Introduction

Young children’s engagement in learning and everyday activities has been of great interest to researchers in the field of education, early intervention, and disability research (Fredricks et al., 2004; Imms et al., 2017; McWilliam & Bailey Jr., 1995; Reschly & Christenson, 2012). Interest in child engagement grew during the 1980s in the literature on school drop-out prevention and early childhood intervention, where engagement was recognized as a mediator between the quality of educational environments and positive academic and developmental outcomes (Finn, 1989; Wolery et al., 1994). Since then, child engagement has become of interest for educational practitioners and for researchers concerned with improving the quality of education in various settings (Eccles, 2016), including early childhood education and care (ECEC).

In early childhood research, child engagement has been associated with other indicators of educational quality such as peer interactions (Diebold & Perren, 2022) and teacher-child relationships (Pietarinen et al., 2014; Roorda et al., 2017). Engagement also predicts learning achievement (Aydogan, 2012; Langeloo et al., 2020; Lei et al., 2018) and well-being (Pietarinen et al., 2014; Roorda et al., 2017). Child engagement can be considered a pivotal behavior that mediates the relationship between supportive environments and other positive behaviors and learning (Chiu et al., 2017; Morales-Murillo et al., 2020). Due to its malleability and the vital role it has for learning and development, child engagement is often considered an outcome in early childhood interventions (Shafer & Wanless, 2022).

While there seems to be agreement about the importance of engagement for children’s learning and development, it is a challenge to precisely define the concept of engagement in research. In everyday language, the word engagement has several meanings, including agreeing to do something or to be somewhere, being interested in something, or having contact with someone (Cambridge University Press, n.d.-a). Notably, the word involvement has a similar meaning (e.g., “the act or process of taking part in something”). Engagement and involvement are words often used interchangeably in everyday language (Cambridge University Press, n.d.-b). Involvement is a relevant concept in the International Classification of Functioning, Disability and Health (ICF; WHO, 2001), where it is defined as “taking part, being included or engaged in an area of life, being accepted, or having access to needed resources.” Circular definitions are also used in research wherein engagement is defined as an aspect of involvement and vice versa (Boekaerts, 2016; Imms et al., 2017). Steinhardt et al. (2022) investigated the use of both terms across several research fields and concluded that the definitions of both involvement and engagement vary, both terms have been used to describe the experience of participation, and there is an overlap in their meaning, especially in the field of educational research where involvement seems to correspond to internal aspects of school engagement. This indicates a potential overlap in the understandings of child engagement and child involvement in research set in ECEC and suggests that both terms should be addressed in a review of children’s engagement.

Research on children’s engagement provides various definitions of the concept engagement (Finn & Zimmer, 2012), and measuring engagement in a way that is comprehensive, objective, and reliable has proven to be difficult, both in school and ECEC settings (Fredricks et al., 2016; Fredricks & McColskey, 2012). Whereas the concept and measurement of engagement in school-aged children have been broadly discussed and problematized (Eccles, 2016; Finn & Zimmer, 2012; Fredricks et al., 2004; Fredricks & McColskey, 2012), the same cannot be said for the engagement of younger children in ECEC settings. Based on a review of school engagement literature, Fredricks et al. (2004) suggested a framework of school engagement where engagement consists of behavioral, cognitive, and emotional aspects. Whereas behavioral engagement can be observed, internal experiences of emotional and cognitive engagement are best assessed through self-reports (Fredricks et al., 2004; Steinhardt et al., 2022). Recently, some attention has been given to investigating the measurement of child engagement in early childhood within the school engagement framework (Shafer & Wanless, 2022). However, it is arguable if this framework is appropriate for young children. In school engagement questionnaires, older children can report about their own experiences, thoughts, and feelings, whereas methods suitable for assessing internal experiences of young children are rare and child-centered techniques for data collection typically rely on qualitative approaches (Conijn et al., 2020; Coyne et al., 2021). Proxy ratings of child engagement rely on observable indicators (Snow et al., 2005), so it is likely that discrepancies between emotional, cognitive, and behavioral aspects of engagement are not obvious in a population of young children. The concept of child engagement in ECEC might be conceptualized and measured differently than school engagement. To the authors’ knowledge, no research offers a thorough explorative review of conceptualizations and measurement of child engagement in ECEC settings.

Engagement is a complex experience and is often described as multidimensional (Fredricks et al., 2004; Wang et al., 2019). It is not surprising that a variety of approaches to conceptualizing and measuring engagement exists as these might be necessary to study all the relevant aspects of child engagement and explore the concept in depth. At the same time, variations in the understanding of engagement also make it challenging to summarize and compare research that involves this “difficult to measure” construct (Whiteneck & Dijkers, 2009).

Why Is It Difficult to Measure Child Engagement?

It is expected that child engagement to some extent will be manifested in behavior and that adult observers can identify if a child is engaged or not (Kishida & Kemp, 2006). Still, child engagement is also a personal experience and external ratings of engagement may be seen as problematic since they can only rely on overt and objective aspects of engagement (Snow et al., 2005). Self-reports about child engagement are common in studies including school-aged children, but it is more challenging to obtain self-reports from young children attending ECEC settings since they have problems understanding complex or abstract item phrasing, to report about emotions, and they display response bias in terms of extreme responses and nondifferentiation (Conijn et al., 2020; Coombes et al., 2021; Varni et al., 2007).

