Strengthening ECEC Workforce Systems in Low-Resource Contexts: Insights from a Delphi Study

Abstract

Even though the majority of the world’s under 5-year-olds live in low- and middle-income countries (LMIC), our knowledge about the systems and architectures which support and sustain ECEC workforce quality is dominated by evidence from high-income countries (HIC). Drawing on a three-round Delphi study of consensus around essential training needs for early childhood staff in low-resourced contexts conducted among fourteen global experts over a period of four months, this paper highlights challenges faced when resources for teacher preparation are severely constrained, and unique approaches for sustaining quality in workforce preparation in such contexts. Key findings that inform efforts to build a strong ECEC workforce in resource-constrained contexts, include consensus among experts around (i) clear professional pathways which offer practitioner-centred flexibility through initial training, followed by on-going mentoring and supervision. A strong emphasis on the importance of strong systems of high-quality and supportive mentoring and supervision reflects recognition that initial training in such contexts is often relatively short-term; (ii) recognition of the importance of diverse roles (for example, community-based practitioners) in working with communities to deliver ECEC in complex communities; and (iii) systems supported by centralised financing, budget allocation and established professional pathways, but open to support and involvement of non-government entities in supporting workforce development.

1. Introduction

The global goals and targets introduced by international non-government organisations (NGSs) (including UNESCO and The World Bank) since the 1990s have resulted in an intense policy focus and a rapid expansion of formalised early childhood education and care (ECEC) across widely diverse contexts. While this welcome focus on ECEC has resulted in increased public policy, investment and research, the field itself has experienced less welcome impacts on young children. These include growing challenges to ensuring that all children have access to ECEC and difficulties with ensuring that ECEC offers all children safe, productive and enjoyable early learning experiences and outcomes. Recent evidence suggests that inequitable access to ECEC may exacerbate existing inequalities (Yang et al., 2024).

Responding to this challenge will require considerable flexibility and innovation in approaches to workforce preparation and training in order to ensure that all ECEC educators are equipped to provide young children with positive, productive and child-friendly early learning.

Recent concern has centred around workforce constraints. According to UNESCO (2024), global ECEC is facing severe workforce challenges in the quality of its training and preparation, workforce capacity and working conditions. For example, whilst the global average of ECEC educators who have achieved minimum required pedagogical training is 85%, the same average in low-income countries is 57%. And, when the global average ratio of trained pre-primary teachers to children is 1:17, the same ratio in sub-Saharan Africa is 1:54 and 1:60 in ‘low-income’ countries (UNESCO, 2024).

To support effective responses, more nuanced understandings are urgently needed of the opportunities and challenges associated with delivering ECEC in contexts where the resources and structural quality (e.g., ratios and qualifications) commonly associated with ‘quality’ in ‘high-income’ countries are severely constrained. The study reported in this paper was commissioned, in part, to provide the United Kingdom Department of International Development (DFID, now the Foreign and Commonwealth Development Office (FCDO), with a framework of non-negotiable, critical training needs for early childhood development (ECD) practitioners working in LMICs, where opportunities for training might be limited but remain critical.

1.1. Global Diversity in the ECEC Workforce

The contexts and working conditions in which early childhood educators across the world deliver early learning are diverse. The nature and location of ECEC settings vary according to local resources, policies and approaches to ECEC, and to the availability of trained personnel. Physical spaces range from community-owned, outdoor and/or purpose-built spaces, facilities attached to local public primary schools, religious settings and modified homes (Neuman et al., 2015). This range is reflected in an equally wide array of qualifications and entry requirements for both training and the profession.

There is significant variation in ECEC teacher backgrounds and qualifications in low- and middle-income countries (LMIC). ECEC educators may have: (i) no academic background or professional training in ECEC; (ii) some non-credentialed informal training in ECEC; (iii) a teaching certificate for ECEC and/or a degree in a non-related field; or (iv) a degree in ECEC or a degree with a postgraduate diploma in ECEC. In most LMICs, ECEC educators are not required to have tertiary level qualifications to meet national benchmarks for certification (see Figure 1). Entry requirements are typically high school completion and an ECEC teaching certificate. However, in many LMICs, a large percentage of ECEC educators in remote and rural areas do not meet even this basic requirement (Neuman et al., 2015).

Figure 1 compares minimum levels of educational attainment required to teach in pre-primary and primary education across 56 countries. In many countries, ECEC educators do not meet the standards which exist. In 2018, only 60% of ECEC educators in low-income countries met the minimum academic qualification required to be employed as ECEC educators, whilst 80% of primary school teachers attained the minimum required qualifications for employment (UNESCO, 2019).

