A systematic literature review identifying inconsistencies in the inclusion of subjects in research reports on STEM workforce skills in the UK

Abstract

The UK’s STEM skills gap is a pervasive issue, manifesting as a marked shortage of skilled workers in these sectors. This shortage poses significant challenges for employers, who find it increasingly difficult to fill job vacancies with qualified candidates. The gravity of this problem has not gone unnoticed, with the government launching initiatives to bridge the gap. These initiatives range from increased funding for STEM educational programmes to the active promotion of apprenticeships that provide hands-on training and experience. However, as efforts to address the gap intensify, there have been conflicting reports on whether a STEM skills gap really exists. To gain deeper insights into this issue, we employed the PRISMA protocol, a recognised methodological approach, to curate relevant research reports for a systematic review. Our meticulous and critical evaluation of these reports aimed to discern their respective definitions of STEM and to tackle the overarching question of what truly constitutes STEM in academic and policy contexts. Our analysis revealed that there are inconsistencies in definitions of STEM and this can lead to interpretations that vary widely, and sometimes conflict with one another. We found different research reports include different subject areas under the STEM umbrella. Recognising the potential implications of such disparities, we advocate for standardisation in the way we cluster STEM skills/subjects. Such an approach not only promises clarity but will also pave the way for coordinated effective interventions. We identify and discuss potential avenues for future research and believe the paper will resonate with researchers and policymakers.

Keywords:

• STEM
• skills gap
• school
• higher education
• policy

PUBLIC INTEREST STATEMENT

The UK faces a growing problem: there aren’t enough people with skills in science, technology, engineering, and maths (STEM). Employers struggle to find qualified workers, and although the government is trying to help by investing in STEM education and promoting training programmes, a new challenge has emerged. There isn’t a clear agreement on what “STEM skills” means in the UK. Our study carefully reviewed many research reports to understand how STEM is defined. We found that these definitions can be very different and sometimes even contradictory. This confusion can hinder efforts to solve the STEM skills shortage. We believe it is vital to have a clear and unified understanding of STEM. A consistent definition will help improve educational and training programmes, ensuring they meet the actual needs of employers and the wider community

1. The STEM skills gap in the UK

Empirical evidence suggests that there is a skills gap in Science, Technology, Engineering, and Mathematics (STEM) in the UK, particularly in certain industries and regions (Edge Foundation, 2018, Wong, 2016). According to a report by the Social Market Foundation, the UK faces a shortage of workers with STEM skills, which is costing the economy £1.5 billion per year in recruitment, temporary staffing, and additional training costs (Rens, 2015). The Royal Academy of Engineering (RAE) found that the UK will need 1.8 million more engineers and technicians by 2025 to meet the demands of the economy (Royal Academy of Engineering, 2018, 2021). A survey by the IET (2023) found that 62% of employers believe that the skills shortage is having a significant impact on their business, and 68% believe that the shortage is set to worsen over the next few years. A report by the House of Commons Science and Technology Committee (2018) identified that the UK has a low level of STEM skills compared to other countries, with a high proportion of adults lacking basic digital skills. The UK Commission for Employment and Skills has also found that there is a shortage of STEM skills in certain regions, particularly in the North East and North West of England (UKCES, 2015).

This STEM skills gap makes it difficult for employers to fill job vacancies in these areas. The UK government has identified this as a priority issue and has implemented various initiatives to address the skills gap, including increasing funding for STEM education and training programs, promoting apprenticeships, and encouraging more students to pursue STEM careers (Brown, 2022; DfE, 2022; HEFCE, 2017). According to the Royal Academy of Engineering (2021), the need for more engineers in the country, is particularly urgent in certain sectors such as construction and manufacturing. Additionally, there need to be more computer science and information technology professionals (RAE, 2021). Overall, the evidence suggests that the STEM skills gap in the UK could have significant economic implications if it is not addressed. This can make it difficult for companies in these fields to find the talent they need to grow and innovate. Efforts are being made to address the gap, such as the government’s recent announcement of increased investment in STEM education and training.

