Subjective wellbeing is the self-assessment of life according to different personal criteria, and is based on a conscious complex and multidimensional cognitive judgment (1). In the search for definitions of wellbeing in the literature it is possible to find different ways of evaluating life or emotional experiences that involve feelings, living conditions, experiences, desire satisfaction, and the balance between pain and pleasure (1, 2). According to the literature it is possible to find two different classifications for wellbeing, which are the hedonic and eudaimonic dimensions. Eudaimonic is driven by the cultivation of wellbeing based on long-term emotional processes, being used to refer to a combination of strengths of character encompassing facets of cooperativism (positive personal relationships), self-direction (life purpose, autonomy, environmental domain and self-acceptance) and self-transcendence (personal growth and self-realization) (3, 4). Hedonic is focused on the search for pleasure and happiness, being related to the experience of satisfaction, that is, what makes life pleasurable, the presence of positive affects and the absence of negative affects (3). Despite having different theoretical traditions, the constructs are closely related and influence each other (3).

Thus, instead of focusing in the difference between hedonic and eudaimonic components, these evaluations can be either in terms of affect (positive and negative) or cognitive reflections (5). Positive and negative affects are more related to the frequency with which people experience emotions than with the intensity of those emotions (6). Positive affects relate to pleasing emotions such as enthusiasm, energy, concentration, and satisfaction (7). On the other hand, negative affects are conditions of displeasure that involve mood swings, anger, and distress. The cognitive component of wellbeing is the self-assessment of life experiences, which is developed according to a set of self-imposed standards (7).

Children's Subjective wellbeing has been widely studied the past several decades (8). Some of the studies in children were carried out from the parents' perspective, as if children are unable to properly assess and understand issues related to their own lives (9), or were based on childhood sociology to predict future outcomes (10). However, considering children as present-moment members of the community moving toward future adults, with the right to immediate wellbeing to optimally develop their skills, studies using children-based data are relevant (10).

Childhood is a time of rapid change, and it is at this time that the trajectories of health and wellbeing are established for life, and that will impact adult life (11). Being a stage for the development of psychological difficulties and mental health problems (12, 13). The main challenge for the study of wellbeing in children and adolescents is to seek more sensitive assessment methods according to the stage of development in which the children are, therefore, the identification of characteristics that lead to the beginning, course and result requires projects that involve a younger age group to understand their vulnerabilities (13). Thus, wellbeing and quality of life measuring instruments for children must account for the stage of development of each age (14). An instrument should have different formats that are appropriate for the age group to be assessed, taking into account the normal age limitations and other factors, especially for very young or sick children who may have difficulty in providing accurate information (15). In addition, the response options should be age-appropriate (16). Some instruments use images so that children can more easily identify the best answer (17, 18).

Despite the importance of the knowledge of aspects that influence children's life according to their own point of view and their perception of interpersonal relationships, few instruments are available for measuring these aspects (19). Among them, the Autoquestionnaire Qualité de Vie Enfant Imagé (AUQEI) was initially proposed in French by Manificat and Dazord (20) and later translated into Portuguese (17) with the objective of assessing the subjective wellbeing of children between 4 and 12 years of age, based on the premise they are capable of expressing their own feelings. The instrument has been mostly used in children with health problems such as autism (21), cystic fibrosis (22), cerebral palsy (23, 24), spinal muscular atrophy (25), orofacial clefts (26), and born prematurely (27). However, no study has used confirmatory analysis to verify the validity of the data obtained with the AUQEI. The selection of the instrument for subjective wellbeing investigation in children must be guided not only by the pediatric context and the stage of development of the study population (17) but also by adequate estimates of validity and reliability of the data collected with the instrument in the sample to guarantee the quality of the obtained data (28).

The knowledge of this construct is relevant in order to allow for the early identification of children at-risk for poor wellbeing in community or educational settings (29). Therefore, to help support the improvement of child services and development of public policies, this study aimed at evaluating the psychometric properties of the AUQEI when applied to pre-school children and estimate the influence of demographic characteristics on their subjective wellbeing.

A total of 443 children participated in the study [mean age: 5.19 Standard deviation (SD = 0.64) years; 52.4% male]. The average age of the participants' mothers was 33.4 (SD = 7.01) years and most were married, had a job, and belonged to economic strata B and C. The demographic information of the sample is shown in Table 1.

The descriptive statistics of the answers given to the AUQEI items by the participants of the two samples (Test and Validation) are shown in Table 2. As none of the AUQEI items presented absolute values of skewness > 3 and kurtosis > 7, no strong deviation from normality was assumed, and therefore the psychometric sensitivity of the items was considered adequate.

