Investigating primary and secondary school staff views on the level of need for educational psychology services in England: “What does good look like?”

ABSTRACT

The research aimed to estimate educational psychologist (EP) service provision required to deliver a “good” level of EP service within primary and secondary schools in one local authority (LA) in England. There is no current shared understanding of the level of demand for EP services, nor the required workforce to meet projected needs. Using a mixed methods design, four index schools were recruited; key staff within each school forecasted potential future commissioning of EP services. Content analysis was used to present key findings from focus groups/interviews to a working group of EPs who assessed the exact time-cost required to supply services. By extrapolation of the data, a model forecasted potential future demand for EP services across all primary and secondary schools in the LA. Findings reveal that all four schools would, with sufficient funding, commission significantly more EP time than they do currently; details and proportions are provided. The model predicts that schools across the LA would commission a fourfold increase in EP time purchased. Under the assumption that other EP service levels remain constant, the model suggests that meeting this need for increased EP services would require more than doubling of the size of the LA EP workforce.

KEYWORDS:

• Educational psychology service
• primary and secondary schools
• action research
• supply and demand
• SENDCo

Introduction

Demand for and delivery of educational psychology services

Educational psychologists (EPs) in England perform a distinctive and valued role identifying and responding to special educational needs and disability (SEND), contributing to statutory Education, Health and Care (EHC) assessments, and upskilling other professionals (DfE, 2019b, 2023). They are reportedly in greater demand than services are able to supply (DfE, 2019b, 2023). Woods (2016) identifies how, since 1978, successive SEND legislation has resulted in increased demand for EP work, and reforms from the 2014 Children and Families Act have been identified as increasing demand for EP services (DfE, 2019b). Also, government focus on improving the mental health of children and young people may have increased demand (DfE/DoH, 2017; DfE, 2019b); as have EPs providing support following significant critical incidents (DfE, 2017; Dunsmuir et al., 2018; Salford Educational Psychology Service, 2017); and, more recently, undertaking COVID-19 recovery work (DfE, 2023; Szulevicz, 2021). Most significantly, requests for EHC needs assessments have increased by 23% since 2020, and the number of new EHC plans has increased each year since 2014 (ONSa Office for National Statistics, 2022). Against this background, a 2019 DfE report on EP workforce found that:

• 68% of Principal EPs (PEPs) reported ongoing difficulty recruiting for vacant posts

• 93% of PEPs said they were experiencing more demand for EP services than could currently be met

• 78% of EPs reported their workload was increasing and they did not have enough time to get everything done in their job (DfE, 2019b)

The problem, then, appears to be one of the availability of supply, rather than underutilisation of an available workforce.

At the same time, as demand for EP services has increased, government funding to local authorities (LAs) fell by an estimated 38% between 2010 and 2014 (Meegan et al., 2014). Financing became increasingly devolved to schools in many areas, providing either opportunities for adaptation and growth, or reduced services (Academies Act, 2010; Gibbs & Papps, 2017). Some LAs endeavoured to identify the “statutory minimum” of EP service they required, often underestimating the need for preventive and consultative work with schools, and capacity to respond to critical incidents, leading to increases in EHC assessments and sometimes to emergency funding of EP services (Fallon, 2018). Understanding the level and types of need expressed by service commissioners is crucial to avoid compounding the impacts of an already understaffed workforce.

EP workforce planning

Following the findings from the 2019 DfE EP workforce report, UK government’s Department for Education (DfE) contracting for the initial professional training of educational psychologists (ITEP) in England increased by 26% from 2020, and efforts have been made to improve retention of newly qualified EPs within the LA educational psychology service (EPS) workforce (DfE, 2019a). Whilst this will provide a boost in numbers entering the profession from 2023, there is no clear picture of how many newly qualified EPs are required each year to meet national demand, not least because there is no current method of measuring demand. Additionally, when issues of retirement, part-time working, and employment outside an LA are accounted for, increasing trainee numbers may not equate to a larger LA EP workforce. The recent report on national EP services clearly states that a purely supply-sided solution to “increasingly urgent capacity issues” is insufficient (DfE, 2023, p. 108).

A “needs-based” approach necessitates an understanding of the needs requiring EP intervention, together with definition of the services provided to meet those needs (Asamani et al., 2021). Since there is substantial diversity within and between LAs and individual settings requiring EP support, great variety in the structure and organisation of service delivery nationwide, and no reference point as to what constitutes an adequate level of service, or to what extent services are meeting growing demand, any endeavour to quantify EP shortages is challenging.