While there is some evidence suggesting a STEM skills gap in the UK, there are also some studies and reports that challenge the notion of a significant gap (Smith & White, 2020; White & Smith, 2022). For example, a report by the Chartered Institute of Personnel and Development (CIPD) found that the perceived skills gap in STEM may be more of a perception issue rather than a real problem. The report suggests that the issue may be more about employers failing to recognise the value of existing qualifications and experience of potential employees (CIPD, 2017). Another report by the Higher Education Statistics Agency (HESA) found that the number of students studying STEM subjects at UK universities has increased significantly over the last few years. The report shows a 25% increase in the number of students studying computer science and a 13% increase in engineering and technology courses (HESA, 2022). Similarly, the UK’s National Audit Office (NAO) has also challenged the idea of a significant skills gap, citing the need for a more nuanced approach to skills shortages (Morse, 2018). The Morse (2018) suggests that there is a need to focus on specific skills shortages in specific industries, rather than a blanket approach to STEM skills shortages. The Organisation for Economic Co-operation and Development (OECD) has highlighted that the UK has a higher proportion of STEM graduates compared to many other developed countries (OECD, 2019). Additionally, the UK also has a higher proportion of people in STEM occupations compared to the European average (OECD, 2019). In conclusion, while there is some evidence to suggest a skills gap in STEM in the UK, there are also studies that challenge this notion. It is important to consider both sides of the argument and take a nuanced approach to the issue.

Having reviewed these reports, one notable observation was the variation in how different research reports, policy documents, and journal articles define STEM subjects. This inconsistency in defining STEM inevitably leads to varied interpretations of STEM skills, ranging from paucity to abundance. The rationale for conducting this systematic review was to ascertain if there are any inconsistencies in how STEM is defined in the UK and then to suggest a unified approach to address these issues.

2. Methods

We followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) protocol to ensure comprehensive, transparent, and replicable methodologies (Moher et al.). It provides a structured approach with a 27-item checklist and a flow diagram, ensuring comprehensive reporting of all aspects of the review process, from study selection to data synthesis. PRISMA is essential for reducing bias, enhancing reproducibility, and facilitating the comparison of different studies. Its widespread acceptance and endorsement by scientific journals underscore its role in elevating the quality and reliability of systematic reviews, making it a preferred choice for researchers in various fields. The stages of our review process were adapted from Bennett et al. (2005), Randolph (2009), and Martin-Paez et al. (2019), as detailed below.

2.1. Search strategy

A comprehensive literature search was conducted using multiple databases, including Google Scholar, ERIC and Education Research Complete. Search terms related to the specific topic of the review—such as “post-16 attainment,” “participation and skills,” and “workforce in STEM”—were used. We limited the search to studies published in English between 2010 and 2023. Additional reports known to us through our previous work supplemented these searches. Variations of the search string are summarised in Table 1.

Table 1. Search strategy

Research question	Keywords	Search string
How do different research reports, policy documents, and journal articles define STEM in the UK?	- STEM - UK - Define	(“STEM” OR “Science, Technology, Engineering, and Mathematics”) AND (“skills” OR “workforce” OR “work force”) AND (“definition” OR “defining” OR define) AND “UK” AND (“research report” OR “study” OR “article”)

2.2. Study selection

We retained studies that met the following inclusion criteria:

1. Studies conducted in the UK or relevant to the UK context.

2. Studies published in peer-reviewed journals, policy reports, or those published by learned societies or public bodies, including PhD theses.

3. Studies written in English.

4. Where different versions of the same study, such as a PhD thesis, conference proceedings, and journal articles were available, we included the latter.

5. Studies published between 2010 and 2023, with the cutoff date for screening being August 20, 2023.

Additionally, we scanned the websites of some public bodies to identify research reports and official statistics relevant to the paper. Reports known to us from our previous work in the area were included. We sought both a large body of literature addressing the topic and a smaller subset presenting conflicting views, specifically to examine how they define STEM subjects.

Studies were excluded if they were not directly related to the review topic or were conducted outside the UK. Also, studies focusing solely on science and math skills or vocational qualifications were excluded, as we aimed to compare various definitions of “STEM” used in these reports. Studies published in languages other than English, as well as books and conference proceedings, were also excluded (See Figure 1).

