (Orpinas et al., 2015). The LPA resulted in five profiles: the participants on profile 1 (moderate resilience) have the mechanisms to cope with adversity but require stimulation to develop their coping skills and abilities -this group has average scores in the evaluated variables and represents 26% of the sample; participants in profile 2 (vulnerable group) exhibit high scores on daily stressors, depression, and negative affect, as well as low resilience -this profile comprises 12% of the sample; participants in profile 3 (functional or homeostatic) depend on the protective mechanisms available at the time of adversity since it presents lower life satisfaction as well as high self-efficacy and resilience -it composes the lowest proportion of the sample at only 3 %; participants in profile 4 (resilient) showed high stress, depression, and higher life satisfaction -it represents 10% of the sample; finally, participants in profile 5 (well-adjusted or positive adaptation) have low scores for daily stressors, depression, negative affect and high positive affect, life satisfaction, self-efficacy, and resilience; they show no signs of maladjustment and comprise most of the participants in the sample at 49% (Table 1).

Table 1 Table 1 LPA model fit summary Note.

AIC = Akaike Information Criterion, BIC = Bayesian Information Criterion, SABIC = estimated for the profile group. All p values were significant at p < 0.001.

Table 2 shows the means and standard deviations for every profile, the classification probabilities for most class membership by profile, and percentages given are within the profile. The Figure 1 shows the chosen solution based on interpretability and statistical significance.

Table 2 Table 2 Five profile model and classification probabilities Note.

AIC = Akaike Information Criterion, BIC = Bayesian Information Criterion, SABIC = estimated for the profile group. All p values were significant at p < 0.001.

Figure 1. Figure 1. Latent profile of resilience.

This research aimed to analyze the latent profile of resilience using the following variables: self-efficacy, life satisfaction, positive-negative affect, daily stressors, and depression in a sample of schoolchildren during the period of confinement. Based on statistical significance, a model of five profiles is supported. The final model reveals the relationships between variables that generate a pattern that defines the profile (Del Valle et al., 2019; Soriano & Monsalve, 2019). Therefore, it is confirmed that resilience has different trajectory patterns, as Infurna and Luthar (2017) suggest, since not all dimensions of resilience have the same weight in the development of this strength (Gifre et al., 2010; Sette Galinari & Resende Bazon, 2024).

Profiles 4 and 5 obtained in this study are similar to the patterns observed by Merz and Roesch (2011) as a well-adjusted or well-adapted profile, which consists of higher emotional stability, sociability, and self-control, as well as with the resilient subtype described by Herzberg and Roth (2006) which "allows them to thrive despite environmental stressors" (Merz & Roesch, 2011, p. 6). Specifically, profile 4 presents better stress management and less depression, which are indicators related to optimal functioning and mental health (Agudelo Hernández et al., 2023; McCrae & Costa, 1991); this type of profile is associated with a successful adaptation to different life areas (Donnellan & Robins, 2010). Moreover, this profile corresponds to 10% of the sample, which is similar to the proportion reported by Chen and Tang (2021), suggesting a current indicator of the prevalence of low resilience during the pandemic (Janitra et al., 2023).

Profile 3 has a tendency for lower scores in life satisfaction, corresponding to 3% of the sample. It is recognizable that many of the everyday life experiences conditioning life satisfaction were limited during the pandemic, which resulted in low levels of satisfaction in the child community of Barranquilla, Colombia, due to the pandemic, academic demands, not being able to meet needs such as food, play, lack of economic resources, and poor socialization (Albor-Chadid et al., 2022). This tendency for low satisfaction may be because children do not have enough tools to resolve interpersonal conflicts yet (Baird et al., 2010; Verdugo & Sabeh, 2002); however, relying on a different point of view, this low satisfaction may be due to schoolchildren within this profile denoting feelings of unhappiness with the current situation, since they are not able to satisfy their desires as expected, which may be a temporary situation and that in some way this nonconformity may become a motivating factor or stimulus for personal development and growth. It is required to analyze the effect of the multiple psychosocial variables that have an influence on life satisfaction (Alfaro et al., 2016). In this context, it is crucial to note that life satisfaction is linked to both internal elements from their world perception and daily life (Várguez Pasos, 2023) and must also be seen within the framework of each culture (Diener, 2009) since “subjective well-being and its expressions cannot be understood outside the context of a particular culture or subculture” (Anguas Plata, 2005, p. 179). In addition, the fact that subjective well-being depends on various factors such as family, health, interpersonal relationships, age, gender, education, and income, among others (Carrillo Punina et al., 2016; Díaz-Loving et al., 2022) leads to the need for further analysis in these age groups.