One basic method of measuring the supply and availability of EPs can be seen in international comparisons of EP: school population ratios. These indicate large differences between countries: for example, English regions report ratios of between 1:2,700 and 1:5,600 (these figures may include a range of within-region variation and do not, in any case, encompass the EP client age ranges 0–5 or 20–25 years) whilst ratios in the US are commonly around 1:1,500 and are sometimes state-mandated (DfE, 2019b; Jimerson et al., 2007, 2016; VanVoorhis, 2020). Where ratios provide broad insight to supply of practitioners, functional and specific detail and measurement is required to allow LAs to plan their workforce, and to inform decisions about ITEP, ensuring there is not oversupply of practitioners for the work available.

Applications from workforce planning tools in healthcare

Since no workforce planning tools exist for EP service delivery, it may be useful to look to other professions for guidance. A recent review of workforce planning models in healthcare identified key principles applicable to EP workforce planning (Hooper, 2023). The review highlights the importance of clearly defining how EP supply is quantified, in terms of available practitioners, full time equivalent (FTE) staff, or using available service hours; this can be complicated where multiple part-time staff often make up one FTE. Also, consideration of skill mix is essential to establish the required balance of skill levels for service efficiency. The review highlights productivity, work efficiency and work intensity as crucial factors in determining available supply.

Jackson et al. (2015) report that tools for managing the nursing workforce in the UK are complex and costly; for example, “Cassandra” was developed using 50 million hours of activity analysed over nine years. Whilst model developers agreed that each nursing action has eight dimensions to consider, they also highlighted complexities resulting from excessive workloads impacting quality of work, staff morale and retention (Jackson et al., 2015). However, the nursing workforce is not directly comparable to EPs, though some principles may be transferable to EP workforce planning.

Research focusing on nursing and midwifery in Scotland shows the importance of identifying standardised, timed activities and benchmarking the workforce against key variables such as mortality rates (Flynn et al., 2010). Fallon et al. (2010, p. 4) define areas of EP work as “ … consultation, assessment, intervention, research and training, at organisational, group or individual level”, which could act as relevant starting points; however, it may be complicated to standardise the time taken for these work areas due to variation of services in what is delivered, to whom and how. Yet it should be possible to sample directly how school commissioners currently use EP time and what EP services they might commission in the future, in the context of what they understand to be possible.

Aims of the current research

In 2021, The Association of Educational Psychologists (AEP) commissioned the host university to research how much EP service may be needed in English LA contexts, with the objective to obtain a clear view of workforce shortages and future training requirements. The aim of the current research is to provide the first step towards creating a national planning tool estimating the EP workforce required to deliver a “good” level of EP service; this, in turn, can be utilised to inform the number of new EPs entering the profession each year. In the UK services sector the question “what would good look like?” is commonly deployed within the field of quality assurance administration as a narrative device to focus upon standards benchmarking. In order to focus upon the design of an appropriate methodology, and to work within the resource constraints, it was decided that investigation at this first stage would incorporate both EP service supply and demand factors within a single LA, with a particular focus upon primary and secondary schools, since these sectors are the predominant service commissioners. The project research questions were:

1. How can a sample of primary and secondary schools be identified to represent the wider schools community (“index” schools)?

2. For each index school, what EP work is currently commissioned and what work might be commissioned if funding was not a constraint?

3. Considering the requirements of each index school (demand), what would be the EP time cost to meet those needs (supply)?

4. How can a prototype workforce planning tool for EP service delivery to primary and secondary schools across the identified LA be developed and evaluated?

Method

Study design

This research uses mixed methods, analysing both statistical and qualitative data; this involves an investigatory, collaborative action research design through a broad application of the RADIO model (Timmins et al., 2003, 2006). Collaboration is with the educational psychology service (EPS) within a large LA, focussing upon all primary/secondary schools that access the EPS. Four “index” schools were chosen to represent all schools. The EPS database was accessed to record the number of EP service sessions each school had purchased in the two previous years including details of how that time was used (one EP session is equivalent to three hours work). The 12 phases of the RADIO model were adapted to five stages for the current research: Stage 1 involves research commissioning, clarification of need, investigating dimensions of the problem, and application for ethical approval; Stage 2 involves identifying stakeholders, agreeing the focus and framework for data collection; Stage 3 is the data collection phase; Stage 4 involves analysing data and determining service allocation for each index school; and in Stage 5 a workforce model is created and evaluated.

