Predictors of Early Childhood Developmental Outcomes: The Importance of Quality Early Childhood Development and Education (ECDE) Services

Abstract

Few studies have established the influence of different aspects of early childhood development and education (ECDE) quality on children’s outcomes in low-resourced settings in sub-Saharan Africa. We examined the impact of different aspects of ECDE quality on school readiness in a low-income context. The current study is a cross-sectional sub-study of the Tayari preschool pilot program evaluation. Baseline data were collected from public ECDE centers. Multiple linear regression analysis was used to establish predictors of school readiness, that is, 4-6-year-old children being mentally, physically, socially and emotionally ready to start and succeed in primary school. Teaching experience, availability of textbooks and school facilities were significant predictors; learners’ school readiness scores decreased with each additional year of teachers’ experience, and were higher where school facilities were better, and in schools where textbooks were available. On the other hand, school enrolment, classroom resources, head teacher support, educational attainment and teacher training did not predict school readiness. Promoting quality preschool programs has important implications for policy as it can lead to improved school readiness and later success for children in disadvantaged settings.

1. Introduction

The importance of improving the quality of early childhood development and education (ECDE) services has received increasing recognition in recent years, given that more women with young children are joining the workforce and the demand for childcare provision has risen [1, 2, 3]. Research has clearly demonstrated that the quality of care and education provided to young children matters for school readiness [4], particularly in poor and disadvantaged settings [5]. High-quality ECDE services promote optimal child outcomes in all domains of development [6, 7]. On the other hand, low-quality ECDE services are associated with negative outcomes for children [8]. There are currently several definitions of preschool quality, depending on what elements are considered. The most commonly referred to elements include structural quality, process quality and educational beliefs of preschool teachers [9, 10, 11]. For the purpose of the current study, preschool quality is defined in terms of structural quality. Structural elements of quality which focus on the characteristics of preschools and preschool classes such as preschool and class size (number of children), teacher/caregiver education, qualifications, specialized training and job experience, child-adult ratios and classroom equipment and materials [12]. Teacher qualification is a key characteristic of structural quality as teachers and caregivers are central to providing quality ECDE [13, 14].

It should be noted that aspects of structural quality are often related to process quality [15, 16, 17], which is concerned with what happens in an ECDE setting. Bronfenbrenner’s ecological systems theory [18, 19] provides a basis for understanding the interconnectedness between process and structural quality. According to this theory, the child’s preschool classroom which is one of the social contexts in which the child operates, forms part of his/her microsystem. Within the preschool classroom, structural variables which are one of the spheres of influence on the child may be proximal (for example, classroom size) or distal (for example, economic conditions) [20]. A review of several studies has noted that highly educated and specially trained caregivers teaching in smaller classrooms with smaller child-to-teacher ratios are more likely to organize materials and activities in such a way that the environments are appropriate for children’s age [17]. Evidence from various studies suggests that high-quality ECDE programs should therefore have: (a) highly skilled teachers; (b) small class sizes with high teacher-to-child ratios; (c) age-appropriate curricula and stimulation materials in a safe physical setting; (d) a language-rich environment; (e) warm, responsive interactions between staff and children; and, (f) high and consistent levels of child participation [21, 22, 23, 24, 25, 26]. ECDE centers with such characteristics enhance young children’s cognitive and social development, particularly those from low-income families [27].

In a review of various studies, Manning and others [28] concluded that high teacher qualifications improve learning outcomes regardless of the culture and context. Further, the review established that there was a correlation between teacher qualifications and support for children’s development. To illustrate this point, the review highlighted that staff with more formal education and specialized training were likely to supervise and schedule age-appropriate activities; ensure that the room was organized and arranged in a way that enhanced learners’ experiences; provide varied social experiences for children; and create a warm and friendly environment for interactions. Similarly, other studies [29, 30, 31, 32] have reported that the high-quality pedagogic practices adopted by better qualified teachers create enriched and stimulating learning environments which are linked to better child development and learning outcomes. One of the essential elements of quality early childhood education programs is, therefore, the availability of qualified teachers with the requisite professional knowledge and skills to provide engaging interactions and classroom environments that support children’s learning [33]. Such teachers are more likely to use a variety of child-friendly teaching methods for individualized learning and small group teaching, which in turn enhances children’s learning outcomes [34]. Whereas these earlier studies illustrate the strong associations between teacher qualifications and child outcomes, other reviews reported poor correlations between ECDE teacher education levels and children’s performance [35]. Such mixed findings suggest the need for further exploration of the relationship between teacher characteristics and child outcomes.