In many countries, the systems for supporting ‘high quality’ ECEC, including workforce preparation, are new and emergent. This has resulted in various challenges for educators working to support young children. The lack of training opportunities available to individuals working in these settings results in lower levels of qualification than their counterparts working in more established, formalised areas of the education system (i.e., primary and secondary schools)—hence, lower status, lower pay (in many cases, pay in kind or voluntary employment) and an undefined career path. This, in turn, results in a shortage of trained, committed educators (Chiparange & Saruchera, 2016; Neuman et al., 2015; SEAMEO/UNESCO, 2016; Sun et al., 2015).

1.2. The Nature of Training for ECEC Educators in LMIC Contexts

The delivery of ECEC training also varies widely across contexts. As highlighted above, minimum entry requirements for teacher preparation can vary from completion of primary education to secondary school certification. Also, the duration of pre-service training tends to be short-term (commonly four to six weeks, but sometimes only a few days). Training, or teacher preparation programmes, are often delivered by NGOs which require teachers to teach in their early childhood centres, rather than by government-approved post-secondary institutions as part of recognised professional pathways (Rao et al., 2022).

In general, training for ECEC educators in LMICs is heavily influenced by approaches adopted in higher income countries across the global North. They tend to focus on enhancing knowledge about early childhood development and learning and are often organised around widely established areas of child development and child development milestones: cognitive; socio-emotional; motor; language and literacy and, increasingly hygiene/nutrition (Abdillah & Karna, 2014; D’Aprano et al., 2015; Fernandez-Rao et al., 2014; Tinajero et al., 2016; Vegas & Santibañez, 2011).

For example, the Science of Early Child Development (SECD, https://www.scienceofecd.com), an online multi-media course designed by Red River College in collaboration with the Fraser Mustard Institute of Human development at the University of Toronto includes an ‘international edition’ with modules on brain development, coping and competence, communication and learning, the ecology of childhood, and developmental health. Resources and materials which accompany these modules, including audio-visual and case studies, refer to a broad range of child development contexts. These materials have been used across the Aga Khan Foundation (Yousafzai et al., 2014), and by other organisations working across different regions of the world. While training based on these approaches has been widely adopted across LMICs, recognition of gaps in knowledge on systemic changes needed to support professional pathways for the ECEC workforce, as well as training needs of ECEC educators working in complex contexts and communities facing severe resource constraints, resulted in the United Kingdom’s Foreign Commonwealth Development Office commissioning the study on which this paper is based (Pearson et al., 2024).

2. Materials and Methods

This study used a Delphi approach to reach consensus about the essential training needs of practitioners working across three different sectors to support the delivery of ECEC programmes in low-resource contexts: education, health, and community-based practitioners. Recent analyses of effective approaches to supporting early learning and care, particularly in low-resource contexts, suggest that in order to thrive in formalised early learning settings, children must have access to ‘nurturing care’ (including early stimulation, adequate nutrition, health and sanitation conditions). This conceptualisation of ECEC recognises the importance of a holistic approach to supporting early learning needs, and suggests that ‘quality’ care requires a workforce that is equipped to address children’s nutritional, and health, as well as learning needs (Britto et al., 2017). The inclusion of community-based practitioners, who may not hold formal qualifications but constitute a critical component of the ECEC workforce in low-resource contexts, is also essential for understanding how the workforce can be strengthened to deliver ‘quality’ ECEC.

The Delphi ‘technique’ is a research methodology commonly used to explore areas, or issues, with limited available evidence by accessing and assessing expert opinion. Delphi techniques are characterised by an iterative process of data collection which incorporates four key elements: (i) a group of experts is invited to contribute their knowledge to developing understanding about an area of interest; (ii) the anonymity of the expert panellists is maintained to avoid bias through pressure to conform to group consensus; (iii) several rounds of enquiry; and, (iv) each ‘round’ is designed to support building consensus on key issues by expanding on and refining the responses to the questions posed in the previous round (Jünger et al., 2017).

This study had three rounds of data collection. This was determined by the recognised conventions around Delphi techniques (Iqbal & Pipon-Young, 2009; Yousuf, 2007). Round One consisted of a series of open-ended questions about the essential components of training. The questions summarised the key areas of concern seen in the literature view: they included desirable qualities and qualifications, knowledge and skills; methods and duration of training; support after training and training materials; monitoring the impact of training, and scaling up training.

The responses to the open-ended Round One questions were analysed independently by two members of the research team using a content analysis approach. They grouped similar responses together into one statement and organised them into key categories—with some adjustments to the original categories to reflect any patterns in the responses. They generated a set of 212 statements which were designed to be rated using a Likert Scale and to reflect the Round One responses.

These 212 statements were then organised into these six sections for the Round Two survey:

• Dispositions;
• Essential skills;
• Essential knowledge;
• Training—systems and delivery;
• Assessing the impact of training;
• Scale-up of training.