Local authority context and participant recruitment

An EPS was recruited by approaching the principal EP and confirming their commitment to support the research over a two-year period. The EPS serves a predominantly urban population, with approximately 180 primary and 60 secondary schools. The DfE commissioned providers of ITEP programmes in England to create an index of need ensuring that LAs with the greatest need have trainees on placement. Using this tool, from a group of 153 LA areas, the chosen LA in this research falls within the top 15 highest in need (NOREMIDSW, 2023) this measurement tool incorporates data from IDACI; 0-25 population; new EHC needs assessments and established EHC plans made within the last year. Within the participating EPS two main grade, one senior, and the principal EP were recruited to form a working group. At the time of recruitment, approximately 40% of primary and secondary schools in the LA did not purchase any EP services and a conservative assumption was made within the wider project brief that this would continue to be the case.

Purposive, convenience sampling was used to recruit two primary and two secondary schools. All schools in the LA were ranked based on “notional SEN budget per pupil”; this is identified monies taken from their designated schools grant which must be used for securing SEND provision (Education and Skills Funding Agency, 2022). Potential schools were identified by the working group from the top of the list (high need) and bottom of the list (low need); schools were selected that were perceived by the group to have previously used EP services effectively and so were more likely to be able to fully engage in the research. School SEND coordinators (SENDCos) were then approached via their school link EP who presented the research opportunity; if there was interest at this stage, the researcher emailed the SENDCo directly. Two primary schools (one high need, one low need) labelled ‘HNPS’ and ‘LNPS’ throughout this report and two secondary schools (one high need, one low need), labelled ‘HNPS’ and ‘LNPS’ throughout this report were successfully recruited.

Ethics

Ethical approval was granted by the host university’s Research Ethics Committee (Environment, Education and Development School Panel, Ref: 2021–12759–20239). The project was evaluated as low risk since participants were non-vulnerable adults who provided informed consent and there was no activity involved that could cause risk of physical harm, psychological stress, or adverse effects. All data collected was fully anonymised before presentation. All participants were provided with information sheets and completed individual consent forms. Ethical practice/research guidance from Health and Care Professions Council (HCPC) (2016), the host university (University of Manchester, 2014) and The British Psychological Society (BPS) (2014) were followed throughout. An identified potential risk was the possibility of a working group member being linked to an index school providing a negative or less favourable forecasting estimate. This risk was mitigated by anonymising forecast data before presentation to the working group.

Data gathering methods

Focus groups and interviews

In primary and secondary schools, the SENDCo has oversight of operationalising the school SEN policy and coordinating provision for children with SEND (DfE/DoH, 2015); therefore, they are very well positioned to understand school requirements regarding EP service. Further to this, Gillies (2018) suggests that sound judgement in educational leadership requires appreciating a number of perspectives, discussing alternative approaches, and collaborative decision-making. Against this background, two focus group meetings were arranged at each of the four index schools, each to include the SENDCo, a member of the senior leadership team (SLT), and a class teacher. The first meeting in each school aimed to discuss previous EP service usage and possible areas for commissioning EP time; the second meeting discussed potential future use of EP services as set out in the school’s co-constructed forecasting template. In two index schools, focus group meetings were not possible, and so they were substituted by an interview with the SENDCo, following the same focus group questions. For consistency, there were two school participation conditions: the SENDCo (or SEN lead) fully participated throughout; and time was prioritised in between meetings for the SENDCo to discuss the needs of the school with the SLT and collaboratively complete a forecasting template for commissioning EP time. All focus groups/interviews were audio recorded and each index school provided to the researchers their forecasting template.

In summary, there were four stages of data collection for each index setting, first a profile of EP time usage was created for the current and previous year; second, the initial focus group or interview was conducted to discuss previous EP usage and potential future use; third, there was a required period of school internal liaison and completion of the forecasting template; and fourth, the forecasting template was returned and discussed in a second focus group or interview.

Working group meetings

There were three, two-hour meetings of the working group, each of which was audio-recorded, these are summarised in Table 1. In meeting 1, potential index schools were identified, research questions were refined, and data collection methods were established. Directed content analysis (Hsieh & Shannon, 2005) of index school focus group and interview data was presented to the working group in meeting 2; the forecasting templates for each index school, including rationale for the forecast provided through the analysed data, were evaluated by the working group to specify the EP time cost required to deliver the forecasted work.

Table 1. Summary of working group meetings.

Meeting order	Working group actions
Meeting 1	Refine research questions, identify index schools and agree data collection methods.
Meeting 2	Scrutinise EP time allocation for commissioned work using collected data and local knowledge of services structure and context.
Meeting 3	Evaluation of the model.

