Parental perspectives on ECEC settings that foster child well-being: a comparison across nine European countries

ABSTRACT
 Parents play a vital role in identifying children’s needs for support and Early Childhood Education and Care (ECEC) features that support children’s well-being. This study examined parental perspectives on features of ECEC that foster young children’s well-being under and above the age of 3 years by interviewing 359 parents across nine European countries (England, Finland, Germany, Greece, Italy, the Netherlands, Norway, Poland, and Portugal). Results revealed that parental perspectives largely converged with quality features discussed in ECEC research. Process quality features were mentioned more frequently than structural features for all children 0- to 6-years-old in almost all countries. However, care-oriented features were mentioned more frequently for under 3 years, and educational-oriented features were mentioned more frequently for the older group. Regarding structural features, patterns of responses across the two age groups were similar in most countries. Age differences were not more pronounced in countries with a split governance system.


Introduction
The attendance of Early Childhood Education and Care (ECEC) settings is sometimes presented as an investment in children's well-being (Burger 2013;Stahl, Schober, and Spiess 2018). Parents may support their children's social, economic, and emotional well-being by sending them to high-quality ECEC settings (Stahl, Schober, and Spiess 2018). Examples of high-quality ECEC criteria include safety, educated staff, responsive and affectionate adult-child interactions, positive peer relationships, and the ability to explore materials (e.g. Rimm-Kaufman et al. 2009;Slot et al. 2015). This perspective requires that parents are fully informed about ECEC quality and know which ECEC features are 'high quality' and what is important to support their children's well-being (Chaudry, Henly, and Meyers 2010).
Parents' views on non-parental care and how their children's well-being can be supported appear to be influenced by cultural perspectives as well (Kim and Fram 2009). There might be certain cultural norms about cognitive stimulation, socialisation, and how the child should be prepared for later academic success (Johansen, Leibowitz, and Waite 1996). Parents vary in their parenting goals and values, supporting the idea that they enact the perspectives and practices of their cultural communities (Keller et al. 2006). Some parents might expect that their child learns to have selfreliance and respect for other people (Einarsdóttir 2010), whereas others expect collaboration with teachers or fun activities that keep their child happy when parents are at work (Laloumi-Vidali 1998). Another factor that might influence parents' perspectives on ECEC and what is important for children's well-being, is the child's age. For example, a warm relationship is often prioritised for the youngest children (Cadima et al. 2020;Ehrle, Adams, and Tout 2001), whereas more formal and structured learning activities are prioritised when children get older (Kim and Fram 2009). This study acknowledges parents as relevant stakeholders in ECEC and aims to understand their views regarding the structural and process quality features of ECEC settings that promote children's well-being, considering children's age. To complement the existing literature examining differences between ethnical and cultural groups within the same country (e.g. Ehrle, Adams, and Tout 2001;Kim and Fram 2009), this study explores differences between countries with different historical and policy contexts.

Children's well-being
Despite the growing interest in children's well-being (e.g. Gleason and Narvaez 2019;Reynolds et al. 2017;Storli and Sandseter 2019), there is an absence of theory-based and specific definitions of the construct as a whole (Forgeard et al. 2011;Lewis 2019). Ben-Arieh (2010) emphasises that perspectives on well-being focus on positive features, such as sense of belongingness, instead of focusing on the presence (or absence) of negative features of well-being, such as discomfort, distress, disability, and death. Further, there has been a shift from an emphasis on children's future 'well-becoming' (e.g. economic well-being) to an emphasis on children's current 'well-being' (e.g. being healthy and happy) (Andrews and Kaufman 1999;Ben-Arieh 2010). Another shift can be seen in the changing focus from a more 'objective' (e.g. health, education, materials, behaviours, and housing) to a more 'subjective' (e.g. personal well-being) evaluation of well-being (Alexandre et al. 2021). These shifts might be explained by the increasing awareness that children's subjective experiences are often absent in models that try to access and support children's well-being in ECEC. Even though there is a lengthy debate on how to define well-being (Dodge et al. 2012), there seems to be a consensus that the definition of children's well-being in ECEC should focus on their social, emotional, and physical wellbeing (Guérin 2012;Statham and Chase 2010).