Figure 1. Papers screened for the review.

2.3. Data extraction

We shortlisted 20 research reports for inclusion in this review. Data were extracted from each study using a standardised data extraction form, which included information on the study as well as the definition of STEM used in it. A summary is provided in Table 2. Data were extracted from each study using a standardised data extraction form, which included information on the study as well as the definition of STEM used in it. A summary is provided in Table 2. We conducted a pilot test to ensure the reliability and validity of the data extraction form. The extracted data were analysed and synthesised by grouping studies according to themes, identifying gaps in the literature, and highlighting key findings. A sample of studies was double-screened by the authors to assess inter-rater reliability.

Table 2. Final list of papers shortlisted after screening for data extraction

Study ID	Title	Country of study	Definition of STEM
Jebsen (2022)	“It’s a whole load of not giving up”: individual and organisational factors that contribute to UK based women achieving professor status in STEM	UK	STEM = HESA JACS codes by subject areas Medicine and Dentistry, Subjects allied to medicine, Biology, Veterinary, Agriculture and related subjects, Physics, Maths, Computer science, Engineering and Technology.
Vecchia et al. (2023)	Vertical and horizontal mismatch in the UK: are graduates’ skills a good fit for their jobs?	UK	STEM = medicine, maths, computer science, biology, engineering, technology architecture, agriculture, environment related, Physics
Archer et al. (2012)	Science aspirations, capital, and family habitus: How families shape children’s engagement and identification with science	Other: UK	STEM subjects included in the study were not defined. Just the term ‘STEM’ is used throughout
Smith (2017)	Shortage or surplus? A long‐term perspective on the supply of scientists and engineers in the USA and the UK	UK	STEM = Biology, chemistry, physics, maths and engineering
Scott (2019)	Gender differences in occupational preferences among young people: evidence from the UK	UK	STEM = three groups STEM traditional (ISCO definition of STEM: “STEM skills expected to be held by people with a tertiary-education level degree in the subjects of science, technology, engineering and maths” (European Commission, 2017), STEM strict (Only maths-intensive occupations, or ‘hard’ sciences, including physical sciences, technology, engineering and maths, and excluding ‘softer’ sciences i.e. less maths-intensive sciences, like biology and chemistry (Gabay-Egozi et al., 2015; Jerrim & Schoon, 2014; Lauermann et al., 2015; Mann & DiPrete, 2013). Lauermann et al. (2015) use information about the required level of knowledge of mathematics and English for each occupation to create their definition of maths-intensive occupations. This includes occupations like architects, engineers, mathematicians and economists.) and STEM wide (all occupations that requiring some scientific or technical training, not necessarily at degree-level. Bagnoli et al. (2014) suggest two definitions; a wide STEM skills definition, including healthcare occupations, building and construction trades, and midwifery).
Holmes (2022)	STEM skills, technical education and career progression in the uk	UK	STEM = only focusses on technical education and progression
Hoyles et al. (2011)	Supporting STEM in schools and colleges in England: the role of research: a report for Universities UK	UK	STEM abbreviation used throughout not defined which subjects are included
Grinis and Grinis (2019)	The STEM requirements of “Non-STEM” jobs: Evidence from UK online vacancy postings	UK	STEM = Biology, biomedical sciences, physics, chemistry, computer science, technology, engineering, maths, statistics
Tripney et al. (2010)	Factors influencing young people (aged 14–19) in education about STEM subject choices: a systematic review of the UK literature	UK	STEM = Maths, Science, Physics, Chemistry, Biology, technology, engineering