In meeting 3, the preliminary outputs from a prototype workforce planning model were evaluated. Having shared with the group at meeting 2 the forecasting templates provided by index schools, three working group members identified two non-index schools each for comparison to the model’s estimation of need. Each member selected one primary and one secondary school that they knew well, and then analysed the school’s needs using a forecasting template in the same way as had been undertaken within the four index schools (working group members had no knowledge of the model’s prediction for their chosen schools at this stage). From this, each member provided an estimate of EP time needed and a brief rationale for their assessment. This estimate was then compared with the model’s prediction and the group identified possible reasons for any discrepancies.

Creating a workforce planning model

By extrapolation of the data gathered from the index schools, a model was created to predict how much EP time other schools within the LA might commission, identifying comparable level based upon the school’s notional SEN budget per pupil and number of pupils on roll. The following gives a description of how this was created.

Schools that did not buy EP time in the year 2021/22 were removed from the overall service delivery estimation. Separate modelling was undertaken for the primary and secondary schools sectors. Then, the number of EP sessions needed at each non-index school was derived by producing a session requirement per pupil for each school, based upon index school estimates, and then multiplying this by the number of pupils at the non-index school. The sessions per pupil requirement between the two known data points was derived by using a sliding scale from one known data point to the other, using the notional SEND budget per pupil as the ratio. For schools outside the bookends of the two known data points, a simple calculation relative to notional SEND of the adjacent school was used to define sessions per pupil requirement. It was then possible to calculate a summation of projected EP service need across the primary and secondary school sectors, respectively. Finally, a calculation of total forecasted EP sessions across all schools in the LA was made by adding primary and secondary school estimates together.

The EPS organises service delivery broadly into three main areas: “Statutory work”, which involves the completion of EHC needs assessments; “school traded”, which is EP time purchased by schools; and “non-school traded”, which is EP time bought by non-school organisations such as The Clinical Commissioning Group and Youth Justice Service. The model’s forecasted estimates necessitated increasing non-front-facing EP time (for example, time for supervision, team meetings, leave) to meet expanded service requirements and a conservative assumption of static levels of non-school traded and statutory demand levels (as at 2021/22).

Data analysis

Descriptive quantitative analysis and summary using spreadsheet tables was used to give an overview of each index school, to include key demographics, number of EP sessions purchased in 2021/22 and EP sessions required to deliver the work identified in the index school’s forecasting templates.

Qualitative content analysis was used to analyse transcripts of audio recordings. This is a process of examining text which is flexible dependent on the theory, area of interest and problem being studied (Hsieh & Shannon, 2005). The specific aims of this analysis were to identify exact descriptions of work commissions from index schools and present findings of an evaluation by the working group. Qualitative content analysis was chosen as it is useful for analysing manifest rather than latent meaning (Graneheim & Lundman, 2004). The specific approach used was directed qualitative content analysis, which takes a deductive approach, building on existing understanding and extending conceptual frameworks using predetermined categories (Hsieh & Shannon, 2005). Two simple coding frameworks were created and used for analysis; both frameworks were checked between the first and second researcher for fitness-of-purpose. The first framework was used to analyse transcript data from each of the four follow-up focus group meetings (or interviews). The second framework was used to analyse transcript data from the third working group meeting when members evaluated the model.

One potential drawback to qualitative content analysis is that researchers may have a bias towards finding particular data, which may be addressed through auditing (Hsieh & Shannon, 2005). Before analysing the full data set, extracts of transcript data were analysed by a second doctoral educational psychology researcher, independent of this project, who applied the same coding frameworks. Analyses were cross-checked with the first researcher’s coding and where ambiguity was identified, the two researchers discussed and clarified. Member-checking was also employed, whereby school staff and working group members were shown narrative summaries of their data to check for accuracy, asking that any omissions, inaccuracies or misinterpretations to be highlighted before finalising. Participants were satisfied with the summaries and did not ask for any changes to be made.