ECEC system features
This study aimed to acknowledge and value parents' voices regarding their children's well-being, by examining their perspectives on features of one particular context -ECEC. For this purpose, trained research assistants interviewed parents from nine European countries: England, Finland, Germany, Greece, Italy, the Netherlands, Norway, Poland, and Portugal. These nine countries were selected because of relevant geographical diversity (northern, southern, western, and eastern European countries) and differences in ECEC systems (European Commission/EACEA/Eurydice 2019).
ECEC systems can be divided in unitary (Finland and Norway), split (Greece, Italy, the Netherlands, Poland, and Portugal), and mixed (England and Germany). In unitary systems, ECEC is organised as a single phase, which means that children have no transitions between institutions before they start primary school. In split systems, younger children go to different institutions than older children, which are often governed by different ministries (e.g. Ministry of Social Welfare versus Ministry of Education). These different approaches to ECEC policy are, to some degree, based on historical and political premises, namely the underlying notion to split systems that younger children require mostly 'care oriented' approaches, whereas older children require 'educationally oriented' approaches. A split ECEC system is the most common organisational structure in Europe. In most countries, the transition from one setting to the next takes place around the age of three (European Commission/EACEA/Eurydice 2019).
Despite this seemingly clear distinction between unitary and split systems, the division between these two types of systems is becoming increasingly blurred. Indeed, many countries with a split system are starting to expand some of the policies that were initially designed to regulate settings for older children to the services provided for younger children (European Commission/EACEA/Eurydice 2019). Through common standards and educational guidelines covering the entire ECEC period at the national level, diverse stakeholders in ECEC (e.g. policymakers and practitioners) may, over time, share perspectives on quality, and expectations of how services should promote children's well-being. The national systems might be reflected on and an influencing factor in parents' expectations for children under and above the age of three years old.

ECEC quality features
To align policies and practices as best as possible with children's and families' needs, this study examined parental perspectives on which ECEC features foster children's wellbeing by considering the constructs of structural and process quality, widely used in the ECEC literature (e.g. European Commission 2014; Slot et al. 2015). Structural quality features include group size, child-staff ratio, staff educational level, standards for hygienic practices and general safety, established routines for communicating with parents, and the availability of play and learning materials. Availability of continuous professional development activities for staff that promote team cohesion and a positive work climate is also considered a structural quality feature (e.g. Slot et al. 2015;Zaslow et al. 2010). It is presumed that structural quality features are associated with children's development and well-being outcomes by setting the conditions for children's daily experiences in ECEC (Burchinal et al. 2002;NICHD ECCRN 2002;Sylva et al. 2006). Process quality features refer to proximal daily experiences (Phillips and Lowenstein 2011;Sylva et al. 2006) and include responsive and affectionate adult-child interactions; stable, positive, and cooperative peer relationships; varied opportunities for peer interactions; children's involvement in decision making; and the opportunity to learn and to explore toys, tasks, and materials according to children's developmental stage (Giudici et al. 2001;Rimm-Kaufman et al. 2009). In addition, well-implemented and pedagogically oriented activities, sometimes separately referred to as curriculum or curricular quality, are part of process quality (e.g. Sylva et al. 2007).
The distinction between structural and process quality features is not always clear, as there are overlaps and fluent transitions between the two categories (Laevers 2005;Pianta et al. 2005;Sylva et al. 2007). Nevertheless, it is a helpful division for policy measures too, as they are usually defined in regulable, structural features. It should be acknowledged though that there is now an international movement to also regulate process quality features (OECD 2019).