Key findings on the first three sections have been previously reported. This paper focuses on the expert consensus around the second set of sections, on building systems, or ‘architectures’, for a strengthened workforce.

A priori definition of consensus is recommended as a key component of rigour in Delphi studies (Diamond et al., 2014). While consensus in Delphi studies is most commonly measured and/or defined as the percentage of participants indicating agreement or strong agreement with statements, this study sought to strengthen the consensus criteria for this study by defining consensus as a measure of ‘the proportion of participants scoring within a range’. von der Gracht (2012) reports similar techniques in a review of consensus measurement in Delphi studies. This study’s approach (detailed below) measured both the extent of agreement or disagreement with statements and the strength of the agreement or disagreement.

In Rounds Two and Three, the study collated the frequency data and descriptive statistics from the surveys to present the levels of consensus per item as a percentage of the sample who responded to that item and rated the item as follows:

• ‘essential’ (7—top importance measure);
• ‘essential’—‘7’ or ‘6’ (top two importance measures);
• ‘essential’—‘7’, ‘6’ or ‘5’ (top 3 importance measures).

Items were categorised as having reached ‘strong consensus’ if (i) >90% of participants rated them in the top 2 levels of importance, or (ii) >80% of participants rated them using the top 2 responses, and if 100% of the participants rated them using the top 3 levels of importance.

Items were categorised as achieving ‘consensus’ if (i) >80% of participants rated them in the top 2 levels of importance or (ii) if 90% rated them in the top 3.

The remaining items were deemed not to have achieved consensus and were labelled ‘low consensus’. These low consensus items were reviewed in the Round Two analyses, and most were removed from the Round Three survey. They were reviewed (by referring to comments from the Expert Panel) to ensure that the low consensus was not due to unclear original statements or confusion about their meaning. This review resulted in minor amendments to a small number of statements.

The consensus was also measured by the extent of the change in responses between Rounds Two and Three. The study tested items presented in both Rounds to ascertain whether any of the differences were statistically significant (i.e., response patterns changed substantially by Round Three). The final items analysed at Round Three all reached strong consensus, consensus, or low consensus. In addition, some were identified as ‘consensus of disagreement’ because more than half the respondents now rated the item as ‘4’ or below.

A small number of items failed to reach consensus and were categorised as ‘non-consensus’ because there were at least two scores of 3 or lower. All but five items received at least one response rating of 7 (essential), and all items received at least a rating of 6. As such, the items which reached consensus of disagreement or non-consensus received a wider range of ratings than the three consensus categories.

Identification of Expert Panel

Reviews of the Delphi Technique suggest that establishing rigour depends on the careful selection of the Expert Panel. However, there is little consensus within the literature on the ideal number of participants, with sample sizes ranging from less than five to several hundred. For example, Akins et al. (2005) reference Delphi studies of competence training for primary care nurses, health promoting interventions and skills in young children involving panels of five to fifteen experts, while other reviews argue that the quality of an Expert Panel is more important than the numbers in assessing representativeness (Hsu & Sandford, 2007; Thangaratinam & Redman, 2005).

To provide a broad overview of expert views on training needs for different ECD cadres involved in implementing programmes in low-resource contexts, this study sought to gather expert views on a relatively complex range of factors. Its decisions about the composition of the Expert Panel were guided by these considerations, which had been identified in the literature review.

• The need to ensure the representation of knowledge and experience in delivering training to a range of different ECD cadres (including education, health and para-professionals) using different approaches.
• The need to ensure the representation of expertise from across the world regions.
• The need to ensure the representation of expertise based on involvement in shaping policy at national, regional and global levels.
• The need to ensure representation of expertise in direct delivery/implementation of training to ECD cadres.

This study’s selection of the Expert Panel was, therefore, made to include (i) experts currently involved in shaping global policy; (ii) a cross-section of experts currently involved in designing and implementing training for practitioners in ECEC settings and sectors across the world regions; and (iii) experts with broad experience in providing training on the ground.

The study invited 22 experts to participate, and 14 agreed to take part. All 14 completed Rounds One and Two. Two of these were unable to complete the Final Round and one opted not to complete items related to health professionals for Round Three on the basis of their expertise. All members of the Expert Panel are internationally renowned for their expertise and contributions to the provision of ECEC in low-resource contexts.

The study received ethical approval from the institution where the first author was employed. Each participant was assigned an identification number from the beginning of the study, in order to ensure strict levels of confidentiality and anonymity. All participants were asked when the study was completed if they would prefer to be identified as having participated, and 12 of the participants indicated that they would be pleased to be identified as having taken part in this study.

3. Results: Building Architectures for a Strengthened ECEC Workforce

The results are organised across three key sections, ‘Training’, ‘Assessing impact of training’, and ‘Scale-up’. The section on ‘Training’ is further sub-divided into four sections: systems; mentoring; teaching methods, and teaching materials. For each section, items from the survey that reached ‘consensus’, ‘strong consensus’, and ‘consensus of disagreement’ are presented in the tables, accompanied by analysis and insights provided by the experts’ open-ended comments.