Findings

Index settings: descriptive statistics, historical EP sessions purchased and forecasted EP sessions

Table 2 shows descriptive data for each of the four index schools, where LNPS is the Low Need Primary School; HNPS is the High Need Primary School; LNSS is the Low Need Secondary School; and HNSS is the High Need Secondary School. It can be seen that the key statistic differentiating the schools is the percentage of pupils receiving free school meals, whereby the low need schools have a significantly lower percentage compared to the high need schools; this follows the distinctive figures for notional SEN budget per pupil which was the statistic used to identify high and low need schools. Other statistics that might be considered in relation to level of need, such as % of pupils on SEN support or % with EHCPs are less closely matched with high/low need categorisations as higher percentages might be expected in high need schools and lower percentages in low need schools. However, low numbers on SEN support could be indicative of unresponsive SEN systems in school; similarly purchasing large quantities of EP time may reduce the need for EHCPs. The HNPS shows an elevated number of pupils on SEN support and a very high proportion of pupils speaking English as an additional language.

Table 2. Descriptive statistics: key demographics for the four index schools.

Index School	Type of school	Notional SEN budget per pupil 2021/22	Pupils on roll	EHCP %	SEN support %	EAL %	FSM %	FTE Teacher	Pupil to teacher ratio	FTE TAs	Previous EP sessions purchased
											2020/21	2021/22
LNPS	Academy	446	433	3.93	10.39	20.3	16.6	20.4	21.1	14.7	4	16
HNPS	Maintained	894	671	1.49	21.16	63	50.75	35.9	17.9	35.7	76	76
LNSS	Academy	646	1663	2.41	18.8	15.6	15.4	91.3	18.2	17.4	19	19
HNSS	Academy	1042	1008	3.08	15.87	21.9	49.3	46	15.9	13.8	0	39

Nb. Population data from 2021/22; workforce data from 2018/19 (DfE, 2022)

Table 3 shows the number of EP sessions requested by each index school based on staff completion of the forecasting template. It can be seen that all forecasts indicate a desire to access more EP time than schools currently purchase. The two high need schools show a larger percentage increase in forecasted EP sessions than the low need schools, with both requesting over 100 sessions more than they currently purchase. The HNSS forecasted a substantially larger increase than the other three; Table 2 shows that this school has the largest notional SEN budget per pupil.

Table 3. EP sessions requested based on forecasting template returns from index schools.

Index School	EP sessions requested based on forecasting template*							Increase in number of EP sessions from 2021/22	% Increase from 2021/22
	Total	Consultation	Assessment	Training	Intervention	Research	Other
LNPS	28	11	6	2	5	4	0	12	75
HNPS	192	49	49	10	78	0	6	116	153
LNSS	45	15	10	5	10	2	3	26	137
HNSS	186	50	50	8	78	0	0	147	377

*These numbers are a combination of the sessions requested by school staff who gave detailed reasoning for requests, with additional sessions identified by the working group to ensure sufficient time to meet identified needs.

School staff rationale for how they would use forecasted EP sessions

Forecasting template returns from all four schools indicate they would like to commission more consultation and assessment work than they do currently; three out of the four schools gave a clear view as to the number of individual cases they would like EPs to work on.

… three children who – and each child I think for a thorough assessment, probably needs, and a report writing, and actually the EP would need to meet with parents of all of these children … So six [sessions] for the assessment and advice. (SENDCo LNPS)

But in the course of a typical year, how many children would I be wanting our EP to come in and really do some productive, impactful work with? Probably about ten, twelve. … I think at present it’s probably more like six to eight. (SENDCo LNSS)

Well, more than that, yeah, thirty to forty students, we’d want to see. If I had a blue sky, I’d like to – there’s a lot of children that are sitting under the radar that have been missed …

(SENDCo HNSS)

Rather than basing estimates on the number of cases, staff at HNPS allocated the amount of time they could potentially use for this work: one day per week across the year. School staff identified a variety of purposes other than individual casework for using consultation with an EP, for example, providing drop-in sessions for parents or finding the most suitable interventions for specified needs.

Very little commissioning for research was envisaged; three of the four index school staff commented that this was because there were designated staff in school who already did this type of work. All four school forecasting templates allocated time for training, ranging from one “all-staff” training session to several across the year, or a mixture of “all-staff” and small group training.

So yeah, so it was a full day training, then I think three additional sessions, or three like half days, with those different groups like our mentors. (SEN lead, HNPS)

There were two areas of EP work, highlighted by all four school forecasting templates, where staff would like to allocate EP time, but under normal circumstances they are unable to due to competing priorities and insufficient time. These were staff training and implementing interventions. Table 3 shows that both high need school forecasting templates indicated 78 sessions for implementing interventions, which equates to one day per week across the year during termtime. School staff envisaged EP time in this area being used to identify suitable interventions and to regularly deliver interventions as well as to train their staff in delivery.