The present study
This study examined the following research question: 'What are the perspectives of parents on the most important ECEC features to foster well-being for children under the age of 3 years versus 3-to 6-year-olds across nine European countries?' This cross-national comparative study provides more insight into whether age group differences are aligned with the structural/process divide proposed in the literature on high-quality ECEC features (e.g. European Commission 2014; Slot et al. 2015), and about potentially changing parental perspectives as children grow older. In addition, findings allow more insight into whether age group differences are more pronounced in parents from countries with a split (Italy, Greece, Poland, Portugal, and the Netherlands), mixed (England and Germany), or unitary system (Finland and Norway) (European Commission/EACEA/Eurydice 2019). We expected that age differences would be more pronounced in countries with a split versus a unitary or mixed system. Based on our literature review, we argue that this is the first study comparing parental perspectives about ECEC features that foster children's well-being across countries. Parents' perspectives on ECEC features associated with children's well-being can help inform national and European policies in the aftermath of the Recommendation on High-Quality Early Childhood Education and Care Systems (Council of the European Union 2019), making children's well-being more visible in future benchmarks and goals for high-quality ECEC.

Method
This study reports on data from the Stakeholder Study of the larger 'Curriculum and Quality Analysis and Impact Review of European Early Childhood Education and Care' project (CARE). The Stakeholder Study is a cross-sectional cross-national comparative survey study into the values, beliefs, and concerns of parents, staff, and policy representatives regarding ECEC services in nine European countries: England, Finland, Germany, Greece, Italy, the Netherlands, Norway, Poland, and Portugal. The aim of the Stakeholder Study was to address issues related to inclusiveness, quality, and individual, social, and economic benefits of ECEC in Europe.

Participants
Participants were 359 parents from nine countries (40 per country, except for Greece with 39). Parents were eligible to participate if they had at least one child from 0 to 6 years who participated in an out-of-home ECEC setting. Participating parents were recruited by e-mails, presentations, and word of mouth recruitment through ECECcentres, schools, community centres, and researchers' networks according to a convenience sampling model, to ensure country specific considerations of relevance (e.g. national political issues, presence of particular groups). Table 1 shows the background characteristics of the participating parents for each country. In Poland and Finland, there were (almost) no participating parents with an immigrant background, which is in line with population characteristics, as reported by Eurostat (2015). In Germany, Italy, Finland, Norway, and Poland the sample was somewhat highly educated, considering national statistics for adults from 30 to 34 years old (Eurostat 2015). However, as shown in Table 1, we still had considerable variation in terms of parental educational level in these five countries. Parents' average age ranged from 33.51 (England) to 38.44 years (Greece), and most were mothers (from 72.5% in Norway to 97.4% in England).

Procedure
This study used data from a larger 118-item survey with structured questions, 16 demographic questions, and 3 open-ended questions. Data were collected between February and May 2015. The survey was translated to all relevant languages, checked by external bilingual professionals, and piloted. The overall survey was administered face to face (e.g. at home or at the ECEC setting) by a trained research assistant, who spoke the preferred language of the parent. Participants were briefed on the goals of the study and encouraged to ask clarifying questions. The research assistants explicitly mentioned that there were no right or wrong responses and that differences among participants, between and within countries were expected. The research assistants obtained active consent from each participant, requesting that they ticked a consent box in the questionnaire. Personal data were not collected, and parents' answers were immediately registered in the online survey by the research assistant. Therefore, the interview was not recorded. The minimum number of face-to-face interviews per country was 40. Some countries conducted more interviews due to additional resources. In these countries, researchers selected 40 interviews based on the demographics of the parents. For the purpose of this study, all types of family backgrounds were represented as best as possible, aiming to increase diversity in the sample (e.g. including parents from lower educational backgrounds). Interviews took between 30 and 45 min.
This study focused on participants' responses to the open-ended question: 'What aspects of an ECEC setting do you think are most important to foster children's wellbeing? Please mention up to three aspects … 1) for children under the age of 3 years, 2) for children 3-6 years'. This was the first question in the survey so that parents would not be influenced by subsequent questions. No definition of well-being was provided. Moreover, parents were not told about ECEC quality features, so that we could study their own perspectives on what is important to foster children's well-being. Educational levels used by Eurostat: Low = ISCED 0-2, early childhood educationlower secondary education; Medium = ISCED 3-4, upper secondary educationpost-secondary education; High = ISCED 5-8, short-cycle tertiary educationdoctoral or equivalent level.