3.1. Training

3.1.1. Systems

The strong consensus around the availability of professional pathways, facilitated by opportunities for both pre- and in-service training, matches the calls within the literature for flexible approaches which support career pathways (Ejuu, 2012; Mangwaya et al., 2016; Pearson et al., 2023, 2024). Strong consensus around the need to include ‘field-based’ components in training supports the consistent messages that an essential aspect of training for practitioners working to support ECEC is the provision of opportunities for site-based practice, supported by on-going monitoring and supervision (D’Aprano et al., 2015) (

Table 1. Training (systems).

Level of Consensus Reached		% Top 3 Ratings (5, 6 or 7)	% Top 2 Ratings (6, 7)	% Top Essential Rating (7)
Strong consensus	There should be opportunities for both pre-service and in-service training.	100	100	92
Strong consensus	There should be clear professional/training pathways.	100	92	75
Strong consensus	Training should incorporate a strong field-based component, where trainees/candidates spend part of their time receiving instruction in formal settings, followed by implementation of what they have learned in their respective professional settings.	100	100	64
Consensus of disagreement	Training should be centralised and administered by government.	25	0	0

).

In open-ended responses, the panel further indicated that it would be desirable to offer some shared learning across the three workforce groups (education, health, and community-based practitioners), but that these elements would only constitute a minor part of overall training and would have to be carefully planned.

Notably, there was consensus that, at least for now, training should not be centralised. Open-ended responses provide some insight into this view:

‘Training and professional standards do not always align easily because of different contextual needs and programmes. Standards often need to be more flexible’.

‘Mixed administration for training appears to be preferable at least until countries have developed the capacity to centralise this role. Collaboration between Govts and NGOs will be needed in this process’.

‘In my country experience, this [Government centralisation] has led to the erosion of a thriving and deeply experienced NGO training sector while inexperienced government institutions are struggling to offer quality training’.

3.1.2. Mentoring

The strongest consensus across all three Rounds was within this section, indicating overall strong agreement that this area is an essential aspect of effective training (

Table 2. Follow up mentoring and supervision.

Level of Consensus Reached		% Top 3 Ratings (5, 6 or 7)	% Top 2 Ratings (6, 7)	% Top Essential Rating (7)
Strong consensus	Supervision and monitoring should be delivered in a non-threatening manner.	100	100	58
Strong consensus	Observations of practice as part of supervision should be followed up by dialogue and reflection sessions.	100	92	58
Strong consensus	Training should be followed by on-site, on-going mentoring and supervision.	100	92	50
Strong consensus	Effective supervisor training is critical for programme success.	100	100	42
Strong consensus	Supervisors should be experienced.	92	92	58
Strong consensus	Training on its own is far less effective than training that is supported by follow-up supportive supervision.	92	92	58
Consensus	Systems of supervision and monitoring should provide opportunities for regular sharing sessions with peers.	92	83	17
Consensus	Supervision and monitoring should promote self-monitoring and reflection (for example, via self-monitoring guidelines).	92	67	25

). As highlighted in the literature review, this view is also reflected in current evidence. Supervision and monitoring are crucial because, as one expert explained:

‘Available human resources, trained or untrained, is a challenge in low resource settings. We need to explore more systems where teachers meet on a regular basis to discuss their teaching practices and receive feedback from others’.

Monitoring and supervision form key aspects of some of the most successful ECEC programmes (e.g., the Madrasa ECD Programme) catered to children in low-resource contexts. In the open-ended responses, however, experts also acknowledged the challenges around this aspect of training, specifically in terms of capacity of supervisors/trainers and commitment/investment:

‘Governments and organisations will do well to invest in a team of trainers to ensure that their skills depict what we want to see in the classroom; and that is determined by characteristics of effective teachers. In other words, if we don’t invest in the trainers of trainers, then we will get as little out as what we put in’.

‘Preferably, a tiered system of reflective supervision would be in place so that, no matter the level of the professional, there are role models to see at other levels; in my view, the issue of supervision and mentoring (which includes monitoring) is the least well developed in most contexts and often because it is seen as too costly. However, getting strong and effective supervision/mentoring in place EARLY is a key to assisting new professionals to adjust and manage their work. It needs to move away from a check list and policing mode to reflective supervision to have an impact on quality in my experience. This can include creating peer groups within a setting or geographic area to support and coach’.

3.1.3. Teaching Methods

The responses in this section (which presents items measuring consensus around appropriate teaching methods) are divided into three sections: (i) teaching methods considered to be highly appropriate for training practitioners working across multiple sectors (health, education, community-based programmes) (

Table 3. Teaching methods—all cadres.