While evidence has shown that highly qualified and professionally prepared teachers promote positive learning outcomes for young children [32], the lack of a supportive environment can hinder them from fostering quality ECDE [36]. ECDE infrastructure plays a critical role in ensuring quality, and a high-quality ECDE learning environment should be characterized by the availability of safe drinking water, appropriate toilet facilities, safe and well-equipped play areas and building structures that provide protection from adverse weather conditions. However, the majority of ECDE centers in many developing countries do not meet these requirements (see [37]), leading to poor outcomes for young children preparing to join primary school.

A developmentally-appropriate curriculum that emphasizes guided learning that is hands-on and language rich is an important determinant of highly effective preschool education [38]. Such a curriculum must be provided in an environment where learners are given opportunities to interact physically with a variety of materials and objects that promote the acquisition of different types of knowledge [39]. In resource-limited settings, the restricted availability of materials may not provide for stimulation in all areas of a child’s development. Furthermore, requiring energetic and curious learners to sit down for long periods doing work from their books may leave them bored, frustrated and unmotivated.

According to the Basic Education Act (2013) of Kenya, every child has a right to free and compulsory basic education [40]. Since the turn of the century, there has been a greater demand for childcare services, and the emergence of different forms of service delivery such as home-based, church-based and school-based care [41]. The preschool-based form of care, delivered through public and private schools which may be attached to a primary school, or established on their own, is the most common in Kenya. The provision of early childhood care and education programs was devolved to the counties in 2010, and the Act outlines that county governments are responsible for early childhood care and education programs. This includes the development of the necessary infrastructure of institutions used for conducting preprimary education. Although access to ECDE increased dramatically between 2009 and 2014, national gross and net enrolment rates remain low, at 78.4% and 77.2%, respectively [42]. Moreover, the provision of early childhood education services in Kenya, as in other countries in sub-Saharan Africa [21, 43] remains far from satisfactory, as the majority of public preprimary schools are characterized by inadequate play and learning materials, shortage of trained teachers, poor and irregular pay for teachers, and lack of health and nutrition services.

As far as the literature has revealed, many of the studies that have been conducted on aspects of ECDE quality, such as teacher qualification and teacher training/professional development, have established the influence of these aspects separately, without comparing which of the elements has a greater impact on ECDE quality. Similarly, although there have been studies investigating ECDE quality in sub-Saharan African settings [21, 26, 43], few of these studies have sought to establish the influence of different aspects of ECDE structural quality on children’s developmental outcomes. Such limitations of ECDE research make it difficult to determine what works in under-resourced contexts. The study reported in this chapter, therefore, aimed to bridge this knowledge gap by examining the impact of different aspects of ECDE quality on child developmental outcomes in a low-income context. Such information will be useful for the development of context-specific ECDE policies and practices to improve ECDE quality.

2. Methods

The study reported in this chapter was a sub-study of an evaluation of the Tayari preschool program. Tayari aimed to develop a cost-effective scalable model of ECDE that ensures children in Kenya are mentally, physically, socially and emotionally ready to start and succeed in primary school. The Tayari program sought to strengthen the existing ECDE model in Kenya through the development of child-centered instructional materials, interactive teacher training and ongoing instructional coaching and support, and a child health intervention that integrated psychosocial and health components to support the holistic development of the child. Tayari was designed as a randomized controlled trial implemented in public and low-cost private (also known as Alternative Provision of Basic Education and Training) ECDE centers. The components of the intervention were delivered through three treatment arms—treatment one received teacher training support through District Centre for Early Childhood Education (DICECE) officers; treatment two received the teacher training support plus books and teachers’ guides; treatment three received all the components for treatment one and two groups, together with a health/hygiene component. The evaluation study aimed to establish the impact of the Tayari program by comparing the performance of learners in each of the treatment groups, with that of learners in the control group.

