Levels of well-being and involvement of young children in centre-based provision in the Free State Province of South Africa

This article reports on a study that examined the levels of young children’s wellbeing and involvement in centre-based provision (birth to five years) at child, group and setting level1 in Free State, South Africa. The study was funded by the Flemish Department of Education and was executed in collaboration with the Free State Department of Education and the University of Free State. Nineteen settings were included in the study. The average setting was registered for 121 children (with ratio’s varying from 30 to 326 children registered). Foundation Phase students from the 2nd and 3rd year of study at the University of Free State collected data through observation tools designed by the Centre for Experiential Education at Leuven University, Belgium. The core instrument uses the Leuven scales for well-being and involvement. Results of the study indicate that overall scores for well-being and involvement are low, but also that there are huge differences between different groups and settings. Thus, indicating that early childhood education in centre-based provision makes a difference.


Introduction
Early childhood development (ECD) services are receiving considerable attention from policy makers, service providers and families across the globe (Britto, Yoshikwa & Boller, 2011).In South Africa policies, law, care and education programming for ECD are framed in support of children's rights to survival, development, protection and full participation in society (Department of Education, 2001;Biersteker & Dawes, 2008;Ebrahim, 2010).A recent study on addressing exclusion and invisibility in ECD in South Africa draws attention to the value of focusing on the well-being and involvement of children in both centre, and home-based, learning environments (Save the Children & Bernard van Leer, 2010).
These constructs, as defined by the Centre for Experiential Education at Leuven University in Belgium, suggest that well-being involves the basic needs of children being satisfied, and refers to the degree to which children feel at ease.Involvement, on the other hand, refers to developmental processes.This concept is closely linked to Csíkszentmihályi's (1979) concept of flow and to what Bronfenbrenner and Ceci (1994, p. 572) refer to as an intense 'proximal process 2 '.
Taking into account the importance of paying attention to well-being and involvement of young children, the study reported in this article aimed at filling the gaps in knowledge about the levels of young children's well-being and involvement in early childhood settings in Free State, South Africa.
Outlining the lives of children from birth to five years in South Africa Ebrahim (2010), while tracing the trajectory of early childhood care and education in South Africa, argues that a complex relationship exists between the provision for the early years and the political, economic, social and cultural features of the South African society.She notes that the advent of democracy led to greater political will and acknowledgement of the complexity of young children's lives.
In mid-2008, South Africa's total population was estimated at 48.7 million, of whom 18.7 million were children under the age of 18 (Statistics South Africa, 2003-2009).Children therefore constitute 39% of the total population.Nationally 43% of children under five are exposed to some form of ECD stimulation (UNICEF, 2010).Approximately 36,8% of children in the Free State are the most likely to have access to centre-based provision (Statistics South Africa: General Household Survey, July 2009).This type of provision is valuable for shaping the outcomes for children in the developing world.Longitudinal studies tracking children's exposure to centre-based preschool education showed improvement in the number of children entering school, children remaining at school and performance at grade level in the developing world (Engle, Black, Behrman, De Mello, Gertler, Martorell & Young, 2007).
In South Africa centre-based provision, although limited, is viewed as an important service contributing to achieving national and global targets in particular aspects of child well-being (UNICEF, 2010).In addition to concerns of increasing access to centrebased provision for children from birth to five, the quality of programmes is becoming a key focus for investigations on improving child and family outcomes (Human Science Research Council & Early Learning Resource Unit, 2010).
In light of the above, the study reported in this article used well-being and involvement as the indicators to measure quality of ECD centre-based programmes in the Free State.The following questions were posed: What is the average score for well-being and involvement for an individual child, 1.
and how are the scores spread throughout?
Which age group showed the highest/lowest levels of well-being and involvement?2.
What is the average score for well-being and involvement on group level, and how 3.
are the scores spread throughout?
Are there significant differences at group level concerning scores for well-being 4.
and involvement?
Is there a correlation between well-being and involvement? 5. What constitutes quality in ECD? From the point of view of the parent, the curriculum developer, and quality improvement programmes, the question is often answered by expressing expectations with regard to the educational context and the practitioners' actions: the infrastructure, the content of activities, interaction style, the daily organisation and so on.Internationally in research, this is often measured by using environment rating scales, such as ECERS-R or ITERS-R (Harms, Cryer & Clifford, 2003).