Level of Consensus Reached	Teaching Methods—All Cadres	% Top 3 Ratings (5, 6 or 7)	% Top 2 Ratings (6, 7)	% Top Essential Rating (7)
Strong consensus	Planned refreshers and continuing professional development sessions.	100	83	42
Strong consensus	Reflection on practice.	83	67	25
Consensus	Participatory/ interactive sessions.	92	92	58
Consensus	Combination of instruction and active learning strategies, such as role-play.	92	83	50
Consensus	Supportive supervision and mentorship by skilled personnel.	92	75	42
Consensus	Interactive sessions (Q & A).	83	67	25
Consensus	Peer to peer learning in groups.	83	67	8
Consensus	Analysing examples of effective practice.	75	58	8

); (ii) teaching methods considered to be highly appropriate for training certified education professionals (

Table 4. Teaching methods—certified education professionals.

Level of Consensus Reached	Teaching Methods—Certified Education Professionals:	% Top 3 Ratings (5, 6 or 7)	% Top 2 Ratings (6, 7)	% Top Essential Rating (7)
Consensus	Educators develop and use practical resources during training.	100	58	33
Consensus of disagreement	Focus on delivery of a specific curriculum/package, as well as strategies for contextualising curriculum content.	58	25	25

); and (iii) teaching methods considered to be highly appropriate for training non-certified para-professionals (

Table 5. Teaching methods—non-certified para-professionals.

Level of Consensus Reached	Teaching Methods—Possible Teaching Methods for Use in Delivery of ECD Training (Non-Certified Para-Professionals):	% Top 3 Ratings (5, 6 or 7)	% Top 2 Ratings (6, 7)	% Top Essential Rating (7)
Consensus	Cadres have opportunity to observe experienced peers ‘in action‘ in home or early childhood settings.	92	58	17
Consensus	Cadres develop and use practical resources during training, in preparation for implementation in the field.	67	50	8
Consensus	Focus on delivery of a specific programme package, to ensure in-depth knowledge of each aspect and accompanying materials.	73	64	27

).

The Round One and Round Two responses indicated that, with regard to teaching methods, specific teaching methods are needed only for certified education professionals and non-certified para-professionals, with no need for particular distinctions for certified health professionals.

There is evidence of some consensus that a specific package or curriculum may not be appropriate for training education professionals (unlike non-certified para-professionals) (Table 4). Between Rounds Two and Three, there was some debate (reflected in open-ended responses) about whether training should focus on preparing to deliver a specific programme or curriculum or on supporting critical knowledge and skills to support contextualisation, or on flexibility in delivery.

For example, in Round Three, one expert indicated that they had reflected on this issue and now believed that training for non-certified practitioners should focus on a specific curriculum/package and that contextual adaptation should be the responsibility of more experienced educational leaders.

However, this was not a view shared by other experts, some of whom suggested that training should involve the discussion of different types of curricula to stimulate exploration among trainees regarding what might work best for children in the particular contexts within which they work.

Feedback from experts suggests a key outcome of training should be practitioners becoming able to adopt and adapt ‘packages’ to suit the characteristics of their setting. While this is challenging, it reflects pragmatic acknowledgment that education professionals are working across vastly diverse settings (i.e., from pre-schools on building sites to serve children of migrant workers to community-based programmes operating in peoples’ homes and serving children from diverse linguistic backgrounds).

The level of debate reflected here could also partially be explained by consensus that a focus on specific packages should form part of initial, ‘tier one’ training for non-certified para-professionals, followed by subsequent training for more qualified cadres which focuses on adaptation and application of curricula to local contexts.

There is a clear distinction between the training needs of education professionals and para-professionals (Table 5), with consensus that training for para-professionals should incorporate a focus on a specific package, to ensure that, by the end of training, trainees are well-equipped to deliver their respective programmes/packages. According to one expert, training could include:

‘simulations of every component of the daily routine; analysis of the daily routine component and how it promotes holistic child development and specific competencies; and materials making and use for that daily routine component; followed by opportunity for the teachers to role play/simulate one component of the daily routine with feedback from colleagues… this is enough to get teachers started if they are provided learning materials’.

3.1.4. Teaching Materials

As with teaching methods, this section (measuring consensus around items related to teaching materials considered to be appropriate for training) is divided into three sections: (i) teaching materials considered to be highly appropriate for practitioners across all sectors (

Table 6. Teaching materials—all cadres.