2.1 Study setting

The Tayari program was implemented as a pilot project in four counties in Kenya: Laikipia, Nairobi, Siaya and Uasin Gishu. The four counties were purposively selected to represent diverse backgrounds. The geographical spread of the implementation zones was determined according to resource availability. Laikipia County is located at the equator, in the former Rift Valley Province. The main agricultural activities include grain farming, ranching and greenhouse horticulture. Nairobi County is located in the southern part of Kenya and hosts the capital city of Kenya, Nairobi. It is cosmopolitan and mainly urban in settlement. Community, social and professional services account for 52% of all the income generated in Nairobi. Siaya County is located in the Lake Victoria Basin and borders Lake Victoria to the south and west. The county is mainly rural in settlement. The main economic activities are crop and fish farming. Uasin Gishu County, whose capital is Eldoret, is located in the mid-west part of the former Rift Valley Province. The main economic activities are large-scale wheat and maize farming.

2.2 Study design

During sampling for the Tayari program, all public preprimary schools across the four counties were listed. Schools were then randomly allocated to treatment and control arms of the study. The evaluation study randomly selected a subset of the preprimary schools in the Tayari program. Power calculations were used to determine the number of preprimary schools required for detecting a mean effect size of 0.20 standard deviations at the program level, and assuming a school attrition rate of 5%.

The study reported here was designed as a cross-sectional study. The findings are based on baseline data which were collected from public ECDE centers during two phases of the main study; Phase 1 was completed in January 2016 while Phase 2 was completed in January 2017. The main study used a “stepped-wedge” design where half the sample of schools (Phase 1 schools) was included in January 2016, while the other half (Phase 2 schools) was included in January 2017.

2.3 Participants

In the case where the selected preprimary school had only one preprimary 2 (PP2—the highest class at the preprimary level before learners join primary school) class, 16 learners were selected at random from that class for inclusion into the evaluation study. If the PP2 class had less than 16 learners, all the learners in that class were involved in the evaluation study. For schools with more than one PP2 class, one class was selected at random for inclusion in the evaluation study. PP2 teachers together with the head teachers of the selected learners were also included in the evaluation study.

The study reported here was based on the data of all the PP2 learners (N = 4190) from 303 public ECDE centers that were included in the evaluation study. Those selected were children who were aged between 5 and 6 years and were expected to join primary class one at the beginning of the next schooling year. By the time the data for this study were being collected, the children were just beginning their second year of preprimary school and had been with the same teacher since the previous year (preprimary schools in Kenya have two levels in which the learner is expected to go through—the baby class and the pre-unit class. In most schools, teachers continue with the same children from the time they join the baby class until they move to the pre-unit class). The number of learners in each classroom across the ECDE centers ranged from 2 to 19, with a mean of 13.8 (SD = 3.44). As shown in Table 1, boys and girls were nearly equally distributed, with no significant differences across all the counties, χ² = 6.37, df = 3, p = .095.

County	Boys	Girls	Total
Laikipia	561 (49.3)	576 (50.7)	1137
Nairobi	369 (51.1)	353 (48.9)	722
Siaya	617 (54.2)	522 (45.8)	1139
Uasin Gishu	595 (49.9)	597 (50.1)	1192
Total	2142	2048	4190

Table 1.

Gender distribution across counties, n (%).

2.4 Data collection tools and procedures

Three quantitative instruments were used to collect the data: a head teacher questionnaire, an ECDE teacher questionnaire and a direct assessment administered to the learners. These tools were developed in consultation with the Tayari program implementers, that is, RTI International and the Ministry of Education. The head teacher questionnaire was used to collect information about center management, education and training background, enrolment of learners, attendance, class size, and facilities in the ECDE center. The ECDE teacher questionnaire captured data on education attained, professional training, access to learning materials in the classroom and classroom facilities.

The direct assessment test (DAT) was adapted from the UNICEF/UNESCO school readiness tool (Monitoring Early Learning and Quality Outcomes – MELQO) and early grade literacy and numeracy assessment tools used by other researchers. The adapted DAT, which was in English, was translated into Kiswahili. During the adaptation process, the DAT was reviewed and validated by a panel of ECDE experts which included ECDE practitioners, ECDE curriculum developers from the Kenya Institute of Curriculum Development (KICD), and academicians involved in ECDE research. The DAT was then piloted in 16 preprimary schools outside Nairobi which were not included in the evaluation study. The pilot data were analyzed and the results from this analysis were used to make further improvements. The DAT was used to assess learners’ progress in literacy, numeracy, health knowledge and psychosocial skills. All data were captured using tablets.