Another problematic aspect of child engagement measurement is the diversity of indicators of engagement which may be related both to the child and to the context of the child. Individual differences between children will reflect on how they experience and express engagement in activities. In addition, young children undergo rapid developmental changes, including motoric, communication, and self-regulation skills (Mascolo & Fischer, 2015; McClelland et al., 2015). Developmental changes may reflect on how children’s engagement in ECEC is experienced and manifested, which is likely to present a challenge for objective measurement of child engagement and comparing engagement between children.

Expressions of engagement can also be contextually sensitive. The context of ECEC settings is less structured in comparison to school settings. In ECEC, activities are dynamic and can include gross motor activities, pretend play, construction and arts, and more traditional didactic activities (Coelho et al., 2021). Defining and measuring child engagement across such broad activities can be difficult. In addition, there is great diversity in how ECEC settings are organized across countries (Coelho et al., 2021), which further complicates the comparison and summary of findings about child engagement across different educational and cultural settings.

Lastly, researchers can have various, often implicit, assumptions about the temporality of engagement, which may reflect on how they choose to measure engagement. In the school engagement literature, engagement can be conceptualized as a momentary, transient state of immersion within an activity (Symonds et al., 2019), e.g., as a state of being caught and held (Skinner et al., 2009). Momentary or state engagement can be assessed by observations and experience sampling methods (Aguiar & McWilliam, 2013; Fredricks et al., 2004). Alternatively, school engagement can also be seen as a stable tendency of a child to behave in a certain way, and children can be described as highly engaged, disengaged, or passively engaged (Skinner et al., 2009). Questionnaires asking about how a child typically behaves can be used to measure engagement as a relatively stable pattern of behaviors (Fredricks et al., 2004; Skinner et al., 2009). We anticipate that diversity in the temporal assumptions about engagement and their impact on the measurement of child engagement is not exclusive to school settings but will also be present in research of child engagement in ECEC settings.

Altogether, the multifaceted nature of child engagement and variations in its measurement create a complex landscape for researchers attempting to objectively quantify child engagement and to synthesize the vast research about engagement of young children in ECEC. These challenges underscore a pressing need for a systematic investigation of the conceptualization and measurement of child engagement in ECEC.

Study Purpose

To improve the conceptual clarity and facilitate child engagement research and studies of interventions that are aimed at improving young children’s engagement, we see a need for a comprehensive review of conceptualizations and measures of children’s engagement in ECEC settings. Since we are interested in operationalizations of child engagement, this literature review will be limited to quantitative studies. The purpose of the current study is thus to explore the measurement of child engagement in ECEC and the concept of engagement the measurements are based on.

A scoping review method was chosen as our aim was to map a wide range of literature, to summarize research in the field of measurements of child engagement and to examine the extent, range, and nature of the research in this field (Arksey & O’Malley, 2005; Levac et al., 2010; Munn et al., 2018). Moreover, a scoping review methodology is suitable when a research area has not been investigated before or is complex (Arksey & O’Malley, 2005) which seem to be the case in the field of measurements of engagement for young children.

With this explorative comprehensive review, we aim to answer these research questions:

  1. 1.

    How is children’s engagement in ECEC settings conceptualized in quantitative empirical studies?

  2. 2.

    What child engagement measures have been used in studies targeting ECEC settings?

  3. 3.

    What are the characteristics of the identified child engagement measures?

Method

The review was developed in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. Figure 1 presents a PRISMA flow diagram (Page et al., 2021), including information about the screening process and exclusion criteria. This review was pre-registered at Open Science Framework (Ritoša et al., 2021).

Fig. 1
figure 1

A flowchart describing the search and screening procedure

Full size image

Search

The search was conducted in March 2021 in the following databases/platforms: ERIC (EBSCOHost), PsycInfo (ProQuest), Scopus (Elsevier), and Web of Science Core Collection (SCI-EXPANDED, SSCI, A&HCI, CPCI-S, CPCI-SSH, and ESCI). Search strings for all the databases are presented in Appendix.

The key components of our systematic search strategy were population, context, and concept (phenomenon). The population of interest was children attending ECEC. Contexts of interest were any non-parental childcare and early education setting that occurred before the start of first grade in school. The concept (phenomenon) of interest was child engagement. Since the terms engagement and involvement are often used interchangeably (Boekaerts, 2016; Fredricks et al., 2004; Imms et al., 2017; Steinhardt et al., 2022), our search and inclusion criteria encompassed both terms. To focus the search more precisely on empirical studies where engagement was measured with quantitative methods, a block of search terms related to designs using quantitative measurement was added to the search string.

The specific search strategies were created by a research librarian with expertise in literature reviews. A reference set of key articles were identified before the search process and used to generate search terms and to test the effectiveness of the strategies in each database. The ERIC strategy was developed with input from the project team, then peer-reviewed by a second librarian, not otherwise associated with the project. After the ERIC strategy was finalized, it was adapted to the syntax and subject headings of the other databases/platforms by the librarian. For our search strategy, we aimed to find as many relevant records as possible. No publication date or language limits were imposed on the search. The search resulted in 10843 hits.

Title and Abstract Screening

Records found during the search phase were exported to a reference management software (EndNote) to enable the identification and removal of duplicates (Bramer et al., 2016). After deduplication, there were 5965 hits, and after excluding articles published in languages not spoken by any of the authors (n = 148), review papers (n = 47), and publications lacking peer-review (n = 4), the final number of articles was 5766. Records were then exported to and screened using Rayyan, a web-based application for systematic reviews (Ouzzani et al., 2016). Studies in English, Bosnian, Croatian, Danish, Norwegian, Serbian, and Swedish languages that could be translated adequately by the reviewers were included. The inclusion and exclusion criteria are presented in Table 1.