Despite the adequate fit of the original AUQEI model (complete - M) to the total sample, the factors Leisure and Autonomy [Variance Inflation Factor (VIF) = 166.91], Function and Autonomy (VIF = 12.75), Function and Leisure (VIF = 125.25), Family and Autonomy (VIF = 55.80), Family and Leisure (VIF = 38.21), and Family and Function (VIF = 100.25) were collinear. In addition, a high correlation was found between the errors of items 2 and 3 of the instrument, which led to the non-convergence of the covariance matrix. Thus, each item (2 and 3) was excluded individually and then both were excluded, however, this strategy did not favor the matrix's convergence. Collinearity may suggest that the AUQEI applied to this study sample has a single factor model and the adequacy verification indices of the one-dimensional model reinforce this suggestion (UniCo = 0.995; ECV = 0.946; MIREAL = 0.161), therefore, this proposal was tested. In the single factor model, was observed high correlation between items 2 and 3, we excluded item 2, based on the modification indices and the theoretical content of the item. The single-factor model showed adequate fit to the sample as well as convergent validity and adequate reliability both in the total sample (λ = 0.71–0.92; χ²/df = 4.47; CFI = 0.98; TLI = 0.98; RMSEA = 0.08; AVE = 0.70; α = 0.98; ω = 0.93) and in the Test and Validation samples (Figure 2). This model has a high probability of being applied in other studies (H-Latent = 0.984; H-Observed = 0.946).

In the analysis performed on independent samples (Test, n = 212 × Validation, n = 231), strict invariance was observed (ΔCFI_{M1 − M0} = 0.001; ΔCFI_{M2 − M1} = 0.000; ΔCFI_{M3 − M2} = 0.001) indicating that the factorial structure found remains in independent samples. The structural model tested presented adequate fit to the sample (χ²/df = 3.58; CFI = 0.98; TLI = 0.99; RMSEA = 0.07). The child's sex, the mother's age, the family's economic level, the fact that the mother is or is not married, and the exercise of work activity did not impact the children's wellbeing (Table 3).

The correlational analysis between the AUQEI and the SWLS (r = 0.70, p < 0.001) pointed to an adequate positive convergent validity of the AUQEI, while the correlation between the AUQEI and the ceo-d index (r = −0.36, p < 0.001) to attest negative convergent validity of the AUQEI model.

Most children [76.7%; (95% CI: 72.7–80.5%)] had positive subjective wellbeing. The mean scores of the items from children (n = 443) and the baseline (P61.5; mean score = 2.46) are shown in Figure 3.

Higher scores of subjective wellbeing were given for items concerning recreation, vacations, and birthdays, and lower scores referred to hospitalization, medication, medical consultation, and being away from the family. This corroborates the theoretical framework of the instrument, pointing toward adequate levels of subjective wellbeing. Only item 14 (hospitalization) presented a lower score than the baseline.

The present study confirmed the validity and reliability of subjective wellbeing data of preschool children obtained with the AUQEI. The found wellbeing estimates highlight the feasibility and importance of assessing subjective wellbeing of young children and expand the knowledge required for developing educational programs for this population.

In recent years, the study of children subjective wellbeing has increased (17, 20, 23), however, it is still a challenge to choose the best measurement tool to assess the wellbeing, especially in young children. Although the AUQEI is an interesting instrument to assess wellbeing in children (17, 20), there were still no studies that supported the validity and reliability of the measure obtained with this instrument. Some important aspects in the validation process, such as item analysis, psychometric sensitivity, and factorial validity, were not performed, which may compromise the conclusions drawn from previous studies.

The AUQEI factorial model that fitted the data adequately was the single factor structure. To obtain the best fit, a new theoretical proposal different from the original one was used. The validation and reliability analysis of the original (four-factor) and the single-factor versions were conducted, and a high correlation among the factors of the original structure was found, which compromises the variances of the parameter estimates. When at least one of the variables is redundant, an estimate of negative variance occurs (28, 37). Therefore, to avoid problems in model estimation, we tested the single-factor model, which adequately fitted the data. This proposal also provided adequate validity of the AUQEI in relation to external variables, ensuring, therefore, three measures of validity to the instrument (content, internal structure and external variables). This finding should raise the alert for future users of the instrument about the importance to obtain estimates of adequate quality for each sample and study context. The process of adapting an instrument for a population is only completed after confirmation of validity and reliability applied to different samples (28, 37). The comparison of the present findings with those of others was not possible since no previous study verified the fit of the AUQEI factor model to other samples using a confirmatory analysis.