Level of Consensus Reached	Teaching Materials—Possible Teaching Materials for Use in Delivery of ECD Training (All Cadres):	% Top 3 Ratings (5, 6 or 7)	% Top 2 Ratings (6, 7)	% Top Essential Rating (7)
Consensus	Video resources (e.g., examples of a range of practices across different contexts that can promote discussion of various pedagogical approaches and interaction styles).	83	67	25
Consensus	A combination of relevant and appropriate materials, including locally developed and accredited resources.	82	55	9

); (ii) teaching materials considered to be highly appropriate for training certified education professionals (

Table 7. Teaching materials—certified education professionals.

Level of Consensus Reached	Teaching Materials Certified Education Professionals:	% Top 3 Ratings (5, 6 or 7)	% Top 2 Ratings (6, 7)	% Top Essential Rating (7)
Consensus	Teacher made resources as examples for cadres to make their own.	92	75	33
Previously low consensus—revised—consensus now achieved	Established ECEC curriculum tailored to level of practitioners and including information on strategies to adapt to diverse contexts.	91	27	9

); and (iii) teaching materials considered to be highly appropriate for training non-certified para-professionals (

Table 8. Teaching materials—non-certified para-professionals.

Level of Consensus Reached	Teaching Materials Non-Certified Para-Professionals:	% Top 3 Ratings (5, 6 or 7)	% Top 2 Ratings (6, 7)	% Top Essential Rating (7)
Consensus	Programmes/manuals/ECEC curriculum (training should closely follow guides and /or curriculum that cadres will be implementing, to ensure that they are equipped to deliver by completion of training)	83	75	33

).

Consensus about the use of audio-visual materials (Table 6) reflects the emphasis on active, engaged learning, with opportunities for cadres to learn a range of possible approaches to delivering programmes and curricula. This also fits with the concern that training should empower cadres by providing access to knowledge, whilst at the same time encouraging them to make informed choices about how best to deliver programmes in their own context. This is supported by using of a combination of materials (with an emphasis on relevance and contextualisation) to model the practice they need to adopt with children in their own settings.

In Round Three, one expert emphasised that introducing practitioners to a range of approaches was only useful if they offered examples of ‘very good practice … to at least inspire reflection and discussions’. Another emphasised the importance of discussing what is appropriate and relevant to the context in which cadres are operating—not making assumptions about what might be seen to be ‘effective’ based on examples from other cultures or contexts (especially not from high-resource or ‘western’ contexts).

Reflecting the debate on whether training for education professionals should centre on a specific programme or curriculum, the second item in Table 7 was revised between Rounds Two and Three as follows: ‘Established ECEC curriculum tailored to level of practitioners’ was revised to: ‘Established ECEC curriculum tailored to level of practitioners and including information on strategies to adapt to diverse contexts’. This revision was made in response to indications that, for education professionals, the use of curricula for early childhood learning needs to focus on encouraging adaptation to local context rather than on strict adherence. This revision resulted in increased consensus around this item.

The consensus appears to be that a focus on curricula is important in training-certified education professionals, and this could mean training around a specific curriculum or package. However, an essential component of training would be an emphasis on analysing the relevance of the curriculum to the context in which cadres are working and exploring ways to adapt and change where appropriate.

In contrast, for para-professionals, there is consensus that training should focus on a specific programme or curriculum to provide support in ensuring that cadres are equipped to deliver by the end of training (Table 8). This contrast acknowledges the different learning and professional needs across the two groups.

3.2. Assessing Impact of Training

The study’s findings for this section indicate considerable complexity in finding ways to assess impact of discreet influences in ECEC settings across diverse contexts: only one of the 21 items developed for Round Two reached consensus.

This lack of consensus around assessing the impact of training can be partially explained through the wide range of possible approaches suggested by members of the Expert Panel in response to the open-ended Round One questions (and also to the wide range of backgrounds represented on the panel). Their suggestions included using randomised controlled trials, gathering child-focused data, and documenting child perspectives on changes in ECEC settings.

There is consensus that, in the short-term, the impact of ECEC training could be measured through subsequent changes in ECEC settings—specifically, in terms of child-friendly approaches. In line with this view, there is a strong consensus that assessments should not use measures which might place undue pressure on ECEC cadres (

Table 9. Assessing impact of training.

Level of Consensus Reached		% Top 3 Ratings (5, 6 or 7)	% Top 2 Ratings (6, 7)	% Top Essential Rating (7)
	Short-term impact
Consensus	Documented changes in creating child-centred, age-appropriate learning environments.	80	70	10
	Long-term impact.
New item (to reflect open-ended comments)—strong consensus	Assessment of impact of ECD training should avoid using high-stakes measures, such as one-off summative testing.	100	80	40
Consensus of disagreement	Documentation of ECD cadres retention rates.	50	20	0

).

A large number of open-ended responses were received in response to the 21 items listed in the Round Two survey. They included these concerns.

Risks associated with the use of ‘globalised’ measures which would not reflect local contexts.