Conceptual framework
From the point of view of policy and government, there is a more direct reference to the expected outcomes.With regular assessments the system of care and education is forced to get results.This leads to the development of curricula, e.g. the Free State preGrade-R curriculum (birth to four years), expressing the expected outcomes of quality ECD programmes.

CONTEXT
Means PROCESS

WELL-BEING INVOLVEMENT
In the middle of this stands the child, living and experiencing activities in centrebased provision.How can he, or she, link context and outcome?Internationally, it is often argued that the most important actor in research in ECD -the child -is neglected: "Children are too often statistically invisible.Countries need to regularly collect more high-quality information on children's well-being that is nationally and internationally comparable.Such information is urgently required to regularly and independently monitor child well-being over time at all stages of the child's life cycle" (OECD, 2009).
In this article, we emphasise that what constitutes quality in early childhood settings can be expressed by focusing on two dimensions namely; the degree of well-being and the level of involvement of children, giving children the central place they deserve.This is the missing link between context and outcomes, and expresses what the learning environment is doing to the children here and now.It also tells us, something about the potential impact.

Well-being
A variety of approaches attempt to explain child well-being by utilising language of academic discourse, such as emotional intelligence, health, social competence, academic achievement or emotional development, to name but a few.It can be argued that although there is a growing body of research that works on the understanding of what constitutes well-being, this multi-faceted and complex term can refer to a variety of terms depending on the discipline, concern, culture and/or society in which the child lives (Gill, 2009, p. 2).Fraillon (2004, p. 19) explains that well-being was initially considered to be a component of an overarching construct of health, but argues that more recently health has been regarded as an overarching construct of well-being.Some researchers, for example Carlisle and Hanlon (2007, p. 262-264), consider well-being as entailing happiness and positive emotions.Theorists, like Pollard and Rosenberg (2003), believe that a wide variety of different factors can have an impact on the well-being of children, and, as a result, provide a more holistic definition of well-being.Child well-being can, however, not be viewed in isolation from an educational context, because the context may act as catalysts that can affect the well-being of children negatively (Fraillon, 2004, p. 17).In fact, Mayr and Ulich (2009, p. 45) emphasise that the well-being of children is a central indicator of the quality of educational institutions and processes.Morrow and Mayall (2009, p. 221) argue that in any educational institution the concept of well-being must be interpreted within the context of cultures and the nation.In order to apply the concept of well-being to the South African context, the proposed conceptual framework for this article is situated in the process-orientated approach developed by the Research Centre for Experiential Education, University of Leuven.We have chosen the experiential education model, because it provides a common basis on which the complex and multifaceted lives of children in Free State educational settings can be investigated.It is hoped that this framework will orientate our critique of the various aspects that influence the well-being of learners.We cannot determine how children are doing in a setting if we do not first investigate the wellbeing of children.
In this article, well-being refers to the degree to which children feel at ease, act spontaneously, are open to the world, express inner rest and relaxation, show vitality and self-confidence, and are in touch with their feelings and emotions; thus, indicating that their mental health is secured (Laevers, 2005).