Table 1 Inclusion and exclusion criteria
Full size table

Before the formal title and abstract screening process, a calibration exercise was undertaken to check for reliability of the inclusion and exclusion criteria. One hundred articles were screened in pairs of reviewers, and the inter-rater agreement was 97%.

In total, six reviewers participated in the screening of the titles and abstracts. To check for, and to sort out any discrepancies in our screening approaches that happened over the course of the screening, inter-rater agreement was checked on a sample of 100 randomly chosen articles at the middle and at the end of the screening phase. Inter-rater agreement was high on both occasions (98% and 99%, respectively).

Out of 5766 articles screened on the title and abstract level, 335 articles were included in the full text screening and 370 were marked as “maybe” since the decision about inclusion could not be made solely on the titles and abstract level. All these articles were included to be screened in full text, in total 705 articles.

Full Text Screening

On the full text screening of included articles, two reviewers screened each article independently, except for two articles written in a language only one author spoke fluently. Frequent discussion meetings were held to avoid observer drift. When deciding to exclude articles, reviewers noted down the most salient reason for exclusion. Out of 705 articles screened in full text, there were 99 instances of conflicts where one reviewer decided to include the article and the other reviewer decided to exclude it. An additional 55 articles had one or both reviewers answer with “maybe.” This corresponds to an inter-rater agreement level for exclusion and inclusion of 78% and a weighted Kappa coefficient of 0.64, indicating substantial agreement. Both “conflicts” and “maybes” were resolved by discussion and consensus between the pairs of reviewers. Most of the conflictual decisions (74%) were resolved by excluding the article in question, indicating that many of the “conflicts” and “maybes” during this stage of screening could be explained by reviewers being generous with including articles. All the included articles were written in English.

Data Extraction, Analysis, and Synthesis

Data extraction was done using a Microsoft Excel spreadsheet. The complete data extraction form can be found at the OSF pre-registration site (Ritoša et al., 2021). Five reviewers extracted the data and instructions were included in the category titles to inform reviewers whether they should answer with “yes” or “no,” or by transcribing text from the articles. Difficult extraction decisions were discussed between the review team. A data-extraction check was performed on a subset of 50 randomly chosen articles to detect discrepancies in the extraction process. Data were extracted about the identified studies, child engagement conceptualization, theoretical framework, study design, method, and population, measurement tool, and child engagement operationalization.

Data about the characteristics of the identified studies, e.g., study design (validation, intervention, correlational, descriptive), setting, and population, were summarized first to provide an overview of the quantitative studies of child engagement in ECEC settings. This was followed by a summative content analysis (Hsieh & Shannon, 2005) of the definitions of child engagement and involvement provided in the identified studies, to answer the first research question. The most cited definitions of child engagement were reported, as well as a narrative summary of the other identified definitions. Theories, theoretical frameworks, and models of child engagement that were explicitly mentioned in more than one study were also reported. The summative content analysis was performed by one author.

To answer the second and third research questions, the analysis moved from the study level to the measurement level. Since some studies used more than one measurement of engagement, data were extracted about each operationalization of child engagement identified in the included studies. If authors had validated or used a cited measure of engagement, this information was extracted. An existing citation indicated, but did not guarantee, that the measure was validated in earlier publications. We termed these measures established, meaning that the measure was published, was available to researchers, and was subsequently used in the identified studies. Accordingly, we use the term unestablished measures to refer to operationalizations of child engagement described in the identified studies that did not appear to have undergone any validation procedure and were not accompanied by a citation. A list of all the identified established measures of child engagement is provided in the results to answer the second research question. A list of references to these measures is available as supplementary material. To answer the third research question, additional information was extracted about the operationalizations of child engagement described in the method sections of the studies, including the measurement method (observation/survey/other), engagement informant/rater, and, when relevant, details about observations (duration, frequency, and reliance on video footage), dimensions of engagement (reported subscales, factors, or dimensions) and if child engagement was only a subscale in a measure.

Results

Studies of Children’s Engagement in ECEC Settings

A total of 286 studies of young children’s engagement in ECEC settings were identified. The studies were published between 1970 and 2021 and were conducted in 29 different countries. An overview of the study locations and time of publication is presented in Fig. 2. A complete reference list to the 286 studies included in the synthesis is available as supplementary material.

Fig. 2
figure 2

An overview of the included studies

Full size image

Different terminology was used in the studies to describe the ECEC settings, and some studies mentioned several settings, leading to the number of extracted ECEC settings being higher than the number of included studies. The most common ECEC settings were preschools (n = 122), kindergartens (n = 80), Head Start classrooms (n = 42), pre-K classes (n = 11), and nurseries (n = 9), while 84 studies used other unique terminology to describe the ECEC programs where the child engagement measurement took place, to indicate other private, university affiliated, non-profit, or special-program organizations that provided care and/or education services to young children before starting school. Seventeen studies included both school-aged children and children attending ECEC and were included as they reported the findings about child engagement separately for children attending ECEC. The great majority of studies (n = 232, or 80%) included children between three and six years of age. Only six studies included infants under one year of age, and only eight studies included children aged seven or older.

An overview of the study characteristics is presented in Table 2. Child engagement was usually studied as an outcome (n = 231, or 81%). More rarely, child engagement had a role of a mediator (n = 18, or 6%) or a predictor (n = 26, or 9%) of other outcomes, such as children’s language skills and other classroom behaviors, or of the support the children received by their teachers.