Challenges around ensuring that measures of impact can be seen to reflect quality of training and not various other factors such as pay and conditions. (This could explain the consensus that ECEC cadre retention rates are not likely to provide a useful measure of the impact of training).

Tracking pupil primary school attendance or monitoring child development outcomes were agreed by four Experts potentially to have some use in monitoring the impact of ECEC cadre training. However, Experts also noted the complexities of managing attendance tracking over the long term.

These findings reflect considerable complexities around the use of standardised assessment tools to assess the impact of programmes. There are particular risks associated with using such tools by untrained assessors, as this can lead to the potential of blaming parents for poor outcomes or holding pre-schoolers back from progressing to primary school.

The findings indicate a consensus of disagreement that ECEC cadre retention rates would not constitute a good measure of training impact. This study’s interpretation of this finding is that there may be multiple other factors which might influence retention rates (including, for example, issues around the recognition of status and remuneration).

Experts also commented on the logistics associated with ‘ideal’ measures. For example, phased target setting and follow up at set time intervals following ECEC cadre training was seen to be a possible ongoing method for monitoring impact, but experts warned that this would require much additional organisation and resources.

These concerns reflect how complex it is to recognise and assess authentic impact. They also match the general consensus in other sections that ECEC cadre training needs to be focused on empowerment and on building up from the ground. Arguably, these objectives are best assessed using programme-specific measures and tools. One expert commented:

‘The items in this section are very light on an ECD professional engaging in self-assessment in concert with input from parents, children and relevant others... Most assume that tools and an objective ‘outsider’ are the way to assess effectiveness’.

3.3. Scaling up

Experts strongly agreed that public support, in the form of financing plans and allocated budgets, is essential for scaling up training to ensure that all practitioners are provided with opportunities for initial training and professional development (

Table 10. Strategies for scaling-up workforce development.

Level of Consensus Reached in Round Two		% Top 3 Ratings (5, 6 or 7)	% Top 2 Ratings (6, 7)	% Top Essential Rating (7)
Strong consensus	Financing plan/budget.	100	92	42
Strong consensus	Availability of a range of trained personnel to support training initiatives.	92	92	42
Consensus	Stable workforce to support scale-up at all levels.	92	83	50
Consensus	Attention to how to scale to remote areas.	92	75	42
Consensus	Alliance of formal and non-formal sectors to ensure reach/coverage of training to all practitioners.	92	75	25
Consensus	Centralised plans for on-going supervision and mentoring.	91	73	45
Consensus	Established, recognised professional standards and clear career pathways that offer progression from basic training through to post-graduate level.	83	83	25
Consensus	Accredited training unit or institute at national/regional level to set policy and procedure.	92	58	25
Consensus	Commitment to intervention and accountability across all levels of administration.	92	58	17

). This indicates that, while there is consensus that the most effective approach for extending training may require the involvement of multiple providers (government and non-government), there needs to be some centralised planning and financing.

There was also strong consensus that trained personnel are needed to support training. This reflects similar views about the availability and quality of mentors/supervisors. Effective training requires a workforce of skilled trainers, mentors and supervisors. There is a consensus that there needs to be a stable workforce at all levels of implementation, from community to national level, and that there must be the capacity to support appropriate training at national level (as well as district and local level) to facilitate commitment and accountability across all levels of administration.

Agreement also exists on the importance of centralised plans for supervision and mentoring, and on the establishment of clear pathways for professional development. Importantly, there is also a consensus that efforts to scale up should ensure that energies are not focused on the mainstream, but on catering for, and reaching, practitioners who work in remote and hard-to-access areas/communities.

Experts recommended a combination of top-down and bottom-up approaches, reflecting concern with the need to ensure ‘buy-in’ and commitment at all levels of implementation.

At the national level, one expert recommends the ‘creation/utilisation of mobile teams of competent trainers composed of government sector, tertiary institutions and CSO to support local training centres/institutions’. This was supported by other experts who particularly identified the importance of mobile teams of trainers who could access marginalised areas.

Another expert suggested that developing a training institute at national level might act as a barrier because of entry requirements. This expert also emphasised the need for centralised commitment and budget to have a direct impact on practitioners at all levels and noted ‘supervisors should also have the authority and accountability to make changes in real time’. Another suggested that diverse training institutions should influence policy and procedure so this could be developed and reviewed.

Four experts appeared sceptical about the impact of policy level planning and inter-ministerial co-ordination. Whilst they indicated this was desirable, they also suggested it was difficult to achieve and that scaling up training could be implemented even when policy co-ordination was unsuccessful.

At community level, another expert recommended measures to ‘strengthen supports for community led early childhood programs rather than imposing programs; Build leadership for children in communities because without that best interests of children will be undermined and resources will be lost…Place greater emphasis on small cadres of very strong trainers who work with network of ECD programmes. What grows out of this bottom-up approach will ultimately fuel a national surge’.

