Uncovering the landscape of cross-national UK education research: an exploratory review

ABSTRACT

Background

Internationally, research comparing education systems across countries and jurisdictions is valuable and can elicit nuanced insights into how particular systems operate. This paper’s interest lies in considering the scope and content of research comparing education systems across the four UK nations (England, Northern Ireland, Scotland and Wales).

Purpose

This study sought to determine the coverage of UK cross-national comparative education research (‘home international’ research) between 2000 and 2022. We chose this time period as 1999 marks the devolution of education policy to each UK nation. We aimed to investigate what educational issues had been discussed in the literature and identify any gaps in the content covered by the research.

Method

An exploratory, high-level review of ‘home international’ education research was conducted, based on the review of abstracts. We searched several research databases using a variety of keyword combinations to identify relevant literature published since 2000. Our search identified 53 studies that met our selection criteria. Using a meta-synthesis approach, we coded the content of each abstract to build a picture of the range and thematic coverage of research involving comparisons between at least two of the four UK nations.

Findings

The analysis of abstracts identified that, over the last two decades, UK ‘home international’ research has tended to include comparisons of all four nations, coverage of multiple educational phases and a focus on national education policy reviews. Furthermore, we pinpointed a number of gaps in coverage that might not have been anticipated (e.g. relatively little cross-national research focusing on assessment).

Conclusion

This high-level review uncovers the landscape of recent ‘home international’ research, allowing us to view issues that are driving the cross-national research agenda in the UK and recognise implications relevant to education systems that may resonate with jurisdictions beyond these four UK nations.

KEYWORDS:

• Home international research
• comparative study
• United Kingdom
• education systems
• educational phases
• policy

Introduction

Comparative, cross-national study is a well-established approach to educational investigation. There is a strong tradition of comparison being used as a way of building a picture of education systems. For example, it can facilitate description of particular features of national systems through comparisons across different jurisdictions (Bray 2023). It has also been employed to establish the relative performance of a given education system through the comparison of standardised test outcomes across nations (e.g. Auld, Rappleye, and Morris 2019). Importantly, comparison does not have to be restricted to cross-national cases, reinforcing the need to consider appropriate units of analysis (see Farrell 1979). Although the state, or nation, has long been recognised as a primary unit for governing social, economic and political life across many contexts (Green 2013), it is a contested concept. An alternative to cross-national comparison is the comparison of subnational units, such as federal states or regions. For instance, comparative studies have considered the distribution of educational opportunities across semi-autonomous administrative regions in Ethiopia (Tesema and Braeken 2018), and racial, ethnic and gender differences in school discipline approaches across federal states in the USA (Wallace et al. 2008). Sellers (2019, 86) draws attention to the ‘subnational turn’. This form of study can be useful for considering differences across territorially distributed units within countries (e.g. ‘state’ institutions such as prisons, located in different federal regions), or subnational units that exist in more than one country (e.g. trade unions).

In the United Kingdom (UK) context, interest in this area was galvanised by a paper (Raffe et al. 1999) that called for ‘home international’¹ comparisons in the UK. Inspired by the name of a series of friendly sporting events that traditionally involved England, Northern Ireland, Scotland, and Wales, Raffe et al. (1999) compared the education systems across these four nations. This comparison rested on some key dimensions (e.g. governance, institutions and curriculum) to derive a more nuanced understanding of how each system worked and to inform policymaking within each nation. Raffe et al. (1999) appears to be a watershed moment for cross-national education research in the UK. The paper argued that, prior to this work, there had been little systematic research that reviewed the four education systems of the UK. Moreover, it observed that when such research was carried out, there was a tendency for it to focus on England or treat the UK as one country instead of four individual nations.

With this in mind, a key question is how cross-national UK education research might have evolved since Raffe et al. (1999). This motivated us to conduct an exploratory meta-synthesis of more recent research in the field of education across the four UK nations. As meta-syntheses can bring together qualitative data to aid interpretation of the research field of interest (Temple University Libraries n.d.), it might allow us to uncover the landscape of cross-national UK education research and better understand what sorts of topics and issues were being investigated through the approach. It would also provide a means by which to identify any gaps in the field that might alert us to where education research might be needed in the future. Before presenting our study in further detail, however, we contextualise our work by considering the notion of comparison in education.