These results provide the possibility of understanding the different paths towards resilience as indicated by Infurna and Luthar (2017), which leads to considering a first approach to the characterization of resilience and analyzing the possible effects of confinement on the children's experience (Tran et al., 2023); therefore, thinking in the application of differential intervention strategies (Marenco-Escuderos et al.,2023; Núñez et al., 2021; Sette Galinari & Resende Bazon, 2024; Suriá Martínez et al., 2015).

Covering the heterogeneity of resilient profiles can be complex, and using the LPA allowed us to simplify its understanding; however, it is necessary to keep in mind the objective of the study, as well as the possible biases in the measurement regarding a likely tendency towards social desirability, resulting in an “overestimation of individuals proportion in the classes interpreted as resilient” (Specht et al., 2014, p.34).

The profile number determination was done empirically, allowing us to guarantee the identification of relevant profiles. In addition, the findings of this research constitute a first approximation of the number of empirically and theoretically plausible types; meanwhile, the solution reflects the differences between the participants on the kind of differentiated profile but with calmness, above all. However, it is necessary to analyze whether this structure can be stable in both adolescence and adulthood since maturing processes may manifest in a different type of stress response, and empirical evidence is necessary to assert whether these typologies are different between men and women.

To our knowledge, this research is one of the few studies in Mexico to identify resilience profiles in children according to variables that are indicators of psychological well-being and distress. Thus, the present study contributes theoretically to consider resilience as a multidimensional construct (González-Arratia Lopez-Fuentes & Torres Muñoz, 2022) constituted by personal, social, and contextual factors, both internal and external, which an ecological perspective assumes that when faced with a certain amount of levels or risk factors, protective factors are activated, resulting in a better adjustment in the face of adversity (Barcelata, 2015).

Regarding the practical contribution of our findings, they focus on educational and health contexts; in the former, teachers or other education professionals use them to be alert to the presence of risk factors, for example, high levels of stress and depression, which can result in school and academic difficulties (e.g., low motivation for learning and behavioral problems). Similarly, resilience identification within the educational environment will allow for soft skills and social and vocational competencies currently in a post-pandemic scenario (Quezadas Barahona et al., 2023). Secondly, the identification of profiles with different patterns generates the need to establish specific interventions, from first-level interventions that allow early prevention aimed at increasing emotional well-being, in other cases, it will be necessary to identify psychological disorders and implement strategies to promote mental health (Folke et al., 2019; Jeffrey & Yamagishi, 2024; Sette Galinari & Resende Bazon, 2024; Shin et al., 2023).

Regarding the specific needs, the participants of profile 2 vulnerable, would benefit from the intervention aimed at the development of coping skills in the face of stress, which is proposed from the cognitive behavioral approach so that from cognitive restructuring it allows them to reduce the management of negative emotions and depression. In the case of profile 3, it is considered that the intervention would be aimed at increasing well-being, which is suggested to be from Positive Psychology, which are aimed at improving well-being and growth through the promotion of positive emotions, personal strengths, and optimism (Bohlmeijer et al., 2017).

In general, the differentiation of these profiles demonstrates the variety of responses to daily stress, the various internal resources (life satisfaction, positive emotions, self-efficacy) and sensitivity to events that can trigger mood disorders (depression) in children, so to understand the effect of traumatic events, and we need to identify protective and risk factors due to the multidimensional and dynamic nature of resilience.

As this investigation is a cross-sectional study, it is necessary to follow up on the recovery of those cases identified as vulnerable groups and their different evolutions. The non-probabilistic sampling does not allow the generalization of findings; hence, caution is recommended in their interpretation. It must be recognized that the pandemic posed various challenges to children, and these changes might induce different coping and recovering response patterns when facing confinement. We think that this is an important stage to identify different resilience patterns in Latin American children. We also identified that our methodological approach has some backdraw, in terms of generalization. First, children did not response themselves directly, but by intermediation of their parents/guardians. Second, we only have a static picture of the resilience profile; to get a more stable results, it will be compulsory to use a longitudinal sample to identify the stability of the profile. Additionally, it is necessary to identify the resilience profiles between boys and girls, as these may be similar, as proposed by Álvarez-Voces and Romero (2024), or there may be differences, which would be the direction of a future study.

It is necessary to highlight that the typologies were based on the variables evaluated, so it would be important to analyze the profiles in relation to other behavioral, cognitive, and affective variables that allow us to broaden our understanding of the multidimensionality of child resilience. Nonetheless, our results contribute to the theorization and characterization of resilience profile in Mexico and a Latin American sample and to consider that the study of resilience involves a multidimensional analysis and not in isolation (González Arratia Lopez Fuentes, 2016).