Involvement
Answering the question of emotional well-being (viz.how is the child feeling at this moment?)alone is not enough.Therefore, we included a second indicator: involvement.This is linked to developmental process and urges the adult to set up a challenging environment favouring concentrated, intrinsically motivated activity.
The concept of involvement refers to a dimension of human activity.Involvement is not linked to specific types of behaviour, or to specific levels of development.Csikszentmihayli (1979) speaks of "the state of flow", of which one of the most predominant characteristics is concentration.Involvement is associated with strong motivation, fascination and total implication: there is no distance between person and activity, and no calculation of the possible benefits.Because of that, the perception of time is distorted (time passes by rapidly).Furthermore, there is openness to (relevant) stimuli, while the perceptual and cognitive functioning possesses an intensity that is lacking in activities of another kind.The meanings of words and ideas are felt more strongly and deeply.Further analysis reveals a manifest feeling of satisfaction and a bodily felt stream of positive energy.The "state of flow" is sought actively by people.Young children find it most of the time in play.Of course, one could describe a variety of situations where we can speak of satisfaction combined with intense experience, but not all of them would match our concept of involvement.
Involvement is not the state of arousal easily obtained by the entertainer.The crucial point is that the satisfaction stems from one source; the exploratory drive, which entails the need to get a better grip on reality, the intrinsic interest in how things and people exist in the world, and the urge to experience and figure things out.Only when we succeed in activating the exploratory drive do we get the intrinsic type of involvement, and not just involvement of an emotional or functional kind.Finally, involvement only occurs in the small area in which the activity matches the capabilities of the person, that is in Vygotsky's "zone of proximal development" (Laevers, 1993).
Involvement entails an intense mental activity, during which a person is functioning at the very limits of his or her capabilities, with an energy flow that comes from internal sources.One cannot think of any condition more favourable to real development.Involvement answers the question of how interested a child is in a given setting.

Status of well-being and involvement as indicators
We favour the view of well-being and involvement as process variables, meaning that it looks at the what, here and now, in interaction with the environment and what happens within the child.As a consequence, well-being is not a stable feature of a child, but the result of complex interaction between more or less stable features of the child (for example, positive self-image) and features of the environment (for example, a sensitive teacher or an environment conducive to learning).One can imagine that a child experiencing a lot of small moments of high well-being in centrebased provision will have an impact on the structural features of a child, which, in turn, will impact on the moments of well-being.Well-being and involvement are constructed and reconstructed in constant interaction with the broader living and learning environments.
Observation of well-being and involvement is an effective and unique measure of programme quality.The observation not only discriminates between levels of quality, but also focuses directly on child behaviour and child experience, thus offering a much needed supplement to global and teacher-related measures.The observation of wellbeing (affective aspect) and involvement (cognitive aspect) is both uncomplicated and expeditious, making it an ideal measure for use by licensing personnel, as well as, childcare staff.These two indicators are the minimum one can expect from every ECD programme.A qualitative ECD programme has to succeed in both.Only paying attention to emotional well-being and a positive climate is not enough, while efforts to enhance involvement will only have an impact if children feel at home and are free from emotional constraints.

Research design and methodology
This project was funded by the Flemish Department of Education in Belgium and was executed in collaboration with the Department of Education in the Free State, as well as the University of the Free State in South Africa.The duration of the project extended from November 2007 till June 2010.This article focuses on the data gathered between February and March 2010 by 79 students in selected ECD settings in the Free State.
The Free State Department compiled a list of 100 ECD sites in poor and rural areas of the Free State.From this list a total of 23 settings and 37 groups were visited.In each group a random sample of 20 learners were observed.Because of different reasons (see "compiling the definitive data set"), not all the gathered data could be used.This article is based on the data of 29 groups within 19 settings.
The observations for this study were done by trained 2 nd and 3 rd year Foundation Phase students.Observation was used as a tool to gather data on experiences of children.Literature shows that this technique is particularly useful for the gathering of data on the behaviour of children (McMillan & Schumacher, 2006, p. 9;Mwamwenda, 2004, p. 13).Davin and Van Staden (2005, p. 243) note that this can be an outcome when careful and meticulous observation includes watching attentively and focusing on specific aspects of activities that children are engaged in.With the use of the Leuven Scales for well-being and involvement (Laevers, 2005) the experiences of children by observation was grasped.This data set was used as a direct indicator for the quality of the ECD setting.
Not everything is grasped by observation.To be able to fully understand the ECD reality and to add qualitative information to the observations, an interview with the responsible practitioner was conducted.It enabled the student observer to clarify misunderstandings that arose during the observation and to gain additional information.Student-researchers were also asked to keep a reflective journal, in order to record their experiences.
Prior to the study consent was sought from the Free State Department of Education, the student-researchers and the participants at the sites.Every attempt was made to adhere to the ethical guidelines for involving human subjects in research, particularly regarding informed consent, voluntary participation and the confidentiality of information.