Field interviewers were taken through a 6-day residential and field training session on data collection procedures. During the residential sessions, they were introduced to the study objectives and trained on conducting interviews, administering direct assessments to children and ethical considerations to be made during fieldwork. A pilot training session in the field provided field interviewers with an opportunity to practice tool administration with teachers and pupils in non-study schools.

During actual data collection, one-on-one interviews with ECDE teachers and head teachers were conducted at a time and place that were convenient for the participants. All interviews took place in a comfortable setting that was free from interruptions. On average, the interviews took about 15 minutes to complete. The responses to the items on the questionnaire were recorded verbatim and were coded before analysis. For instance, the items on educational attainment were coded from 1 to 3, with 1 denoting primary-level education and 3 denoting university education.

The direct assessment tool was administered to learners on a one-to-one basis. Each assessment was completed in about 15 minutes and was preceded by an introductory 1–2-minute interaction between the assessor and the learner to establish rapport. A few practice items were administered before the test items to ensure that the learner understood the test requirements. Responses to the DAT were coded as 0 = incorrect, or 1 = correct.

2.5 Ethical considerations

The study was reviewed by the African population and health research center’s (APHRC) internal Scientific Review Committee. Ethics approval was sought and obtained from Amref Health Africa’s Ethics and Scientific Review Committee (ESRC). A research permit was provided by the National Commission for Science, Technology, and Innovation (NACOSTI).

Pre-visits were made to sampled ECDE centers to sensitize the county education officials and head teachers about the study, and to seek their permission to visit the schools. Informed consent was obtained before the interviews were conducted. Head teachers provided informed consent on behalf of the learners as many head teachers in Kenyan schools are authorized to do so by parents. Verbal assent was obtained from learners.

All participants were informed that their participation was voluntary, and that they were free to withdraw their participation at any point during the study. Participants were assured that their identities would be kept anonymous and all data would be confidential.

2.6 Data analysis

Data were analyzed at the ECDE center level. Descriptive statistics were used to provide information on school and classroom characteristics, teachers’ educational attainment, professional qualifications and years of teaching experience, head teacher support and school readiness scores.

The Tayari School Readiness Index (TSRI) was the primary outcome measure. The TSRI was a composite score derived by computing weighted scores from the DAT. First, items in the DAT were grouped into 10 subtasks. Learner percentage scores on each of the 10 subtasks were computed and multiplied by a weighting factor of 0.1, resulting in ten weighted scores. The ten weighted scores were summed to produce a composite score of the DAT. The maximum possible TSRI score was 100%. We then generated TSRI mean scores for each school.

Independent t-tests and chi-square analysis were run to check for differences in the variables of interest between Phase 1 and Phase 2 schools. Multiple linear regression analysis was used to establish the most important predictors of school readiness. Variables that were known to be predictors of school readiness, according to past studies in Kenya [44, 45, 46], were entered into the model. Teachers’ educational attainment was defined as the number of completed years of education. School enrolment was the total number of boys and girls enrolled in the school. Dummy variables were constructed to denote teachers’ professional qualifications at two levels (reference = untrained), certificate-trained and diploma-trained. Head teacher support was established by recording the number of times the teacher reported that the head teacher had observed a lesson. Availability of textbooks was a dichotomous variable coded as “0” for “not available’ and “1” for “available.” Composite scores were developed by summing the scores awarded on component items for school facilities, classroom facilities and learning materials. The school readiness and composite scores were standardized to allow direct comparison across the two phases.

3. Results

As shown in Table 2, the mean age of teachers in both Phase 1 and Phase 2 schools was fairly similar (39 vs. 38 years). The majority of teachers in both Phase 1 and Phase 2 schools had attained college-level education and had certificate-level professional qualifications. Teachers in Phase 1 schools had slightly more teaching experience than those in Phase 2 schools; however, these differences were not significant. The main employer for teachers in Phase 1 schools was the county government, while those in Phase 2 schools were mainly paid by parents. There were significantly more teachers paid by the county in Phase 1 schools compared to Phase 2 schools, χ² = 6.143, df = 2, p = .046.