Analysis
The first step in the analytic process was to create a coding scheme to classify parental responses in all nine countries for both age groups. For this purpose, the responses of eight parents in each country were translated into English, ensuring the representation of parents from lower and higher educational levels, and from native and immigrant groups. Parental responses were segmented into units of meaning. Using steps from the constant comparative method (Boeije 2002;Glaser 1965), we first relied on an inductive approach by openly labelling the answers and identifying units of meaning that could be assigned the same codes. Next, higher-order codes were developed to identify the most important messages from the participants (Boeije 2010). Finally, we deductively used the existing literature on ECEC quality features and on developmental goals for young children to inform the final stages of development of the coding scheme.
Based on Burla et al. (2008), the coding scheme consisted of three levels. The highest level considered the distinction between structural and process quality features. The second and third levels included broad codes and specific subcodes, respectively. At the subcode level, specific units of meaning retrieved from the data were used as an illustration to support coders. In addition, we developed a glossary to ensure accurate interpretations (available upon request). Subcodes were mutually exclusive (i.e. units of meaning could only be assigned one single code). Every country had at least two coders (fluent in the local language and English) who translated parents' answers from the local language to English. The development of the coding scheme and glossary was done by both the authors and coders, who were all researchers in the larger CARE study.
In addition to extensive discussions among researchers from all countries to ensure the cultural sensitivity of the initial coding scheme, researchers in each country coded additional interviews to test the applicability of the coding scheme. Comparing parents' answers of the pilot and the additional interviews showed that we reached saturation, since there was no need for new codes. Moreover, researchers within all participating countries were able to apply the coding scheme to the data of their country. Only minor adjustments were made to codes that might be interpreted differently in some countries and this process continued until agreement was reached. Finally, based on theoretical considerations, some subcodes were merged to create a distinctive, meaningful, and comprehensive coding scheme. For example, pedagogically oriented activities might be considered as indicators of curriculum quality (e.g. Sylva et al. 2007). However, answers focusing on stimulating learning activities, such as reading and math activities, were coded under 'supporting the child's development'. In addition, reading and math were merged under the subcode 'development of thinking, language, and math'.
Final codes and subcodes, including an example per subcode, are presented in the Appendix. The structural quality category consisted of four higher-order codes (i.e. staff characteristics, centre organisation, staff-parent communication routines, and physical environment), and the process quality category included two higher-order codes (i.e. relations and interactions, and supporting child's development), which were divided into subcodes.

Interrater reliability
To check for intercoder reliability, 30% of the data were double coded by an independent coder in each country. Intercoder agreement was calculated by dividing the units that were assigned the same subcode by the total number of coded units (Reynolds and Livingston 2011). Assignment of a different subcode or assignment of more/fewer subcodes were considered disagreements. Except for Greece (79.1%), reliability was higher than 80% in all countries (range 82% -94.5%). In accordance with guidelines from Burla et al. (2008), disagreements were resolved through consensus. Figure 1 displays the proportion of the two higher-order categories across countries and age groups (i.e. under and above 3 years of age), capturing the extent to which parents in the nine countries identified structural and process ECEC quality features as the most important to foster young children's well-being.