Docherty and Fernandez (2014)	Career Portfolios and the Labour Market for Graduates and Postgraduates in the UK	UK	Graduate and post graduate focus. The criteria for choosing what occupations and percentage of the total employment of STEM graduates. 1). Density - The proportion of STEM degree holders within that sector or occupation; and 2). Proportion - How large employment of STEM in that occupation or sector is as a
Patrick and Smith (2022)	From subject choice to career path: Female STEM graduates in the UK labour market	UK	UKCES (2015) definition of STEM
Smith and White (2019)	Where do all the STEM graduates go? Higher education, the labour market and career trajectories in the UK	UK	STEM = adopted the classification used by the UK Commission for Employment and Skills (UKCES) which uses the criteria of whether an occupation has a highproportion of graduates, a high proportion of STEM-degree holders and a high proportion of STEM-degree holders among graduate entrants.
Zhu et al. (2023)	Genetic basis of STEM occupational choice and regional economic performance: a UK biobank genome-wide association study	UK	Chemistry and related, Biology, Medicine and related, Agriculture, Physics, Geology, Maths, Engineering, Water, Sanitation, Drainage and Public Health
Smith (2011)	Women into science and engineering? Gendered participation in higher education STEM subjects	UK	STEM = UCAS grouping Medicine and dentistry (medical sciences), subjects allied to medicine (allied medical sciences): including nursing and pharmacy, Biological sciences: including biology, microbiology, biochemistry, psychology, sports science and combinations thereof. Veterinary sciences, agriculture and related (agricultural sciences). Physical sciences (physical sciences): including chemistry, physics, physical geography and environmental and forensic science and combinations. Mathematical and computational sciences (mathematical sciences), Engineering and group J technologies (engineering sciences), Architecture, building and planning (architectural sciences), Sciences combined with social sciences or arts (combined sciences and other), Combined sciences (combined sciences)
d’Aguiar and Harrison (2016)	Returning from earning: UK graduates returning to postgraduate study, with particular respect to STEM subjects, gender and ethnicity	UK	The definition of STEM subjects used here is the one followed by the Higher Education Funding Council for England (HEFCE). It covers medicine and medical technology/science, biological and physical sciences, mathematics and computing, and engineering and technology. STEM or non-STEM first degree (STEM): a binary variable denoting whether the content of the individual’s first degree comprised at least 50% STEM subjects (i.e. single subject, at least one subject in a joint degree, the major component in a major/minor combination, or at least two of the three in a triple combination), following HEFCE’s definition (coded Yes); others were coded No
Wilson (2009)	The Demand for STEM Graduates and Postgraduates. London: CIHE.	UK	The Council for Industry and Higher Education (CIHE) considers that sports, science, psychology, architecture and building and planning fall within the category of STEM (CIHE Citation2009), all subjects which HEFCE consider not to be STEM.
House of Commons (2018)	Delivering STEM skills to the economy	England	“STEM stands for science, technology, engineering and mathematics. In education, it means the study of these subjects, either exclusively or in combination.” “In employment, STEM refers to a job that requires the application of science, technology, engineering and mathematics skills or a qualification in a relevant subject, or that is located in a particular industry or sector, such as pharmaceuticals, construction or aerospace.”
UKCES (2015)	Reviewing the requirement for high level STEM skills	England	Scientists, Engineering professionals, IT professionals and technicians, Quality professionals, R&D managers, Production, Health and Safely officers, Environment conservation professionals (Specified clearly that Maths was not included)
DfE (2019)	Attitudes towards STEM subjects by gender at KS4	England	Science, Technology/Mathematics
Okay-Somerville et al. (2020)	Disentangling the impact of social disadvantage on ‘becoming employable’: evidence from STEM student university-to-work transitions	Scotland	Science, Technology, Engineering, Mathematics