In scaling up standardised training materials or modules, five experts commented that they must allow for innovation, contextualisation, and adaptation.

These comments reflect the considerable complexity involved in developing systems for appropriate training. Importantly, however, there is consensus in the importance of centralised commitment to the provision of training, reflected in the allocation of budget and in measures to ensure accountability at all levels of provision.

4. Discussion

The current workforce constraints outlined in the introduction section pose a significant challenge to achieving ‘quality’ ECEC globally, with implications which include the risk of ECEC exacerbating educational inequalities, despite widespread claims regarding its importance as an ‘equaliser’. Successful advocacy during discussions around construction of the Sustainable Development Goals (SDGs) ensured that access to some form of pre-primary education for all children was included as a key target of SDG 4. However, recent studies indicate that the global dependency on private investment in ECEC as a key government strategy for expanding access to pre-primary learning opportunities has resulted in growing levels of inequity in some contexts (Yang et al., 2024). Stronger, more equitable systems that provide more quality learning experiences for young children are supported when government policies around compulsory ECEC are accompanied by public investment in building strong systems of support (UNESCO, 2021). To support this work, a more nuanced understanding of how to build these systems in resource-constrained contexts is needed.

The findings presented here provide some useful insights into (i) the diversity in training needs which global and national systems need to address and (ii) possible strategies for moving forward in meeting workforce challenges through strengthened architectures for supporting quality ECEC.

The findings indicate that a combination of high expectations of practitioners entering the workforce, along with strong levels of contextual awareness and pragmatic responses, are needed to support workforce preparation. Experts referred consistently, across training components, to the importance of clear professional pathways which offer practitioner-centred flexibility through initial training followed by on-going supportive mentoring and supervision.

The emphasis on mentoring and supervision reflects broad consensus around some important components of quality training for educators (Darling-Hammond et al., 2017) in High Income Countries. Importantly, though, this focus also recognises that ECEC practitioners in LMIC enter the profession from a wide range of educational backgrounds, ranging from primary schooling through to post-secondary training. Accounting for this variability requires flexible pathways which offer multiple opportunities for up-skilling and on-going professional development.

There is growing consensus within the field that multiple pathways into ECEC are also important, and not only for attracting practitioners who may have not attained formal educational qualifications into the workforce. Many of these practitioners offer important life-skills and competencies associated with their knowledge of, and respect within, local communities. Research indicates that community-based practitioners play a significant role in delivering ECEC in complex situations, which require well-established relationships, including supporting children with disabilities (McConachie et al., 2001) and those in conflict settings (Murphy et al., 2015).

Successful examples do exist of short-term training followed by intensive mentoring. For example, the Ghanaian National Nursery Teacher Training Centre has offered five-day experiential, in-service training, followed by two-day and one-day refresher courses at four- and eight-month intervals. Initial evaluations indicate that this approach can result in positive impacts on classroom quality (Wolf et al., 2019). Similar approaches have been used in Jamaica and Chile, resulting in positive changes in children’s interest and enthusiasm among teachers who participated (Neuman et al., 2015).

However, the evaluation studies in Jamaica and Chile did not record significant impacts of training on children’s language or literacy skills (Yoshikawa et al., 2018). Risks associated with assuming links between enhanced training and child outcomes were highlighted by some experts, and this was reflected in consensus that the assessment of improvements in supporting child-centred, age-appropriate environments may constitute a more valid measure of training impacts than summative assessments of child outcomes—which are subject to various confounding influences.

A second set of findings relates to consensus that systems should consist of centralised financing, budget allocation and established professional pathways, but should also support the involvement of non-government entities in delivering training. This consensus appears to be linked to two important considerations. First, that non-government organisations concerned with supporting equitable access to ECEC are often in a better position to reach remote communities than government departments, both in terms of logistics and established relationships. Second, that non-government organisations have historically played a significant role in supporting ECEC. Recognising this experience and expertise, particularly in contexts where the workforce is constrained, is key to building strong architectures for quality training.

A final insight is the consensus around training focused on a specific curriculum for non-certified para-professionals, with graduation to a more critical stance for certified education professionals, centred around adaptation to context. This distinction reflects the complexity of training in LMIC, where training should be flexible in responding to different practice-based requirements. Even though non-certified para-professionals may not be ready to engage in critical approaches to curriculum delivery, there should be opportunities for professional development which enable more experienced practitioners to support adaptation and creativity.

Overall, the findings indicate that architectures for supporting workforce growth and development in LMICs should be grounded in strong systems of governance and public support but should also offer sufficient flexibility to respond to the diverse backgrounds, working conditions, and training needs of ECEC practitioners entering the profession. Evidence suggests that systems that are unresponsive to contextual challenges and opportunities are less likely to meet the considerable challenges faced by ECEC globally.