Table 2 An overview of the methodological characteristics of the included studies
Full size table

Conceptualizations of Child Engagement

Out of the 286 identified studies, only 105 studies (37%) provided a definition of child engagement or involvement, or both, in their introduction/background section; 90 studies provided a definition of child engagement, 14 studies included a definition of child involvement, and one study defined both child engagement and involvement. Fifteen of the definitions of child engagement included the word involvement (e.g., engagement is behavioral involvement). Four of the definitions of child involvement included the word engagement (e.g., a degree to which children are engaged).

Child Engagement in Academic Activities

While all studies were set in an ECEC setting, a large group of studies defined child engagement in relation to academic (n = 38) or literacy (n = 5) activities, as visible in the child engagement definition, or stated in the study purpose. The terms academic engagement, school engagement, learning engagement, and student engagement were sometimes used interchangeably, and are here referred to as child engagement in academic activities. The focus on child engagement in academic activities was predominant in US kindergarten settings.

Twelve of these studies relied on a definition of school engagement proposed by Fredricks et al. (2004), where engagement includes a behavioral, emotional, and cognitive dimension. For example, Searle et al. (2013, p. 1113) define engagement as “children’s behavioral involvement, cognitive investment, and emotional commitment during learning activities.” In an additional 12 studies, child engagement was conceptualized as containing only behavioral aspects, e.g., as a “a composite of specific classroom behaviors, including attending to the teacher, participating in lessons, and asking and answering questions about the topic at hand” (Clarke et al., 2004, p.133). The remaining studies (n = 21) conceptualized child engagement in academic or literacy activities as a combination of behavioral, and social, cognitive, or emotional aspects, or as more general interactions with peers, teachers, tasks, or activities.

Examples of behaviors included in the definitions of child engagement in academic activities were effort, attention, focus, persistence, proactive behavior, participation in lessons, following instructions, completing assignments, listening carefully, meeting deadlines, manipulating instruction materials, asking and answering questions, not interrupting, and not skipping classes. As can be seen, some of these behaviors do include a social aspect of adhering to rules, as well as indicators of cognitive investment. Moreover, words related to emotions, such as curiosity, dedication, optimism, passion, and enthusiasm, were sometimes included in the definitions of child engagement in academic activities, but they were often accompanied by a verb (e.g., the quality of emotion and dedication that learners demonstrate, the degree of optimism and passion the students show), indicating the interest in observable behaviors. Children’s internal, covert, and potentially more stable characteristics, such as attitudes towards schooling and identification with the academic environment, were included in four definitions of child engagement in academic activities. Five definitions included a child’s ability or capacity, such as cognitive control, social competence, ability to follow instructions, to self-organize, and to complete tasks, or as a reflection of the fit between the child’s abilities and the environmental demands. In sum, the conceptualizations of child engagement in academic activities were varied. They mostly included behavioral indicators, but internal aspects of engagement were also represented as well as children’s stable characteristics.

General Child Engagement

Thirty-seven studies defined the concept of child engagement across various activities, hereafter referred to as general child engagement. The most popular definition (n = 32) of general child engagement was that by McWilliam and Bailey (McWilliam & Bailey Jr., 1995), who defined child engagement as the amount of time children spend interacting with the environment in a developmentally and contextually appropriate manner. When using this definition, environment is sometimes further elaborated into including social and physical aspects, or, more specifically, adults, peers, materials, and symbols.

Examples of general child engagement consisted exclusively of descriptions of behaviors. For example, Ponitz et al. (2009, p. 104), defined child engagement as “correspondence between the child’s observable behavior and the demands of the situation, including attending to and completing tasks responsibly, following rules and instructions, persisting in the face of difficulty, and exercising self-control.” Even when a cognitive aspect of child engagement was included in the conceptualization, this was done in a way so that the behavior was in focus (e.g., engagement refers to observed attention, watching in a concentrated way, or listening attentively). In sum, general child engagement is dominantly associated with behaviors of the child, and interactions between the child and the context the child is in.

Child Involvement

Most of the studies providing definitions of child involvement used variations of the definition by Laevers (n = 10), the author of the Leuven Involvement Scales (see Table 4 and Supplementary file for references). For example, Bjørgen (2015), p. 307) relied on Laevers’ definition of involvement stating that “involvement is a quality of human activity, characterized by concentration and persistence, a high level of motivation, intense perceptions and experiencing of meaning, a strong flow of energy and a high degree of satisfaction.” Overall, our analysis showed that child involvement was usually described as a process, activity, or quality of a child’s activity, characterized by strong intrinsic motivation, exploratory drive, deep concentration, focus, persistence, positive energy, high degree of satisfaction, and amusement.

Other Conceptualizations

The other identified conceptualizations included concepts of task engagement (n = 5), characterized by active participation and behavior that is appropriate given the demands of the task, joint engagement (n = 2), defined as interactions or as the ability of a child to maintain social interactions, play engagement (n = 1), characterized by cognitive aspects of focused attention and concentration of a playing child, peer engagement (n = 1), defined as observable displays of positive and negative social behaviors, and negative engagement (n = 1) described in terms of undesirable interactions with teachers, peers and tasks. Social aspects of engagement were represented in conceptualizations of general child engagement and engagement in academic activities, but surprisingly, no explicit definition of social engagement was identified.

Theoretical Frameworks

Explicit theoretical frameworks were rarely provided in the identified studies (14%; see Table 2). Where theoretical frameworks were mentioned, both child engagement and involvement were linked to proximal processes defined in the bioecological model of human development (Bronfenbrenner & Ceci, 1994; Bronfenbrenner & Morris, 2006). Child engagement in academic activities was sometimes studied within self-determination (Deci et al., 1991) and self-regulation (Zimmerman, 1990) theories. Csikszentmihalyi’s (1990) theory of flow, and Vygotsky’s (1943/Vygotsky, 1987) theory of proximal development were more commonly mentioned in relation to child involvement than child engagement. Attachment theory (Bowlby, 1979) was discussed in studies investigating the teachers’ role for child engagement or social interaction in relation to child engagement.