Compiling the definitive data set
In total data of 79 different observers covering 23 settings and 37 different groups was received.In order to get rigid data, a triple selection was undertaken: To check validity and reliability of the observations, all observers were asked • to perform a reliability test (by scoring 15 video excerpts for well-being and 15 for involvement) to check how reliable and valid their scores for well-being and involvement were.Data of observers who did not perform well on the test (this is a result with less than .70correlation, using the ICC-norm whereby the results on the test are compared with the expert scores) were excluded.
Observers who did not conduct a reliability test were excluded.• If several observers observed the same group, an at random selection of data of • only one observer was included.All data of the other observers in the same group were excluded.
The consequence is that the data presented is based on a much smaller sample than initially gathered.Data of 29 observers composed of 19 settings and 29 different groups were retained for analysis.
The smallest setting was registered for 30 children, the biggest one for 326 children.The average setting was registered for 121 children.The adult/child ratio was also registered.In general there was one adult available for 25 children.For the youngest age group (0-2 year) the ratio was slightly better with one adult for 14 children.

Child level
An at random sample of children was observed.In total, the definitive dataset consisted of 409 individual scores for involvement, and 224 scores for well-being.The average involvement was 2.96 on a five-point scale.Well-being was with 3.04, on a five-point scale, slightly higher.In both cases, the standard deviation was high, indicating the big differences in scores between children.There was a significant, but moderate correlation (r = .66,p < .001) between the two quality indicators; well-being and involvement.This interlinking of well-being and involvement is understandable, and is also evident in other research. 3

Indoors-Outdoors
In a standard observation, the observer did one observation (= this is one scanning round of 10 children) indoors and one observation outdoors.Both for well-being and involvement, the results outdoors were lower than the results indoors.The independent samples t-test, however, did not reveal significant differences (wellbeing: t(182) = 1.216, p = ns; involvement: t(367) = 1.391, p = ns).

Spreading of well-being and involvement
For well-being (N = 224) we concluded that almost 32% of the children did not feel good at the moment of the observation (scores 1 and 2).Approximately a third of the children (34%) felt rather neutral (scores 2+ to 3+).Another third (34.4%) felt good to very good in the setting (scores 4 and 5).
For involvement (N = 409), we concluded that 38.4 % of the children were not busy during the observation (scores 1 and 2); approximately a fifth of the children (22.5%) were doing active things without real investment of their possibilities (scores 2+ to 3+).More than a third of the children (39 %) were performing in top gear (level 4 or 5) at the moment of the observation.

Age groups
For all observations, the observers noted down the age of the observed child.For wellbeing we received data about the age of 190 children, for involvement 375 children were included.To know if well-being and involvement are influenced by age, we did an ANOVA 4 on the data of single age groups.In total, 190 scores for well-being and 375 scores for involvement were included in this analysis. 5For well-being we did not get indications of significant differences, linked to age (well-being: F(5,184) =.1.307,p = ns).For involvement, however, there was an age difference (F(5, 369) = 2.621, p < 0.05).
When looking closer only the observed difference in mean involvement between the youngest group and the 4-5 year olds was significant.Older children displayed significantly higher levels of involvement.Yet, the number of observations per age group is quite limited, and future analyses including more observations, is needed to draw any firm conclusions on this matter.

Group level
In total, 29 different groups were observed.In most of the groups, both an observation indoor and an observation outdoors took place.We noticed no significant differences between indoor-and outdoor observations.For this reason, we will base the following analyses on the group mean of indoor and outdoor scores together.Most of the group means were based on 20 individual scores (= 2 scanning rounds of 10 children/ observation).

Categories
Which mean level can we consider as sufficient or even excellent?The answer cannot be given by statistics alone.An approach is to imagine what kind of quality we want to receive at the level of experiences of children.Amongst users of the scales (Laevers, 2009) there is a consensus that 3.50 is a critical point and can be considered as a minimal acceptance.
For well-being, almost all of the observed groups have a group mean fewer than 3.50.This is no surprise, since also in the spreading at child level; the majority of the scores (66%) are beneath 3.50.
For involvement we get a similar picture.The majority of groups (81.5%) have a group mean beneath 3.50.