So intervention is where, within our own provision, I think there’s a significant gap, and I think in an ideal world, that’s what we would have loved to have done with our EP … having a morning with our EP, where they might do three different intervention groups for different needs throughout the school. (SEN lead, HNPS)

We don’t know what interventions are out there or what interventions might be suitable for particular students, but if an EP was to come in … they might pick up on areas of need that we might not necessarily have picked up on, and then those students could be targeted for intervention work. (SENDCO, LNSS)

Working group analysis of index schools’ forecasted commissioned work: calculating corresponding EP time costs

Deciding how much total EP time the forecasted work required involved careful deliberation with the working group and resulted in a pragmatic approach for predicting time allocation. Consideration was given to the exact nature of work identified and additional time was given in some areas where requirements had been overlooked, such as, planning, preparation and writing reports. For example, the group differentiated between drop-in or brief consultations requiring no written report, where 0.5 sessions were allocated, and more complex cases involving consultation, assessment and written reports where between 2 and 4 sessions were allocated. Similarly, for interventions that were short or familiar and involved only training/coaching staff, between 2 and 5 sessions were allocated; those requiring delivery, modelling or more bespoke production, between 5 and 10 sessions were allocated. Exact session numbers depended on the details provided by school staff.

Workforce planning modelling across the EPS

Table 4 shows data taken from the EPS database for 2021/22 and projected data based on the findings from the four index schools. It can be seen that there is a projected more than fourfold increase in commissioned EP time compared to 2021/22 levels in the area of traded work with schools. Maintaining the same level of delivery for statutory and non-school traded work (since these areas have not yet been researched, and it is conservatively presumed that there is no change to demand) whilst increasing demand for school traded work in line with the index school forecasts, including corresponding “non-front facing” time, requires slightly more than double the current service staff levels.

Table 4. EP sessions for current and projected delivery to the four index schools.

Year	School traded	Non-school traded	Statutory	Non-front-facing time*	Total sessions	Total staff required**
Current year (2021/22)	2366	2366	3154	5368	13254	31
Projected year	10700	2366	3154	11041	27261	63

*CPD, management time, supervision, team meetings, leave, travel, etc.

**This includes a mixture of assistants, main grades, trainees, specialist seniors, seniors and principle EPs.

Evaluation of the workforce planning modelling

Table 5 shows a comparison of the working group’s assessment of need in six non-index schools (in terms of forecasted EP sessions), without prior knowledge of the model’s predictions for those six settings but with knowledge of the forecasts for index schools. These estimates were then compared with those predicted by the model. There followed a discussion as to reasons for any disparity.

Table 5. Projected EP sessions for six identified schools compared with working group estimates.

School	Model projection (EP sessions)	Working Group assessment (EP sessions)	Difference
Primary 1	31	40	WG estimate +9
Primary 2	67	59	Model projection +8
Primary 3	121	80	Model projection +41
Secondary 1	164	106	Model projection +58
Secondary 2	100	124	WG estimate +24
Secondary 3	152	88	Model projection +64

Based on notional SEN budget per pupil, primary schools 1 & 2 have relatively low need, the other 4 schools represent a high level of need. Estimates made for the two lower need primary schools were close to the model; however, when schools had a higher level of need, EPs tended to estimate significantly less EP time needed than would be predicted by the model. The group provided reasoning for these differences. In summary, EPs had a specific understanding of school contexts and identified barriers to implementing very much larger quantities of EP time, such as ineffective senior leadership or restricted SEN teams; where they judged capacity to be restricted they underestimated the number of sessions compared to the model. However, in one case the fact that the school had extremely competent senior leadership led to an assessment of less EP time required.

Where the working group over estimated EP time compared to the model, they identified three reasons: new leaders, SENDCos or teachers requiring extra support; schools ranked as “inadequate” by OFSTED; and specialist knowledge of EPs making the estimation able to identify greater potential for applying EP time. Based on the evaluation of the working group, it is possible that a school-based estimation of need may slightly overestimate realisable supply to meet that need.

Discussion

Findings summary

The four research questions will now be addressed individually. First, index settings were identified by ranking all schools in the LA by level of need, apportioned by their notional SEN budget per pupil. Two high need and two low need schools were recruited; these were used as a sample to represent all primary and secondary schools in the LA. Second, by collating data held by the LA EPS it was possible to summarise previous EP sessions purchased by the four index schools. Key members of school staff from each index setting used a forecasting template to collaboratively estimate how many EP sessions they might commission without funding restrictions, also providing details of the requested work. Third, an established working group of EPs from the LA appraised the detailed requests made by school staff and determined the amount of EP time necessary to deliver identified services, giving a final figure for each index setting. Fourth, these figures were extrapolated across all primary and secondary schools in the LA, creating a statistical model (not including 40% of schools who did not buy any time from the EPS at the time of sampling). Projected EP time requirement for individual schools, as calculated by this model, were evaluated by comparing predictions with independent estimates made by the working group of EPs.