Measures of Children’s Engagement in ECEC Settings

Out of the 286 studies included in the current review, 21 used two ways to measure child engagement, and in two studies, child engagement was measured in three different ways. Altogether, 311 engagement measurements were included in the analysis. All but two of the identified measures could be categorized as observations or survey measures of child engagement (see Table 3).

Table 3 Observational and survey measures of children’s engagement in ECEC
Full size table

Of the identified engagement measurements, 202 relied on established measures of child engagement. Some established measures were used in two or more studies, and in total, 77 unique established measures of young children’s engagement in ECEC settings were identified. A list of these measures is presented in Table 4. A list of references to these measures, as cited in the identified studies, is available as supplementary material. In Table 4, the child engagement measures are grouped based on the measured concept (e.g., engagement, involvement, academic engagement, social engagement, task engagement, negative engagement, joint engagement, engagement in reading, engagement in play, engagement in music activities, and engagement in computer activities), as reported in the engagement operationalization in the identified articles, usually in the method sections. Five of the established measurement tools were general observational protocols and software programs that were adapted to include the variable child engagement but were not specifically developed for the purpose of measuring child engagement.

Table 4 Established measures of children’s engagement in ECEC grouped by the measured concept
Full size table

Observations of Child Engagement

The most common method of measuring child engagement in ECEC was observations by study authors, research assistants, trained observers, or other external observers not previously familiar with the children (see Table 3). Surprisingly, more than one-third of the observational studies used only general terms, such as, “observer,” “coder,” or “data collector” to describe who was rating the children’s engagement without providing any additional information about the observers’ expertise or training.

The majority of observations relied on observations over more than one day (n = 147). Direct coding of observations (on-site, n = 159) was twice as common as relying on video recordings (n = 76). The length of the observations ranged from three seconds up to full session lengths expressed in minutes or even hours. The most common observation time frames were 10- or 15-second time sampling in 10-minute-long observations. Short observations were usually repeated and averaged to obtain child engagement scores for individual children. In 55 observational studies, the length and/or the frequency of observation were not reported (23%).

Most of the observational measures could be cited, and 51 unique established observational measures of child engagement were identified in the current review (for a complete list of the identified established measures of children’s engagement in ECEC, see Table 4). Most of the observational measures included child engagement as a subscale alongside other constructs. About one-third of the established observational measures were explicitly described as requiring some training of the observers (n = 16). Eight observational measures measured various dimensions of the child engagement concept (e.g., engagement with teachers, peers, and tasks; active and passive engagement, see Table 4). Four observational measures were intended for measuring group engagement, and three observational measures were primarily intended for use in a school population but used in ECEC settings in the identified studies.

In observations, child engagement was typically rated on a scale from low to high (n = 101) or using the dichotomy “engaged versus non-engaged” (n = 89) which allowed for the calculation of duration or proportion of time the child was engaged during the observation. Descriptive categories of child engagement indicating which activity the child was engaged in during an observation, rather than a level of the child’s engagement, were used less often (n = 40). Observations relying on behavior frequency counts (n = 4) and teacher’s activity logs to describe the engagement of the whole group of children (n = 1) were rare. The established observational measures were the most likely to use ordinal scales for the child engagement ratings (62%), whereas in most of the unestablished observational measures identified, child engagement was rated on a dichotomous scale “engaged versus non-engaged” (67%). The operationalizations of child engagement in the unestablished observational measures varied from more objective behavioral indicators, such as, the child’s eye gaze, location, body orientation, talking, or being in physical contact with specific materials, to more ambiguous behavioral indicators of attention and interest, which often included value-based descriptors of behaviors such as “appropriate,” “relevant,” or “purposeful.”

Surveys of Child Engagement

Surveys of children’s engagement were the second most common method of measuring child engagement in ECEC. Most surveys were answered by the children’s teachers, and more seldom by the children’s parents or caregivers, the children themselves, or by both teachers and parents using the same survey (see Table 3).

Most of the survey measures could be cited, and 26 unique established survey measures of children’s engagement in ECEC were identified in the current review (see Table 4). About half of the established surveys measured child engagement as a subscale, and in less than half of the survey measures, child engagement had dimensions (e.g., engagement with teachers, peers, and tasks; emotional and behavioral engagement; school liking and school avoidance, see Table 4). About one-third of the survey measures was primarily intended for measuring child engagement of school-aged children but could also be used for children in ECEC, as they were in the identified studies. As seen from Table 4, teacher surveys were the dominant method for the measurement of child engagement in academic activities.

Other Child Engagement Measurement Methods

Two measurement methods could not be categorized as either observations or surveys. One study used a sociometric method where peer nominations of likability were an indicator of social engagement (Altman et al., 2020). The other study used teacher categorization of children into three groups based on their engagement levels (Almeqdad et al., 2016).