Results
For the age group under 3 years, parents in seven countries mentioned process quality features more frequently than structural features (from 51% in Finland up to 66% in England). Exceptions were parents in Italy and Poland, which identified slightly more structural quality features (51% and 54%, respectively). Importantly, parents from all countries identified process quality features more often for 3-to 6-year-olds than for children under 3 years of age (up to 16 percentage points more). Only Polish parents mentioned process and structural quality features almost equally (51% and 49%, respectively) for the older age group. Over two-thirds of the parents in Norway (68%), England (78%), Poland (80%), and Portugal (88%) made no distinctions regarding what they considered important to foster children's well-being in the two age groups (i.e. they provided similar responses for both age groups). In Germany and Finland, this was the case for less than one-third of the parents (28% and 30% respectively), and, in the remaining countries, this applied to about half of the parental group (the Netherlands, 50%; Greece, 54%; Italy, 60%). Note. Deviations from 100% in the total are due to rounding issues.
Identifying specific structural and process features that foster children's wellbeing Figures 2 and 3 display the relative frequency of specific structural and process quality features of ECEC settings that parents across countries considered to be the most important to foster the well-being of children under and above 3 years of age, respectively.
Structural features that foster well-being for children under 3 years Figure 2 shows that parents from Greece, Italy, the Netherlands, Poland, and Portugal identified characteristics of the 'physical environment' as the most important structural  Note. Deviations from 100% in the total are due to rounding issues. features to foster the well-being of the younger group, compared to parents from other countries. Parents from Portugal emphasised the quality of 'indoor spaces' and 'health' and 'safety-related' features of the physical environment. Parents from Poland also emphasised the safety of the physical environment.
'Center organization' features were more frequently mentioned by parents from Germany, Finland, and Norway, compared to parents from other countries. Specifically, parents from Germany considered 'staff-child ratio' and 'routine and rhythm' as the most important features to support young children's well-being within centre organisation, while parents from Finland emphasised 'group composition' and parents from Norway valued 'group composition', 'routine and rhythm', and 'stable staff' equally. Parents from England pointed out 'staff characteristics' (15%) over other structural quality features likely to foster the well-being of the youngest children. In all countries, 'staffparent communication routines' was the least frequently mentioned structural quality feature for the youngest children.

Structural features that foster well-being for children aged 3-6 years
Even though structural quality features were mentioned less frequently for 3-to 6-yearolds, patterns were similar to those reported for the younger group (see Figure 3). Only parents in Germany, Finland, and Norway considered 'center organization' as less important for the older group compared to the youngest group (6-9 percentage points), but it was still frequently mentioned. Whereas some parents from Finland (4%) mentioned 'staff-parent communication routines' for children under 3 years old, they did not address this quality feature as important for the well-being of the older age group.

Process features that foster well-being for children under 3 years
As displayed in Figure 2, when considering the well-being of children under 3 years of age, parents from almost all countries identified quality features related to 'relations and interactions' more frequently than features related to 'supporting the child's development'. The exception were parents from Greece, who emphasised the latter. Importantly, there was considerable variation in the percentage of ECEC features related to 'relations and interactions' across countries, ranging from 25% in Italy and Poland to 43% in Finland.
There was variation in parents' perspectives on the most important features within 'relations and interactions'. Parents from Greece, Finland, and Norway emphasised a 'secure child-staff relationship'. Conversely, parents from Italy, the Netherlands, and Poland emphasised 'focused attention to individual children'. In turn, parents from Germany, England, and Portugal emphasised a 'positive climate'. Only a few parental responses addressed 'focused attention to child-group relationship' (ranging from 0% in Norway, Poland, and Portugal to 5% in the Netherlands).
Regarding 'support of the child's development', parents from Greece, England, Italy, the Netherlands, Norway, and Poland emphasised the 'development of thinking, language, and math'. In turn, parents from Germany, Finland, and Portugal emphasised 'social-emotional development'. Within 'support for the child's development', both 'physical development, skills' and 'personal skills' were less frequently mentioned by parents from all countries.

Process features that foster well-being for children aged 3-6 years
Overall, process quality features were identified more frequently for the older age group. When parents considered the well-being of children aged 3-6 years old, the frequency of features related to 'supporting the child's development' was higher across all countries (see Figure 3), with increases between 3%, for parents from England and 22% for parents from Finland. Parents from Finland focused more on the 'development of thinking, language, and math' and on 'social-emotional development', when considering the well-being of 3-to 6-year-olds. Parents from Greece mentioned the 'support of the child's development' more frequently than parents from all the other countries. Parents from Germany seemed to value the role of 'social-emotional development' for the younger group, and emphasised the 'development of thinking, language, and math' for the older age group.
Importantly, the frequency of features related to 'relations and interactions' was somewhat lower across all countries when parents considered the well-being of children aged 3-6 years, with decreases between 1% for parents from Portugal and 12% for parents from Finland. The relative importance of specific features within 'relations and interactions' when considering the two age groups was similar for parents from six countries. Slightly different patterns were found within countries for parents from Germany, Finland, and the Netherlands. For example, parents from Germany emphasised 'focused attention to individual children' for the well-being of the older children versus 'positive climate' for the younger group.