A method for measuring schools’ future need for EP services was devised involving consultation with SENDCos and senior leaders using focus groups, interviews and a forecasting template. Applying this approach in four index schools revealed that all schools would commission more EP time if they had sufficient funding; schools with a high notional SEN budget would increase commissioned time far more than schools with a low budget. Whilst all four schools projected that they would commission more consultation and assessment work than they do currently, schools’ projections to commission substantial amounts of EP time for training and implementing interventions was striking in that they do not currently allocate any resources to these types of EP services due to competing priorities. This is in line with the recent report on national EP services where school staff shared a desire for increased EP direct intervention work with children, young people and families (DfE, 2023). These findings demonstrate that schools may require increased funding to commission the EP time they need or changes to how they currently allocate budgets.

By extrapolation of the data gathered from the index schools, a model was created to predict how much EP time other schools within the LA might commission, establishing comparability with reference to settings’ notional SEN budget per pupil and number of pupils on roll. Projected outputs from the model estimate a more than fourfold increase in commissioned EP time in the area of traded work with schools compared to current levels. In turn, this would require the EPS to more than double its current EP staffing levels to accommodate projected demand across all of the primary and secondary schools within the LA. However, this only accounts for traded service time; statutory SEND work and non-school traded service delivery were not evaluated and were conservatively assumed to remain at current levels of demand. Overall, findings from this research confirm reports of shortages in EP services nationally (DfE, 2019b, 2023), and offer a first step towards quantification of the resource deficit. This innovative research has directly gathered and quantitatively evaluated data related to service user demand for psychology services, an identified need within educational psychology workforce planning in England (Hooper, 2023).

There has been rapid growth in the EP profession in England over the past 50 years. In 1968 it was recommended that there should be a ratio of 1:10,000 EPs to children (Department of Education and Science 1968), whereas the DfE (2019b) reports ratios of between 1:2,700 and 1:5,600 (these figures may include a range of within-region variation and do not, in any case, encompass the EP client age range 0–5 or 20–25 years). However, EP: population ratios do not provide any rational understanding of demand or need for EP services. This research has for the first time started to answer the question “what does a good level of EP service look like?” and provides data to suggest that even with the recent government-funded 26% increase in training capacity from 2020 (DfE, 2019a), the effects of which will filter through to service delivery sites from September 2023, there may still be insufficient workforce supply to meet national demand. Furthermore, increasing numbers in training does not directly equate to a larger LA EP workforce, when issues of retirement, part-time working and EPs moving into non-LA practice are considered; to accurately measure the required workforce, a clear picture of the whole system of demand and supply is needed (Hooper, 2023).

Given the complexity of workforce planning and the innovation required in a field where there is no existing precedent, it was necessary to take a pragmatic approach to both providing a first step towards a national understanding of the demand for EP services and demonstrating an initial method for calculating need across schools in a LA. The findings of this research are indicative not definitive and will contribute to a wider project; as outlined below, further research is necessary to fulfil the original commission.

Limitations

In the present research, there are six limitations to consider. First, the research did not seek to engage critically with school commissioners of EP services and chose to report an independent prospective estimation of service need. Some barriers to utilising EP time were identified by the EPS working group but there was no moderation to school staff forecasts, for example, with respect to potential difficulties organising, applying or enabling commissioned work. Critical insight could be sought from EPs who know the school, with comparison of EPs’ perspectives on projected need with school staff’s assessments providing a counterbalance; however, on the question of “need” per se, it seems more appropriate to trust the judgement of SENDCos and senior leaders who know the SEND needs of the school first hand. These staff are experienced in commissioning and allocating EP time and whilst EPs deliver services they do not have this experience and are often only in schools for brief periods. From this perspective, it may have been more useful to evaluate the model using estimates from non-index school staff focus groups.

Second, the forecasting template was devised in a local context and does not include a comprehensive list of possible EP services, allowing school staff to highlight their own identified needs; however, there may be areas of work which have been missed, for example, “follow-up” work reviewing previous EP involvement.