General Discussion

This scoping review is aimed at identifying the existing quantitative studies and measures of young children’s engagement in ECEC settings and at investigating variations in how engagement has been conceptualized and operationalized in these studies. A total of 311 reported measurement methods were extracted across the 286 identified studies published in the last five decades. The large number of identified studies indicates that children’s engagement in ECEC settings is of great interest to researchers. This may not be surprising considering the importance of early experiences for learning and development (Institute of Medicine and National Research Council, 2000). On the other hand, the actual diversity and the number of unique established measures identified, as well as the number of studies relying on unestablished measures of child engagement, surpassed our expectations. Although 77 unique established measures of children’s engagement in ECEC settings were identified, more than one-third of the identified studies relied on unestablished measures of engagement created by the study authors. As seen from Table 4, only a handful of the established measures have been used repeatedly or were revised over time (e.g., ICER developed by Kishida and Kemp in 2006 and ICER-R revised by Kishida et al. in 2008, see Table 4). The great majority of the identified established engagement measures were used in only one or a few studies. The abundance of approaches to child engagement measurement might indicate a problematic fragmentation in this field of research and makes it challenging to systematically summarize the findings in research on young children’s engagement in ECEC. Still, it is hard to imagine a universal measure of child engagement that would fit ECEC settings across the world, and variations in child engagement measurement can partly be justified as a necessary adaptation to variations in research interests, study designs, assumptions about the temporality of child engagement, or different educational and cultural contexts of the identified studies.

Interest in the engagement concept seems to have emerged in parallel within different research traditions. Variations in the conceptualization and measurement of child engagement may to some degree reflect different research paradigms and priorities that have shaped research traditions of child engagement in different parts of the world. For instance, the Leuven scales of involvement (LIS-YC, see Table 4), developed in Europe and based on Vygotsky’s sociocultural theory (1987) and the concept of flow (Csikszentmihalyi, 1990), were developed in an educational context with the ultimate goal of deep learning. At the same point in time, the Children’s Engagement Questionnaire (CEQ; McWilliam, 1991), developed in the USA, was intended for use in the context of early childhood intervention. These examples, while not necessarily indicative of the current landscape, provide valuable insights into the diverse perspectives that have shaped today’s research practice.

The different cultural and educational contexts may reflect the use of the concepts of child involvement versus child engagement. We noticed an overlap in meaning between the terms engagement and involvement and no consistencies in definitions using one of the terms to define the other. When comparing conceptualizations of involvement and engagement, an emphasis on behaviors is obvious in definitions of engagement, while involvement is typically described as an internal experience. This corresponds to other findings about engagement and involvement as aspects of participation (Steinhardt et al., 2022). Interestingly, most measures of involvement are observations (see Table 4). Looking into operationalizations of both concepts, involvement is typically approached as a general state, while engagement is more often (but not always!) contextualized. Some measures of engagement are intended for a specific activity (e.g., play engagement and engagement in reading) while some contain subscales differentiating between different contexts the child can engage with (e.g., engagement with peers, teachers, and tasks).

Other cultural differences may be related to the organization of ECEC and children’s age when attending ECEC and could be additional reasons for the variation in the measurement of child engagement. Child engagement is often considered context-dependent, and it may manifest differently depending on the context and the activity of interest. Thus, the different focus on academic versus play-related or social content in ECEC across countries (e.g., Coelho et al., 2021) is likely to reflect on how the researchers conceptualized and measured child engagement.

Furthermore, the variations in the observational schedules, such as duration and number of observations in measuring child engagement, reflect a variety in the conceptualizations of child engagement on a state-trait continuum (Yoder et al., 2018). The specific concept of academic engagement stood out as more commonly measured by teacher surveys, indicating that a certain stability and consistency in children’s academic engagement were expected. Notably, only a few definitions of academic engagement described it as a stable trait of a child, indicating that conceptualizations do not always match the measurement approach. Based on the findings of this scoping review, recommendations for future research on child engagement in ECEC will be presented next.

Recommendations for Research on Child Engagement in ECEC

Clarifying the Concept of Child Engagement in ECEC

Child engagement was rarely studied within an explicit theoretical framework or a model of engagement in the identified studies. Additionally, only one-third of the studies offered an explicit definition of child engagement although it was a central research concept, measured in all the identified studies. The commonly cited definition of child engagement by McWilliam and Bailey Jr. (1995) is very broad. What is considered developmentally and contextually appropriate interactions between a child and the environment is expected to vary between children and cultural contexts. The general lack of explicit definitions of child engagement indicates that researchers presume the meaning of the concept ‘engagement’ is clear.

On the other hand, the identified variations in the definitions and measurements of child engagement indicate a lack of consensus on what child engagement in ECEC is. Research differed in how much importance was given to emotional and social aspects of child engagement, in comparison to reliance solely on behaviors. Self-reliance and self-determination aspects of child engagement were only rarely mentioned in the identified studies but are also worth considering (Kohn, 2022). Other variations reflect differences in norms and values of different educational contexts, for example, how much focus there is on academic activities and tasks, or on play activities and social connectedness. Additionally, some of the variations in the definitions and measurements of child engagement seem related with researchers’ assumptions about the generalizability and temporality of child engagement. For example, whether engagement is a temporary state co-created in the interaction between the child and the environment, or a stable behavioral tendency of the child, and, on a related note, whether engagement is considered to be a general tendency or if it depends on the context or the activity domain.

To empirically study child engagement, one needs to define it in concrete and precise terms (Bhattacherjee, 2019). This review shows that the meaning of the concept cannot be assumed as researchers have different understandings of what constitutes engagement. Whereas there is value in having multiple approaches when studying complex phenomena such as child engagement, empirical studies of child engagement should provide an explicit definition of engagement that is relevant for the context of their study, which can serve as a basis for operationalization of the same construct.