Child level
As a province, the received image out of the child observations indicates that quality ECD-programmes are in urgent need.A third of the children are feeling neutral and for another third of the children, well-being is worrying (score 1 or 2) at the moment of observation: emotional development of these children may be at risk.A similar picture is received for involvement.Only 39% of the children are developing and learning at the moment of observation (score 4 and 5).More than a third of the children are bored when we observed them at a random moment (score 1 and 2).

Group level
One could argue that this can all be explained by the deprived background of most of the observed children, thus, this would neglect the influence of the educational context and the fact that this context may act as a catalyst that can change well-being of children, as stipulated by Fraillon (2004).As argued above we see well-being and involvement not as child features, but as the result of complex interaction between features of the child and features of the environment, thus, indicating something of the quality of the learning environment, in turn.If so, differences between groups will be evident, since the sample of ECD-settings are all recruited in similar poor and rural areas.This is linked to differences in the quality of the educational context as well.Analysis on group level reveals these differences.Both for well-being and involvement we get a broad range of mean scores (see Table 5).In other research linking environment rating scales to the scores for well-being and involvement we have confirmation that the learning environment has a strong influence on wellbeing and involvement.The learning environment (defined in five dimensions; offer, group climate, room for initiative, organisation and style) is responsible for 28% of the differences in well-being and 40% of the differences in involvement at group level after multilevel analysis (Laevers & Declercq, 2011).

Discussion
In this study we attempted to gain a picture of well-being and involvement of young children in centre-based early childhood settings in the Free State.The findings show that the overall scores for well-being and involvement are low, but also that there are huge differences between individual groups.The average adult-child ratio in a Free State ECD-setting is worrying.Often there are a lot of children in a setting with only a few adults available.The average number of adults/child is 1 adult for 25 children.For the baby and toddler group (0-2 year), the average is, with 1 adult for 12 to 14 children, slightly better.In qualitative notes, two observers (different settings) mention that teachers do not get a wage, but work for charity.Another observer states that "Teachers are underpaid and this is what discourages them."A lot of practitioners mention, during the interview, that it is the love for children that keeps them going.This underlines the urgent need to invest in staff for ECD.
Both well-being and involvement of a lot of children, observed at an at random moments is worrying.The mean 'well-being' is with 3.04 low.If we look at the spreading, we get a picture where more than a third of the children (32%) are feeling at unease in centre based provision (score 1, 1.5 or 2 for well-being).Another third are feeling rather neutral (score 2.5, 3 or 3.5 for well-being).In the qualitative data, we get indications to explain this.Often basic needs of children are at risk: Some children are orphans, live with grandparents and are HIV positive.Some children are underfed and sometimes the only food they receive is at school (breakfast and lunch).
No food, troubled family-situations and too many children in the class to get individual attention might explain some of the low levels of well-being.
For 'involvement' we get a similar picture: the mean is with 2.96 low and the spreading is high (1.32):38.5% of the children are doing nothing of significance at the moment of observation, 22.5% of the children are doing routine activities, 39% are developing and learning when observed.
Although the deprived background of a lot of children can explain some of the scores, it's not all.Given the fact that observations are done in settings with a similar (and rather deprived settings), we see enormous differences at group level concerning well-being and involvement.This is hopeful.It means that the quality of the learning environment impacts directly on the levels of well-being and involvement.In other words, it is the ECD setting that makes the difference.Groups with a group mean for well-being and/or involvement higher or above 3.50 are achieving good quality within the hardest conditions.Further qualitative research is needed to investigate the critical success factors of those settings performing well.
For teacher development and education purposes well-being and involvement are valuable.The concepts match the intuitions of many student-observers and practitioners and give them a scientifically-based confirmation.When we can get children in that 'flow state' (this is a level 4 or 5 for involvement), development must and will take place within the area(s) addressed by the activity.In contrast to effect variables -the real outcomes are only seen in the longer run -the process variables give immediate feedback about the quality of interventions and tell us on the spot something about their potential impact.Furthermore, foregrounding involvement and well-being as key indicators for quality, engenders a lot of positive energy and synergy: the enthusiastic responses of children are very empowering and give the practitioner deep satisfaction both at the professional and the personal level.Many observers have implemented the concepts afterwards in their own lessons and approach towards children.Well-being and involvement empower people.