Discussion
This study aimed to investigate parents' perspectives on ECEC structural and process quality features that are most important to foster young children's well-being across nine European countries, while examining the extent to which parental perspectives varied as a function of children's age (i.e. under 3 years and for 3-to 6-year-olds). In addition, this study provided insight into whether age group differences were aligned with the structural/process divide proposed in the literature on high-quality ECEC features, and into potentially changing parental perspectives as children grow older. Moreover, findings provided insight into whether age group differences were more pronounced in parents from countries with a split, mixed, or unitary system. Process quality features were mentioned more frequently than structural quality features in both age groups, in almost all countries. Parents seemed to focus on both explicit features, wherein the staff promotes the personal skills and competencies of the child (e.g. 'focused attention to individual children', 'development of thinking, language, and math', and 'social-emotional development'), and more implicit features, regarding the provision of emotional support (e.g. 'secure child-staff relationship' and 'staff characteristics'). Parents' appreciation of process features is consistent with research that suggests that process quality is more closely related with children's outcomes than structural features, with the latter providing the conditions for the former (e.g. Melhuish and Gardiner 2019; Slot et al. 2018). This underlines the changing focus on a subjective evaluation of personal well-being (Alexandre et al. 2021). In addition, there was more focus on children's current well-being for the younger children and more focus on well-becoming for the older children.
For structural quality, patterns of responses across the two age groups were quite similar across most countries, with most parents focusing on the 'physical environment', except parents in England, who focused on 'staff characteristics', and parents in Norway, Finland, and Germany, who focused on 'center organization'. Interestingly, these latter countries have mixed and unitary ECEC systems (European Commission/EACEA/Eurydice 2019), which may account for variation in parents' perspectives. It is possible that in countries with unitary systems, 'center organization' is especially important from the perspective of parents, to respond to the needs of children from a wider age range and their families. The physical environment was especially valued by parents from Portugal, namely when considering children under 3 years of age. This may be related to the fact that the ECEC provision for infants and toddlers in this country is still relatively scarce, although rapidly expanding (from 26.7% in 2006-49.1% in 2017Folque and Vasconcelos 2019). 'Staff-parent communication routines' was rarely addressed across countries, which might imply that parents emphasise structural features that are related to the child directly, such as the 'physical environment' and 'center organization'.
We found a clearer age distinction within process quality features. Generally, for all countries, process quality features were more frequently addressed than structural quality features for both age groups, except for Italy and Poland, where structural quality features were more frequently mentioned for the younger group. Regarding process quality, features of 'relations and interactions' were mentioned more often than 'support of children's development' for the younger group in all countries, expect Greece. 'Secure child-staff relationship', 'focused attention to individual children', and 'positive climate' were most frequently mentioned for the younger group. These results are in line with the historical tendency to focus more strongly on 'care' features for under three years of age, with strong links to attachment theory, and on 'education' for 3-to 6-year-olds (Sylva, Ereky-Stevens, and Aricescu 2015).
However, for the older age group, in most countries, parents prioritised features related to 'support for the child's development'. Exceptions were parents from Norway, England, and Portugal, countries with quite diverse ECEC systems. Parents likely focus more on ECEC features related to support for children's development in the older age group because of increased well-becoming (c.f., school readiness skills, e.g. Duncan et al. 2007). Indeed, these findings might be explained by the national and European ECEC guidelines and frameworks that increasingly focus on development as children get older (Moser et al. 2016).
Our findings suggest that parental perspectives on ECEC features that foster young children's well-being largely converged with the quality features mentioned in ECEC research and policies (e.g. European Commission 2014; Slot et al. 2015). Most of the features that were mentioned by parents, such as staff-child ratio, staff characteristics, and physical environment are factors reflected in studies on structural quality (Slot et al. 2015;Zaslow et al. 2010). Parents also mentioned support of children's development, which is in line with ECEC goals (Council of the European Union 2019) and with previous studies focusing on the provision of opportunities to learn and explore materials, tasks, and toys according to the developmental stage (Giudici et al. 2001;Rimm-Kaufman et al. 2009). Moreover, parents frequently mentioned a secure child-staff relationship and focused attention to individual children, which are considered crucial features of process quality (Phillips and Lowenstein 2011;Sylva et al. 2006).
We expected that age differences would be more pronounced in countries with a split versus a unitary or mixed system, because parents in such countries tend to experience distinct service provision for younger children to a higher degree. However, we found no indications that systems' governance might explain observed differences in parents' priorities across age groups. An explanation could be that the distinction between unitary and split systems is becoming more blurred (European Commission/EACEA/Eurydice 2019). The introduction of educational guidelines for younger ages, and the increase in qualification requirements for professionals in ECEC working with the youngest children (European Commission/EACEA/Eurydice 2019) may account for a harmonisation of practices and for changes in parental perceptions. In fact, it has been argued that, as far as systems' governance is concerned, European systems can now be placed in a continuum, rather than at two ends of a spectrum (European Commission/EACEA/Eurydice 2019).
These findings raise important issues regarding the extent to which parents prioritise quality features most likely to directly influence young children's developmental outcomes (i.e. process features), in a period when young children's neurodevelopment is particularly sensitive to environmental influences (Council for Early Child Development 2010; Naudeau et al. 2010). In addition, this cross-national comparative study is a complement to cultural studies by providing insight regarding historical and policy features of countries.