Adequate Operationalization of Child Engagement and Description of the Methodology

When operationalizing child engagement, researchers should be careful to ensure that conceptual and operational definitions of child engagement are aligned (White et al., 2022). Our findings show that when defining child engagement in ECEC, the predominant focus lies on behavioral aspects of engagement. Observations are a suitable method for the assessment of dynamic child behaviors (Neyer, 2006). On the other hand, psychometric approaches have been deemed adequate for evaluating stable latent traits (Wright & Feinstein, 1992). Notably, engagement in academic activities is sometimes portrayed as a stable ability, competence, or a trait, contrasting the prevailing characterization of young children’s engagement predominantly as observable behaviors and interactions. Questionnaires that inquire about a child’s typical behavior may be suitable to measure stable aspects of engagement, but reliance on questionnaires to gauge dynamic aspects of child engagement remains doubtful.

In conclusion, researchers of child engagement face the imperative to either adjust their definitions of engagement and to acknowledge an anticipated consistency in a child’s behavior over time if they choose to measure child engagement in a manner that implies stability. Alternatively, if child engagement is understood as a dynamic state contingent on interactions between a child and their environment, it is ill-suited for measures that assume latency and consistency of engagement. To further advance this field, an exploration of the psychometric dimensions of general child engagement questionnaires, including reliability over time, internal consistency, and discriminant validity, would be of immense value (Hughes, 2018).

Besides validity concerns, researchers must consider the broader context in which child engagement is measured as the implementation of measurement should be feasible and ethical (Woodhead & Faulkner, 2008). Across the studies, we have observed a discernible trade-off between the meaningfulness of measures and their objectivity and practicality for use, a trend previously discussed in intervention research (Castro & Yasui, 2017). For instance, in some of the identified studies, child engagement was operationalized as easily observable behaviors, such as eye gaze and physical contact with certain materials. Such behavioral indicators of child engagement can be easy to objectively measure, but at the same time, they can be accused of being reductionist and simplifying the complex construct of child engagement to simple behaviors (Bhaskar & Danermark, 2006). On the other hand, complex descriptions of behavioral indicators that acknowledge emotional and cognitive aspects of child engagement and rely on less tangible behavioral indicators can easily suffer from low objectivity and reliability (Dijkers, 2010). Increased awareness and discussion of these measurement challenges, as well as improved reporting on the methodology used and the operationalization of child engagement, are welcome to improve the research about child engagement in ECEC.

Many observational studies rated child engagement on a dichotomy of “engaged versus non-engaged.” Even some established observation tools and software programs measuring child engagement were described only in terms of duration and number of observations but lacked an operational definition of child engagement (measures 72–76 presented in Table 4). In this way, it is up to the observer to rely on their own definition of engagement, and it is questionable how objective such measures are. No comparisons of these single-item measures in relation to more complex measures of child engagement were identified in this review. It should be investigated if single-item measures asking whether the child is engaged or not correlate with more complex measures which tap into several aspects of engagement.

Additionally, several observational studies failed to report about the duration of the observations, the number of observations, or the time sampling information, when describing the data collection method. Surprisingly, many observational studies provided only limited descriptions of who performed the observations and rated children’s engagement. Such basic methodological information should be described in the research to allow for replicability but also to inform about the quality of the measurement (Portell et al., 2015).

Relying on Validated Measures of Child Engagement

Various established measures of child engagement have been identified in this review, as well as several validation studies of child engagement measures. It is important to rely on measures that are objective, reliable, culturally relevant, and measure the construct they are intended to measure. Psychometric evaluation of measurements can investigate these aspects (Mokkink et al., 2010). However, more than one-third of the identified studies, and most of the intervention studies, measured engagement with measures that were unestablished, indicating that the measure had not been validated through conventional psychometric methods. This is common in other research fields as well (Bentler & Kramer, 2000; Fayers & Machin, 2007; Karlsson & Gustafsson, 2022) and an issue to be aware of when choosing to measure social constructs. It should, however, be stressed that the presence of a citation, as in the 77 identified established measures in this review, does not imply that the measure has undergone a rigorous validation process, only that more information about the measure could be found elsewhere (see the Refence list). Unvalidated measures are often based solely on face validity, and while commonly relied on, Kazdin (2021) stresses that face validity alone is not an adequate criterion for a measure. Measures relying on face validity alone could be measuring a single construct of child engagement, several aspects of it, or capture some other constructs besides, or instead of, child engagement.

As mentioned earlier in this article, engagement is a common spoken word that has many connotations. In educational practice the word engagement has a high face-value and is easy to comprehend. However, this should not diminish the importance of and need for using measures of child engagement that have acceptable psychometric properties. When choosing a measurement of children’s engagement, it is therefore of importance to be aware of if the measurement is validated (both the original version and eventual translations) and to use measures that show good measurement properties. If this is not the case, it is recommended to conduct a validation study prior to using a measure in other empirical studies (Fayers & Machin, 2007).

With this review, identifying 77 established measures, we aim to support researchers in choosing, and possibly further validating, existing measures of child engagement, and we hope to see a decreasing trend in reliance upon improvised unestablished measures.

Acknowledging the Differences Between Child Engagement in ECEC-Settings and Older Children’s Engagement in School

The prevalent framework in school engagement research views child engagement as encompassing emotional, cognitive, and behavioral facets (Fredricks et al., 2004). It is not surprising that this framework was also represented in the studies of young children’s engagement in academic activities in ECEC. Notably, the identified measures that included emotional, cognitive, and behavioral aspects of child engagement as distinct dimensions were exclusively used to measure engagement in academic activities. These measures were developed for school engagement measurement, and in the identified studies, they were used mainly in kindergarten and Head Start classrooms with children older than four years of age. On the other hand, child engagement measures developed specifically for the ECEC population primarily capture overt behaviors and interactions between the child and environment, in accordance with the most common definitions of child engagement (McWilliam & Bailey Jr., 1995). When measuring young children’s engagement in ECEC, as seen from Table 4, it was common to differentiate between simple engaged behaviors also referred to as undifferentiated behavior, passive or unsophisticated engagement, and more complex engagement, also referred to as sophisticated engagement, which included competence, persistence, challenge-seeking, or self-reliance.