2.4. Limitations and elimination of bias

The search strategy employed in this review could have missed some relevant studies, potentially leading to publication bias. Furthermore, by focusing on studies published in English, we may have introduced a bias towards research published in English language. However, as we were only looking at UK focussed research reports, we expect to have identified relevant literature. We controlled for selection and coding biases by independently conducting the search and selection of articles, achieving an 84% agreement rate among authors. Articles were selected based on predefined inclusion criteria and were double-screened. A protocol guided our data collection, and coding was performed collaboratively to minimise coding bias.

3. Findings

3.1. Defining STEM: The need to have a consistent definition

Some research reports show a growing demand for STEM graduates, which the country is struggling to meet (UKCES, 2015; Wakeham, 2016). In order to cope with the required skillset, policy-makers, academics, and stakeholders in the industry have tried to understand the problem and suggested measures from time to time (BIS, 2009; CBI, 2014; UK Parliament, 2022). The government’s objective itself is evidenced by incentives, such as higher bursaries for teachers of STEM subjects (DfE, 2023), financial incentives for students, investments aimed at increasing and widening participation of young people in STEM courses through initiatives and activities planned at the local, regional and national level (BIS, 2009; CBI, 2014; House of Lords, 2012).

The Social Market Foundation report shows young people are reluctant to choose STEM subjects when they are no longer compulsory. This has led to a shortage of 40,000 workers in STEM skills (Codiroli, 2015). The same report predicted that these shortages would increase significantly if steps were not taken to close the gap and urged the government to take immediate action to deal with the predicted STEM crisis, which could likely impact the national economy. However, these claims have been contested by other research reports. They contradict this proposition and suggest that there is no STEM skill shortage (Harris, 2014; Smith, 2017; Smith & Gorard, 2011). In the light of these conflicting claims, it is not clear which of these analysis is more reliable and why there are inconsistencies in the claims being made.

As the debates around skills shortages and subject choices continue, we carried out a conceptual analysis (Hamami & Morris, 2020, p. 3) of the term “STEM” to see if there were any differences in how different reports classify subjects into STEM and non-STEM categories. Our analysis shows that during the last few decades, there has been a lack of consistency in defining what constitutes STEM. This inconsistent application of STEM definitions in research presents not just an interpretation challenge but also leads to confusion and misrepresentation of findings for interventions and practices designed to support pupils in school and beyond (Manly et al., 2018). This misunderstanding/inconsistency is further compounded at different levels of qualifications in schools, vocational and higher education as they cluster different subjects under the STEM umbrella. This also made some of the comparisons between research more complicated and brought apparent incongruity into some of the findings. Our analysis shows that the following main approaches are taken in classifying STEM subjects:

3.2. The UKCES classification

The United Kingdom Commission for employment and skills (UKCES) was an organisation that provided insights and advice concerning skills and employment issues in the UK. It was closed down in 2017, but the legacy of its work and reports continues to be cited and used. UKCES provided guidance on various aspects of the labour market, including the classification of subjects in STEM and non STEM fields. In general, STEM classifications by organisations like UKCES typically include:

1. Sciences: including but not limited to physics, chemistry, biology, environmental science

2. Technology: information technology, computer science

3. Engineering: all branches including mechanical, electrical, civil, chemical, etc.

4. Mathematics: including statistics and data analysis

Non-STEM subjects include:

1. Humanities: English history languages philosophy

2. Social sciences: psychology, sociology, anthropology

3. Arts: Fine Arts, music, theatre

4. Business and commerce: Management, economics, finance

Smith (2019) makes use of the UKCES classification for STEM subjects when analysing the career trajectories of STEM graduates from higher education to labour market. To categorise an occupation as STEM, the adopted UKCES classification uses the criteria of whether an occupation has a high proportion of graduates, a high proportion of STEM-degree holders and a high proportion of STEM-degree holders among graduate entrants. For a more detailed discussion, see also key labour market data for identified STEM occupations in Appendix B of UKCES report (2015). A similar definition of STEM is adopted by White and Smith (2022) when studying career paths (from subject choices) of female STEM graduates in the UK labour market.

3.3. HEFCE Classification

The Higher Education Funding council for England (HEFCE) was responsible for the distribution of funding to universities and higher learning colleges in England. Although the organisation was replaced by the office for students (OfS) and Research England in April 2018, its research and classifications have had a lasting impact on higher education in the UK. HEFCE’s classification of STEM subjects is generally aligned with common understandings of these fields, but the specific classification could vary depending on the context of the researcher report. Generally, STEM classifications in their reports included:

1. Physical sciences: such as physics, chemistry, earth sciences

2. Biological Sciences: such as biology, biochemistry, microbiology

3. Mathematical sciences: including mathematics and statistics

4. Computing: computer science and software engineering

5. Engineering and technology: all forms of engineering as well as technologies related to them

6. Medicine and allied subjects such as Medical Sciences, nursing, pharmacology

These classifications were used for various purposes including the allocation of research funding, education analysis and labour market studies. Their classifications also informed higher education institutions internal policies and funding allocations for STEM versus non-STEM subjects. d’Aguiar and Harrison (2016) adopted HEFCE’s definition of STEM for their study, focussing on the UK graduates returning to postgraduate study from earning. HEFCE’s definition of STEM is different from UKCES in that it covers medicine and medical technology/science in addition to biological and physical sciences, mathematics and computing, and engineering and technology. To classify STEM or non-STEM first degree, d’Aguiar and Harrison (2016) use a binary variable denoting whether the content of the individual’s first degree comprised at least 50% STEM subjects (i.e. single subject, at least one subject in a joint degree, the major component in a major/minor combination, or at least two of the three in a triple combination). Following HEFCE’s definition, these were coded “Yes” for STEM and all others were coded “No” for STEM.

3.4. CIHE classification

The 2009 report from the council for industry and higher education (CIHE), ‘the demand for STEM graduates and post-graduates used a new system of classification for STEM courses (Wilson, 2009). The individual subjects were grouped within the report in the following ways:

1. Medicine: medicine and medical related subjects,

2. All other STEM: Biological Sciences, agricultural sciences, physical/environmental sciences, mathematical sciences and computing, engineering and technology.

For this report, STEM was defined as Medicine plus other STEM. The CIHE considers that sports, science, psychology, architecture and building and planning fall within the category of STEM (Wilson, 2009), all subjects which HEFCE consider not to be STEM.

3.5. Classifying a level STEM subjects

The DfE (2019) report on attitudes towards stem subjects by gender at KS4 included science technology and mathematics-related subjects under the STEM umbrella (DfE, 2019). However, the list of qualifying A level STEM subjects released by the government for the Civil service fast track apprenticeship included the following: science, advanced science, physics, chemistry, biology, maths IT, ICT, engineering, computing, further maths, applied ICT, design and technology (product design), applied science, computer science, electronics, human biology, pure maths, statistics, construction and built environment, electrical engineering, electronic engineering, mechanical engineering, manufacturing engineering, operations and maintenance engineering, pharmaceutical science, vehicle technology and psychology. Though referring to the same qualification, the list of subjects included were different.

3.6. The classic approach

A classic structure posits a set of necessary and sufficient criteria that a qualification should satisfy to be STEM. This is what has been done by most researchers. When operationalising concepts, the focus has often been on the hard sciences (physics, chemistry and biology) and Maths (Grinis & Grinis, 2019; Tripney et al., 2010). Some research reports have included more subject groups with no clear explanation of why they were included or why it was necessary to expand them. One subject which is included in some reports is architecture (Vecchia et al., 2023). Similarly, Zhu et al. (2023) study the genetic basis of STEM occupational choice via a genome-wide association study. In addition to chemistry, biology, physics, medicine, agriculture, geology, maths and engineering, they also include water sanitation drainage and public health under the STEM umbrella. Disentangling the impact of social disadvantage on STEM student’s university to work transitions, Belgin Okay-Somerville et al. (2020) take a similar approach. Furthermore, some reports do not specify which subjects were considered under STEM definitions, this lack of justification makes it harder to interpret the findings (Archer et al., 2012; Holmes, 2022; Hoyles et al., 2011).

This is one of the biggest reasons we see inconsistencies in analysis conducted and conclusions being drawn by researchers despite using the same datasets. This agreement is crucial because, without a consistent definition, there could be problems with comparing findings for outcomes in STEM subjects before university. While the definition of STEM in the higher education sector remains fluid in schools, it invariably refers to science and Maths subjects. The terminology “core STEM” is often used to separate those subjects considered most strongly affiliated to the STEM group, such as Physics, Maths, Further Maths, Chemistry, Computing, ICT, Design and Technology and Other Sciences (Smith, 2011). There have, of course, been other approaches, such as the three-level grouping of Medicine and related STEM, Core STEM and non-STEM from the UK Commission on Employment and Skills, but there is little consistency across the approaches in the way these groupings are allocated.