	Phase 1		Phase 2
	N	%	N	%
Number of ECDE centres
County
Laikipia	50	32.5	42	28.2
Nairobi	27	17.5	24	16.1
Uasin Gishu	41	26.6	36	24.2
Siaya	36	23.4	47	31.5
Total	154		149
ECDE teacher’s highest level of academic education
Primary	8	5.2	9	6.0
Secondary	39	25.3	35	23.5
College	107	69.5	105	70.5
Highest level of professional training
Untrained	28	18.2	26	17.4
Certificate-trained	65	42.2	76	51.0
Diploma-trained	61	39.6	47	31.5
Source of teacher salary
County	86	55.8	62	41.6
Parents	65	42.2	83	55.7
School/church	3	1.9	4	2.7
Head teacher trained in school management
Yes	50	32.5	41	27.5
No	104	67.5	108	72.5
School facilities
Toilet facility
None	3	1.9	1	0.7
Pit latrine	96	62.3	86	57.7
VIP latrine	28	18.2	41	27.5
Flush toilet	27	17.5	21	14.1
Source of water
None	2	1.3	2	1.3
Carried from home	25	16.2	29	19.5
Well or borehole	43	27.9	0
Piped water	33	21.4	49	32.9
Rain water	19	12.3	17	11.4
Surface water	19	12.3	14	9.4
Water from vendors	13	8.4	9	6.0
Electricity?
No	44	28.6	46	30.9
Yes, but not working	14	9.1	17	11.4
Yes, working	96	62.3	86	57.7
Availability of textbooks
Yes	69	44.8	71	47.7
No	85	55.2	78	52.3
Multigrade classrooms
Yes	38	24.7	60	40.3
No	116	75.3	89	59.7
	M	SD	M	SD
Teaching experience (years)	13.09	7.36	11.89	7.52
School enrolment	28.23	19.55	23.86	15.96
Teacher age (years)	39.31	8.76	37.94	8.71
Number of times head teacher observed teacher	0.85	1.38	0.71	1.56
TSRI scores	33.12	8.13	31.66	8.07

Table 2.

Background characteristics of ECDE centers and teachers.

Significance of bold value is actual p values.

More than half the schools relied on pit latrines for their toilet facilities. Whereas more than one-quarter of Phase 1 schools used water from wells or boreholes, nearly one-third of Phase 2 schools used piped water for their daily needs. More than half of the schools in both phases had working electricity. In more than half of the schools in both phases, there were no textbooks available. There were significantly more multigrade classrooms in Phase 2 schools compared to Phase 1, χ² = 8.415, df = 1, p = .004. School enrolment was significantly higher in Phase 1 schools than in Phase 2 schools, t(301) = 2.13, p = .034. The majority of head teachers had not received any training in school management. The mean number of times that head teachers had observed teachers in the classroom was higher in Phase 1 than in Phase 2 schools. The mean TSRI scores for learners in Phase 1 schools were slightly higher, but not significantly, than those in Phase 2 schools.

In the multiple linear regression (Table 3), a significant regression equation was found (F(9, 210) = 3.426, p = .001), with an R² of .128. Teaching experience, availability of textbooks and school facilities were significant predictors of school readiness. Learners’ school readiness scores decreased by 0.143 points for each year of teachers’ experience and increased by 0.29 points where school facilities were better. Learners in schools where textbooks were available had 0.148 points higher than those in schools without textbooks. Based on the standardized coefficients, the school facilities variable was the strongest predictor of school readiness, followed by the availability of textbooks and finally length of teaching experience. School enrolment, classroom resources, head teacher support, educational attainment and teacher training were not significant predictors of school readiness.

	β	S. E.	t	p value
Constant		.714	1.762	.079
School enrolment	.125	.003	1.827	.069
Classroom resources	−.013	.083	−.174	.862
Teaching experience	−.143	.009	−2.031	.044
Availability of textbooks	.148	.136	2.199	.029
School facilities	.290	.071	3.837	.000
Head teacher support	−.090	.141	−1.354	.177
Years of education	−.158	.057	−1.867	.063
No teacher training	Reference
Certificate trained	.116	.227	1.036	.301
Diploma trained	.025	.250	.210	.834

Table 3.