A critical note on issues related to the study
The Free State province is a large area to conquer.Transport was an issue.A taxi-• shuttle was organised.In small buses groups of four to five students, with one student responsible for the group, were put together to make the observation possible and in some cases, accommodation needed to be arranged.
At the beginning of the study safety was an issue.Most of the ECD settings are • situated in areas, which are perceived as being dangerous.After the studentresearchers were trained some parents raised their concerns with the Department of Education.A few requested that safety agents escort student-researchers for field visits.Some student-researchers refused to participate due to the location of the ECD settings.
At the time of the study there was no qualification directly related to early • childhood education at the university.The observers were students in Primary Education (linked to the Department of Curriculum Studies).They were not familiar with specificities of early childhood education.They were more geared towards education of older children and structured teaching driven by outcomes rather than the child's perspective and the value of free play.Whilst they were given a whole days training, which helped them to familiarise themselves with the basics of early education, more intense training before the actual observation would have made observations even stronger.
The field visits proved to be a profound experience for most student researchers.

•
There is often a significant cultural distance between the (mainly) Afrikaner girls as observers, and (mainly) Sesotho practitioners in ECD settings.One could ask about the levels of well-being of the student observers.Did the observers feel at ease in the setting?The same question can be asked about the observed practitioners: were they comfortable with strangers visiting their settings?
Language was an issue that complicated the study.Sesotho practitioners were • interviewed in English by mostly Afrikaans native speakers.Both for interviewer and interviewee English was an additional language.In this context nuances are lost or more difficult to grasp.
The research tool was developed in close collaboration with researchers at the • University of the Free State.However, the concepts underpinning the whole tool, well-being and involvement, were developed at the Research Centre for Experiential Education.It was also tried out in countries where culturally complexities are not as proliferated.Both concepts are broad and openly defined and suitable for a range of situations.In a country where the richness of culture and ethnicity matters, it is crucial to explore how these foundations influence the concepts of well-being and involvement.
Setting level refers to the preschool setting as a whole.Group level refers to the (age) groups or classes within the preschool.In most cases one preschool setting contains several groups.
2. Defined as "In the early phases and throughout the course of life, human development takes place through processes of progressively more complex, reciprocal interaction between an active, evolving biophysical human organism and the persons, objects and symbols in its immediate environment.To be effective, the interaction must occur on a regular basis over extended periods of time.Such enduring forms of interaction are here forth referred to as proximal processes.Examples of enduring patterns of these processes are found in parentchild relationships, child-child activities, group of solitary play, reading, learning new skills, problem solving, performing complex tasks and acquiring new knowledge and know how (Bronfenbrenner & Ceci, 1994). 3 . From other research we can see that both dimensions are interlinked with a moderate correlation of .50.In other words, we expect to see some correlation (cf.Laevers, F. et al. (2009).Werken aan kwaliteit vanuit het kindperspectief: welbevinden en betrokkenheid als richtsnoeren.Ziko II (eindverslag).Leuven: ECEGO.
4. Analysis of variance (ANOVA) is a collection of statistical models, and their associated procedures, in which the observed variance in a particular variable is partitioned into components attributable to different sources of variation.ANOVA provides a statistical test of whether or not the means of several groups (in this case, the data of the single age groups) are all equal or not.
5. We have limited the database to the data of single age group groups going from birth to six years.Mixed groups are not included, since the exact age of those children is unknown.
One exception is the mixed age group 0-2 years, since there is no overleap possible with the other distinct age groups.

Figure 2 :Figure
Figure 2: Histogram with normal curve for well-being (child level)