Whereas the emphasis on observable behaviors is understandable due to the above discussed practicality of data gathering, it carries certain unfavorable implications. By disregarding the internal experiences of engagement, there is a risk that the engagement of more reserved or introverted children is being overlooked. When the focus is solely on observable behaviors, observers might misinterpret or fail to notice that a child displaying overt behaviors might not necessarily be deeply engaged, while a quiet child might be highly engaged in their own introspective way. To encompass children’s internal experiences and capture a more holistic picture of young children’s engagement, the development of self-report methods tailored for this age group emerges as a promising solution.

Developing Self-Reports for Young Children

Child engagement measurements in school-aged children often rely on self-reports (Fredricks et al., 2004). Only two identified measures relied on children self-reporting about their own engagement in ECEC and were primarily intended for the measurement of academic engagement of school-aged children specifically. The importance of self-reports to capture a holistic picture of children’s engagement has previously been highlighted in the school engagement literature. This is based on the idea that objective measures are ineffective in capturing children’s emotional or cognitive engagement since these dimensions involve experiences which cannot be directly observed (Fredricks & McColskey, 2012). Allowing children’s voices to be heard and taken into consideration about matters that affect them is also stressed by the Convention on the Rights of the Child (United Nations, 1989). This is clearly a potential problem in the field of ECEC given that self-reports are more difficult to collect from younger children (Conijn et al., 2020; Varni et al., 2007). However, research indicates that children as young as 5 years can successfully use simplified self-report pain scales (von Baeyer et al., 2017), report on their health (Almqvist et al., 2006; Coombes et al., 2021), and even on topics more closely related to engagement, such as academic orientation (Ruzek et al., 2020). Development of self-report measures for young children requires specialist knowledge but is nonetheless worthy of consideration. In combination with other measures, self-reports can provide a more complete picture of young children’s engagement in ECEC. Although we believe that the abundance of different measures of child engagement in ECEC presents a challenge for child engagement research, and our general intention is to problematize this, the development of new measures of engagement suitable for young children to report about their own engagement in ECEC could be considered a significant contribution to child engagement research.

Limitations

Although our intention was to identify as many quantitative studies of children’s engagement in ECEC as possible, the applied limitations on language and publication type imply that some relevant studies might have been excluded from this review. Our search was limited to peer-reviewed papers to ensure the quality of included empirical studies (Kelly et al., 2014). We also only included studies where child engagement was quantitatively assessed. Whereas we think that limiting this review on quantitative studies of engagement was appropriate due to our interest in measurement of child engagement, a broader inclusion of literature would have allowed for a more comprehensive investigation of conceptualization of child engagement. Including qualitative research, as well as discussion papers, would provide material for a more detailed and nuanced overview of the understandings of child engagement.

Most studies included in this review were conducted in the USA or in Europe, and all the included studies were written in English language. Our findings indicate that the interest in measurement of child engagement and involvement originated in the early intervention field in the US (McWilliam et al., 1985) and European early-years pedagogy (Laevers, 2000). It is thus not very surprising that most of the included studies come from the USA and Europe. After concluding that child engagement is contextually and culturally defined, it would be valuable to learn how child engagement is viewed in non-western countries. Whereas we recognize that our focus on peer-reviewed papers and language limitations might have introduced a bias against research from non-western countries, we also acknowledge that additional inclusion of non-peer reviewed literature written in other languages could be problematic due to potential translational ambiguity (Schwieter & Prior, 2020), or even a lack of equivalent term when translating child engagement to other languages. Furthermore, we have identified only 25 validation studies, some of which validated the same measures, whereas the total number of the identified established measures was 77. This means that most of the established measures’ manuals and validation procedures were not reported in the identified studies. After looking into secondary references for all established measures, we have concluded that this may be because child engagement measurement tools are often published as manuals in non-peer reviewed publications or presented at conferences and are not available in peer-reviewed journals. We consider that the identified studies provided substantial information about child engagement measurement so that this review has the potential to assist researchers with the choice of method for measuring child engagement in ECEC. A future study might be aimed at gathering papers and reports from the authors of the identified measures and systematically investigating the psychometric properties of the identified child engagement measures in more detail.

Conclusion

As young children’s engagement is often of interest as an indicator of ECEC quality or as an outcome of ECEC interventions, it is important that child engagement is clearly defined and that the measures of engagement are reliable and valid. Our findings demonstrate that many quantitative studies lack an explicit conceptualization of child engagement in ECEC settings. The existing conceptualizations show variations in researcher’s understandings of child engagement, and there is even greater variety in how engagement is operationalized. A relatively high use of unestablished measures and a lack of self-report measures were also noted. Self-reports could be especially valuable for validating observational and proxy measures of engagement and for complementing engagement measures focusing mainly on behaviors.

We encourage researchers who choose to study this appealing but complex concept to be careful and explicit in defining the concept of child engagement, to operationalize it accordingly, describe the measurement procedures in detail, and critically discuss the measurement challenges. Although specific research interests, differences in the ECEC settings and cultural norms likely require flexibility in the child engagement measurement, as well as adaptations of existing measures, we hope that this overview of identified measures of child engagement will assist researchers in choosing appropriate measures and to prevent further confusion and fragmentation in this field of research.