Predictors of school readiness.

4. Discussion

While research has been conducted to examine the adequacy of ECDE infrastructure in developing contexts, and the influence of ECDE infrastructure on learning, studies on the most influential aspects of school quality are non-existent. The study reported here sought to fill this gap. We examined different elements of school and classroom infrastructure, as well as characteristics of ECDE teachers to understand their role as predictors of young children’s school readiness.

Our study revealed that school facilities played a big role in influencing children’s school readiness scores, supporting the findings of earlier studies [17]. Noteworthy is that the majority of schools sampled for this study had toilets, water sources and electricity, suggesting that efforts had been made to ensure that these were available for young learners. The finding that school facilities impact children’s outcomes has implications on the need to improve the infrastructure in public preprimary schools in Kenya, so that provision of these facilities is harmonized across all schools.

Whereas the availability of textbooks had an impact on learners’ school readiness scores, more than half of the preschools studied here did not have textbooks. School materials such as textbooks enable teachers to implement the curriculum according to their learners’ needs, and it is imperative that these are made available within schools in order to facilitate early learning.

The multiple regression results suggest that teachers who have taught for a long time produce poorer school readiness scores in young children. In an analysis not reported here, teacher age and teaching experience were found to be highly correlated. It may be that older teachers rely on “old school” pedagogical methods which emphasize rote learning and memorization. Such methods are not suited to young children, hence do not enhance their school readiness. One earlier study has reported that the types of engagement that young children encounter in prekindergarten settings impact their learning outcomes, and those that focus on individual instruction tend to be more influential [47].

Whereas other studies have concluded that ECDE quality is better when teachers have higher levels of education [48, 49], our results, although only significant at the 10% level, suggested that the more the years of education a teacher had, the poorer the outcomes for children. A review of earlier studies suggested that low associations between ECDE teacher education and child outcomes could be related to the content and quality of higher education programs [35]. Another perspective that has been offered is that perhaps the manner in which teachers interact with children, and their ability to “effectively implement appropriate curriculums” have a bigger influence on child outcomes than their qualifications [50]. Moreover, teachers’ willingness to apply what they have learned into practice may be influenced by individual, social-cultural and structural factors [51]. Based on the null findings on the associations between ECDE teacher education and child outcomes, we suggest, similarly to the views advanced by Early and colleagues [35], that improving teacher quality requires a broad range of professional development activities and specialized ECDE-focused training to improve pedagogical skills and interactions between ECDE teachers and children. Further, Siraj et al. [32], in their evaluation of an evidence-based in-service professional development program, found that it is critical to enhance teachers’ professional knowledge, skills and attitudes in order to improve pedagogical quality and child development outcomes.

In the multiple regression results, school enrolment was only significant at the 10% level, which is in line with previous research that reports that group size is a more critical element of quality for children of younger ages (less than three years) than those of preschool age (four to five years). Surprisingly, ECDE teacher pre-service training, classroom resources and head teacher support were not significant predictors of children’s school readiness in the regression model. These findings contrast those of other research which, for instance, reports that higher caregiver/teacher training [52], is associated with children’s school readiness. The contrasting findings may be related to the differences across the contexts studied.

The results of the current study are important for researchers in the early childhood development field, as well as policymakers, as they highlight the aspects of structural quality that are most critical for school readiness in low-income settings. Promoting quality preschool programs has important implications for policy as it can lead to improved school readiness and later success for children in disadvantaged settings. Based on the results of our study, we recommend that in contexts with limited resources and competing budgetary needs, improving facilities in preprimary schools would go a long way in enhancing outcomes for preprimary school children. Apart from what happens within the school, it is also important to consider other environments within which the child interacts with others. As has been suggested by Rimm-Kaufman and Pianta [53], in order to improve school readiness, there is a need to leverage the resources within the home, school, neighborhood and community. One of the ways in which this can be done is by strengthening the connections between the home and school environments [54], as well as encouraging parents’ involvement in school activities, especially among low-income families where such interventions are likely to have the greatest impact.