3.7. UCAS grouping

The House of Lords Select Committee on Science and Technology (2012) suggested that a consistent definition should be agreed upon for Science, Technology, Engineering and Mathematics (STEM) subjects (Lords, 2012). The ambition was to align the higher education statistical agency (HESA), higher education institutions (HEIs), research councils (RCUK) and other professional bodies’ definitions of STEM. For many government reports into participation at HE, STEM was defined as those HE subjects that are in groups A-K in the Joint Academic Coding System (JACS3) as maintained by the Universities and Colleges admissions service (UCAS) (Gill et al., 2018; Pawson & Hulme, 2013; Stagg, 2009). See summary in Table 3.

Table 3. Definition of STEM in higher education

Groups with STEM (Groups A-K)	Non-STEM subjects (Groups L-X)
A - Medicine and Dentistry	L - Social Studies
B - Subjects allied to Medicine	M - Law
C - Biological Sciences	N - Business and Administrative Studies
D - Veterinary Sciences, Agriculture and related subjects	P - Mass Communication and Documentation
F - Physical Sciences	Q - Linguistics, Classics and Related Subjects
G - Mathematical Sciences	R - European Languages, Literature and related subjects
H - Engineering	T - Eastern, Asiatic, African, American and Australasian Languages, Literature and related subjects
I - Computer Sciences	V - Historical and Philosophical Studies
J - Technologies	W - Creative Arts and Design
K - Architecture, Building and Planning	X - Education

Source: Based on JACS 3 listing of the highest-level Groups in HE in the Science and Technology Select Committee report (2012).

This definition is based on subjects taken at university by students aged 18. It allows for consistent comparisons across universities and within the university sector and comparisons with the department of education data. For example, it provides a useful starting point to estimate how many science teachers have a STEM degree or to find out how many students who have been in the care of the Local Authority started studying for a Maths degree at the university.

Despite the initial agreement summarised above, HEIs cluster different subjects under the STEM umbrella (Morse, 2018). However, these definitions vary across HEIs, affiliated government bodies, and subject-based associations (Manly et al., 2018). For example, two of the most commonly misplaced subjects are—Geography and Psychology. Some universities and academics cluster them under STEM subjects (Al Mamun et al., 2015; Caldis & Kleeman, 2019). Others treat them as non-STEM subjects more closely aligned to “category L” in the above table—“Social studies”. Some professionals and educators have clustered them under Humanities and Social Sciences (HASS), while there have been suggestions that these act as bridge subjects between HASS and STEM (Caldis & Kleeman, 2019). There is also danger in using an expansive definition of STEM because a significant proportion of growth or variation reported in STEM uptake is made in subjects with limited STEM content and may not directly contribute to STEM skills. Broad definitions may suggest that progression rates to STEM courses in HE are satisfactory. However, the subject-wise end-of-cycle report released by UCAS shows that Mathematical sciences saw the most significant proportional fall of −9.9% (915 acceptances) to 8,285, the lowest number since 2012, coinciding with the launch of the new Maths A level 6, which saw nearly 6,000 fewer students taking the subject in summer 2019 and this has continued to increase (UCAS, 2021).

4. Introducing the cluster approach

With so many different protagonists with their ideas about STEM, a classic structure of defining STEM is impossible, for example, by aligning it to JACS codes. One of the other options could be to use a cluster approach. For example, there are debates for mathematicians to define what mathematical proof is. Czocher and Weber (2020) offer a cluster definition of mathematical proofs. They argue, “What is novel about our cluster definition is that we are not claiming that any of these criteria are necessary for category membership. The core assumption in the cluster account is that properties ‘count toward’ category membership but that no individual property from the cluster is necessary.” They suggest five properties for mathematical proofs that are the possible foundation for a STEM cluster definition. This proof is a convincing, perspicuous, transparent, and a priori justification sanctioned by the mathematical community. A justification that will remove all doubts is comprehensible and gives the reader an understanding of why it is true. It follows a logically necessary deductive consequence.