Limitations and strengths
This study has several limitations. One limitation is that we did not ask parents to define well-being. Parents may have had different perspectives on what is well-being (Alexandre et al. 2021) and, therefore, may have interpreted the question differently. Therefore, it could have been useful to collect data on parents' definitions of well-being to provide further context for understanding their priorities regarding ECEC features.
The fact that we only collected data from parents is also a limitation, as it did not allow for triangulation of informants (e.g. considering the perspectives of professionals and children). However, parental perspectives are typically neglected in the literature on ECEC quality and studies that focus on their perspectives are warranted, since parents enrol their children in ECEC. Nevertheless, future research might include children's perspectives on the ECEC features that foster their well-being.
Moreover, a deeper parental explanation regarding the chosen features would have been helpful. For example, parents' perspectives regarding why each quality feature is important for children's well-being might help further our understanding on parents' priorities and understanding of children's needs. In addition, future research might examine the extent to which parents consider that each quality feature is ensured in their children's setting. Finally, since the European Union is in continuous change, other countries from, for example, Eastern Europe, should be included in subsequent studies.
Importantly, this study has multiple strengths. This study was, to the best of our knowledge, the first to examine cross-national variation in European parental perspectives regarding which ECEC features are important to foster young children's wellbeing. It did so with a diverse sample of parents within countries from different backgrounds. Methodologically, the research team developed a robust coding scheme regarding parents' perspectives on children's well-being in multiple countries. The coding scheme reached saturation, suggesting that important features to foster children's wellbeing were not missed, and that our sample was of adequate size. Finally, the interrater reliability was high in all countries.

Conclusion
Parents' perspectives on how to foster children's well-being can provide helpful input on regional, national, and European level policies to improve ECEC quality. Policies for quality improvement are generally focused on structural features, whereas policies focused on process features are scarcely mentioned in policy. However, this study suggests that process quality aspects such as secure child-staff relationships, positive climate, and attention to the individual child, are central to parents' perspectives across both age groups, and considered more important than structural aspects. Investments in process quality are also an evidence-based mechanism to improve children's development and learning, with especially strong evidence in the case of disadvantaged children (Anders et al. 2013;Van Belle 2016). Therefore, parental perspectives regarding which ECEC features are important to foster young children's well-being provide support for further investments in policies that regulate and monitor process quality.
Our findings suggest that staff in ECEC centres could engage in conversations with parents about what they think is important to foster their child's well-being. Parents' responses may be used to evaluate the care that is provided at the centre level and adapt and tailor individual and group activities that will improve well-being for every single child and family. Finally, we recommend that ECEC centres collect information on parental perspectives on a regular basis, because both parental perspectives, ECEC practices, and policies are changing continuously.