Item 2 (“Say how you feel: at night, when you lie down”) presented a high correlation with item 3 (“Say how you feel: if you have siblings, when playing with them”), and we found that item 2 had a high theoretical similarity to item 4 (“Say how you feel: at night, when sleeping”) causing collinearity, which may have resulted in the high correlation between items 2 and 3. Considering that such young children, in our sample, could not differentiate the concepts in items 2 and 4, item 2 was excluded. Evidence from the literature on child development and psychology shows that children under 6 years of age have a limited distinction of language and verbalization of emotions, especially when referring to past events, and abstract thinking capacity develops after the age of 6 (50, 51). Future studies that include children from different stages of development are suggested to assess the best theoretical/factorial proposal for the AUQEI in each of these phases.

The strong invariance of the AUQEI observed between independent samples confirmed the stability of the single-factor model applied to young children. Thus, a valid and reliable set of data obtained with the AUQEI model is provided.

Currently, investigation of wellbeing has shifted, aiming at individuals without a specific complaint or disease, focusing on promotion of wellbeing and not just its impair (52). However, most of the studies are carried out on children with health problems (20, 23, 24). So far, only the study by Assumpção et al. (17) was performed in healthy children; however, the authors did not present evidence related to the validity of the proposed model. In addition, their model was developed for a sample of children at different stages of development (4–12 years old), which may explain the difference between the data obtained in the present study in relation to the factorial model. In the study by Assumpção et al. (17) only 49 children were aged 4–6 years and, therefore, their data may have a low contribution in the 4-factor model presented. Thus, the present study contributes to the field of study by providing a specific model for children under 6 years of age; further studies may confirm or contest our findings. In addition, future studies aiming at defining the best model for assessing subjective wellbeing using the AUQEI in children at different stages of development are recommended.

No difference in the scores of subjective wellbeing was found between mothers' age, marital status, work activity, and income (see structural model Table 3). Although income has a role in wellbeing, it is not its main indicator (19), and other conditions such as culture and social aspects should be considered in the relationship between income and wellbeing (53). The lack of an effect of the above variables in the responses shows that children do not have a clear perception of the interference of these conditions in their daily life, and provide a more self-focused assessment, based on the effects of their own world on the daily routine, without considering broader aspects.

Children presented a greater wellbeing score for items related to recreation, vacations, and birthdays, a finding that corroborates with previous studies (17, 54). Activities that break the routine and provide fun moments, such as free-time activities, are valued in childhood, being significant for wellbeing, especially among younger children (19). In contrast, the items with the lowest scores were those that refer to hospitalization and being away from the family, which is in line with the work of Assumpção et al. (17) not only in relation to the items, but also with regard to the overall positive wellbeing for most children. Despite presenting lower scores, the average subjective wellbeing was positive for most participants, which is explained by the concept of subjective wellbeing being a balance between positive and negative emotions, which not always is affected by some conditions, be they material, health-related, financial, or security-related. The influence of these factors depends on expectations and values of each person, the people around him or her, and the community in which he or she lives. Therefore, subjective wellbeing involves a global judgment of all aspects of life, and although some conditions can affect wellbeing, the emphasis is usually placed on the general judgment of a person's life (55). Despite the high prevalence rate of subjective wellbeing found in the sample, approximately one-fourth of the sample presented lower values than expected. Thus, further monitoring and investigation of the factors involved is suggested.

As limitation of the study, the data were collected in a single national context, namely, Brazil. To broaden the scope of research, it is recommended that further studies be conducted using samples from different national contexts. While the main factorial structure is expected to be consistent, cross-cultural comparisons may unveil variations in factor loadings and associations with other variables, fostering discussions on sociocultural factors' role in children's wellbeing. In addition, the cross-sectional design of the study does not allow the establishment of a cause-and-effect relationship. Despite this limitation, this study provides information about the psychometric properties of the AUQEI model for pre-school children and identifies the subjective wellbeing of preschool children, which can contribute to health professionals and researchers from different areas of knowledge in guiding and developing preventive strategies focusing not only on treating health problems, but also in promoting wellbeing.

The AUQEI model provided valid and reliable data, being thus an interesting instrument to assess the wellbeing of young children. This assessment can broaden the view of education and health professionals in order to better estimate, re-establish and monitor the wellbeing of children. The demographic characteristics of the mother or family analyzed in the present study, or even the child's sex, had no significant impact on the children's wellbeing. The study observed a high prevalence of children reporting a high subjective wellbeing.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The studies involving human participants were reviewed and approved by Research Ethics Committee of São Paulo State University (UNESP). Written informed consent to participate in this study was provided by the participants' legal guardian/next of kin.

BS and BM involved in conception and design, data acquisition, analysis and interpretation, and drafted and critically revised the manuscript. JM, JC, and LC involved in conception and design, data interpretation, and drafted the manuscript. All authors contributed to the article and approved the submitted version.