One of the propositions we make in this paper towards developing a consistent definition of STEM is by taking this clustering approach. This could be done by listing out what properties “count toward” category membership of STEM. So, for example, psychology might satisfy some of the criteria to be a STEM subject but may appear weaker than mathematics. In HE, where there is such a range of subjects and subject combinations, we could consider some of the following as part of that cluster definition:

• It has a STEM JACS code at University

• It has a “high” proportion of modules underpinned by Maths

• It requires more than one STEM subject at A level/Level 3 as an entry requirement

• It requires Maths or Science at GCSE as an entry requirement

• Participants, graduates or institutions are eligible for Government recognition or funding as a STEM subject

• It appears on a recognised list of high STEM content (such as the WISE core-STEM criteria)

• It is taught by teachers or lecturers with a specialist STEM qualification

• It has a significant number of graduates entering a “STEM profession”

• It would allow a graduate to enter a course in a STEM Initial Teacher Training without completing a Subject Knowledge Enhancement (SKE) course

The suggestions we make are neither all-inclusive nor mutually exclusive. Many subjects include aspects of STEM-related learning, and this contribution to STEM from so many areas of learning makes it such a rich and motivating context for learning. We recognise that the criteria listed above will not remain static. New STEM-related pathways are likely to be developed to meet the changing needs of society and industry, and from time to time, the criteria listed above will need revisiting. STEM definitions in schools and sixth-form colleges are perhaps easier to define as the variety of subjects is limited. Nevertheless, we could include similar cluster definitions concerning the future educational trajectory that STEM subjects support, the definitions used by the DfE when collecting school-wide data and the proportions of mathematical content.

5. Discussion and conclusion

In this paper, we have examined some prominent reports and shown the different ways in which they cluster subjects under the umbrella of STEM. Inconsistencies in definitions of STEM have led to inconsistent findings and hindered effective policy-making. We highlight the importance of having a consistent definition of STEM in research and policy and show how this is important to accurately map participation in STEM fields, particularly for underrepresented groups.

To address this issue, the review proposes a possible way forward. It recommends a collaborative effort among stakeholders, including educators, policymakers, and industry leaders, to develop a comprehensive and inclusive definition of STEM that reflects the current and future needs of the STEM workforce. It is crucial to get the definitions right first. It is possible to increase the number of STEM students by a creeping, ever-broadening definition of STEM at high school and HE (in all its incarnations, such as STEM with the emphasis on the high school subjects, or with Medicine (STEMM) or with Art (STEAM)). Increasing STEM participation in this way is unlikely to encourage policymakers to give sufficient attention to the endemic and institutional barriers to STEM participation. A more robust and transparent definition of STEM will help address this displacement activity and focus attention on those key STEM subjects and skills if that is the policy objective.

Once the definitions have been agreed upon, a new analysis will help us to see whether the STEM skills gap has narrowed down or widened during the last few years. The other aspect that may interest future research will be conversion rates. While there may have been an increase in STEM participation for 16–18-year-olds, these students may not necessarily be making informed choices or pursuing STEM careers. A student studying a single STEM subject at A-level does not necessarily follow-up on this subject for a specialist degree or career. The conversion rate from A-levels to career choices declines for Arts and Humanities, and none of the subjects has a 100% conversion rate; the dip may be more pronounced for STEM subjects. This exploratory analysis will be possible only when we have defined STEM consistently and agreed on which subjects are included in the concept.

Confirmation

All authors have approved the manuscript for submission.

We confirm that the content of the manuscript has not been published, or submitted for publication elsewhere

Disclosure statement

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

Data availability statement

The data used in this paper is from research reports in the public domain so does not involve third party rights

Supplementary data

Supplemental data for this article can be accessed online at https://doi.org/10.1080/2331186X.2023.2288736

Additional information

Funding

This work was supported by the University of Exeter.

Notes on contributors

P Banerjee

Pallavi Banerjee My research focusses on addressing inequalities in educational outcomes for children and young people. I conduct robust evaluations, utilising experimental and quasi-experimental research designs, linked administrative datasets, longitudinal surveys, and systematic literature review to synthesise evidence aimed at improving educational effectiveness and strategic decision making.

Luke Graham

Luke Graham is the PGCE course lead for Science in the School of Education. He has taught in several schools across England for the last 20 years and has been the head of the department and the deputy head teacher.

Gemma Given

Gemma Given is pursuing her MSc in Educational research at the University of Exeter and has an interest in STEM education.