Third, many schools in the participating LA do not currently buy “traded” time from an EPS. Understanding the reasoning and reluctance to commission EP work would provide further information to give a comprehensive view of relevant school needs. Fourth, having only two data points in a large sample of primary schools (>100) weakened the predictive power of the model used. A third index school, taken as a mid-point between the HNPS and the LNPS, would, theoretically, provide greater predictive accuracy of extrapolations from the index schools to the non-index schools. Secondary schools were fewer in number and so having two data points is less problematic for assumed predictive accuracy, though more index settings would be preferable. One possible alteration to this process would be to amalgamate primary and secondary schools into one data set, thus providing additional data points for greater accuracy. However, looking at the differences in percentage increase in EP sessions between secondary and primary schools and the differences in needs between KS1/2 and KS3/4 learners, it seems appropriate to maintain separate modelling for the primary and secondary schools sectors.

Fifth, variability in school structures are not accounted for across the primary and secondary sectors, for example, “through schools” (KS1 to KS4) may require different levels of service with reduced concerns around transition work at the end of KS2. Sixth, the participating EPS in this research is a “partially-traded” service; modelling adaptations may be needed to replicate findings to different EPS service models where, for example, LA statutory work may be incorporated with other services to schools.

Implications for practice and policy

In this research, a framework has been developed and demonstrated for gathering data on the need for EP services in primary and secondary schools that could be replicated and scaled up nationwide to provide an understanding of demand across all areas of EP work, including Early Years, the 16–25 age group and the special schools sector. However, if there were sufficient EP practitioners to meet higher levels of need indicated in this research, it is possible that schools or LAs would need increased, “ring-fenced” funding to enable purchasing. The main implication for policy at government level is that the estimates suggest a need for approximately twice the current staffing of EP service in England (although further evaluation of demand and consideration of skills mix is needed – see below). At a local level, this research’s framework and model has the potential to be used by EP service leaders to gather data when making the case for local service expansion.

Future research

In respect of the initial, broader question about sufficiency of EP services across the children’s services sector, there remain other significant areas of EP work for investigation regarding required EP workforce, including replication and extension of the current findings to confirm a reliably representative national picture. The current research has investigated EP traded work with schools which does not include statutory SEND work commissioned separately by the LA, for which increased national demand is reported (ONSa Office for National Statistics, 2022). Other areas of EP work include pre-school (Early Years), post-school (the 16–25 years age group), special schools, and non-school traded work; for example, The House of Commons Education Select Committee (2022) recently recommended that all prisoners be assessed for learning needs, which would likely involve EPs working closely with The Ministry of Justice. There should also be some consideration of the views of service users such as children and families who do not currently directly commission EP work.

When estimating the EP workforce required to meet identified demand, a more comprehensive model than the simple prototype this research has created will better inform local recruitment and annual training requirements; for example, Hooper (2023) found that most workforce planning tools used in healthcare project service need over a 10-year period presented alternatives to furnish a more comprehensive workforce planning model. In the case of educational psychology this might include: feedback loops; calculation of differences between rural and urban populations; workload intensity and efficiency measures that are known to affect productivity (cf. Jackson et al., 2015); consideration of full-time versus part-time staff and numbers leaving the profession; service structure efficiency regarding skills mix, including the use of trainee/assistant/locum EPs; and consideration of potential system “shocks”, such as rapid increases in demand for EHC needs assessments.

When research findings are combined and extended to form a national picture there are four factors to consider. First, this research has estimated perceived need for EP services; whilst this has been used as a proxy for demand, the actual national demand for EP services will need to consider to what extent need is converted into demand. Second, consideration regarding different models of service delivery is required; supply of EP services differs widely between English LAs. Third, the seven key considerations identified by Hooper (2023) will need to be incorporated into a national workforce planning model: one of these insights relates to including feedback loops in the model; one such example that must be addressed is the impact of high EHCP workloads which reduces capacity for early intervention work within LAs. Fourth, the recent report on national EP services and workforce (DfE, 2023) stated there were insufficient data to create a workforce projection model: this is to be expected when undertaking workforce modelling (O’Malley et al., 2020); a reliable method of estimation will be required, especially regarding the size and characteristics of the available workforce and inflows/outflows to the profession (DfE, 2023).

Disclosure statement

No potential conflict of interest was reported by the author(s).

Correction Statement

This article has been corrected with minor changes. These changes do not impact the academic content of the article.

Additional information

Funding

This project was funded through England’s Department for Education (DfE) National College for Teaching and Learning (NCTL) ITEP award 2020-2022.