Background

Comparison in context

Comparison-making has an important function in educational science. According to Farrell (1979), single observations cannot prove a proposition in a complex social situation. Contextual comparison can facilitate a better understanding of the conditions in which observed phenomena occur: for instance, in educational situations it can help tease out whether observed features are associated with learner demographics or otherwise. Comparison may assist us to notice salience, which brings us to consideration of the distance between cases of comparison. It should not be assumed that close cases are necessarily the best option for comparative study design. On the one hand, most similar case comparison can allow linear causality between factors to be studied, since too much variability can lead to noise that may result in rich contextual detail in a particular locality/system being overlooked. On the other hand, some have argued that the demands for close case comparison may be problematic, as ‘similarity’ is a construct imposed on the study by the observer and not inherent to the objects being studied (Farrell 1979). Such debates reinforce the need for study designers to make clear their underlying assumptions and rationales when developing comparative studies.

Cross-national comparison in the UK context

In the UK, education policy became devolved across the four nations at the end of the 1990s (Sibieta and Jerrim 2021). In the period prior to and immediately following devolution, it was noted that there was a tendency in education research to focus heavily on highlighting the conditions of education in England over the other three nations (C. Donnelly and Osborne 2005; Raffe et al. 1999). This over-representation of England in policy analysis meant that ‘traditional’ analyses of social policy neglected the local and regional expressions of, and variations to, government policy throughout the UK. Local, regional, and (in the case of the UK) cross-national influences on policy are important to identify, as insight may be gained into how the systems of the UK work. This is particularly significant when considering education policy, because it is broadly recognised that the systems have many commonalities, rendering them ‘close cases’ and making the use of comparative analysis particularly useful. For example, Paterson and Ianelli (2007) observe that the different education systems across the UK nations share important structural features. The phases of education are similar across the four nations, and these feed into broadly similar social and economic structures (e.g. there is a high degree of labour mobility across the national borders of the UK, with mutual recognition of educational achievements). This point is reinforced by Raffe et al. (1999), who highlight that, although the systems of administration are different across the UK nations, there are similarities in the social contexts and social relations that do not hold when comparing the nations with other non-UK nations. These similarities include the UK nations having similar levels of social mobility and of class inequality in education, and similar profiles of gender differences in education (Raffe et al. 1999). This suggests that the UK systems have an interdependence, with the four education systems interacting with an economy that can be regarded as integrated and organised at a UK level.

Given this contextual similarity, Croxford and Raffe (2014), Phillips (2003) and Raffe et al. (1999) argue that a cross-national comparative approach can enable researchers to identify policy outcomes that are achievable in similar circumstances, as well as to observe how different nations deal with similar issues that they are experiencing. For Paterson and Ianelli (2007), this ability relates to how UK cross-national research conforms to a ‘most similar comparisons’ (Paterson and Ianelli 2007, 332) research design in political science. The approach better allows researchers to test for the independent effects of different systems of education, or different educational policies, than when these variables are confounded by multiple and ‘noisy’ features of the economy or society.

In line with the idea of comparative research as policy lab, Avis et al. (2012) indicate how cross-national research can provide insight into the ways that governments deploy differing strategies to address similar problems. As Raffe (2005) suggests, this may have policy implications, as the similarity of different systems could mean that the lessons from such observations have transferable qualities across borders. Despite its clear potential for learning about ‘what works best’ through exploring the use of different approaches to tackle similar problems in parallel, Paun et al. (2016) note that cross-national comparative research approaches are underutilised. This observation spurred us to look more closely at the state of cross-national research in the UK since the original work by Raffe et al. (1999).

In their opening sentences, Raffe et al. (1999, 9) state that ‘the differences among the four education systems of the UK are often perceived as a nuisance by comparative researchers … they are also an opportunity’. This perfectly sums up the inspiration for our exploratory study – the opportunity to uncover the landscape of cross-national UK education research over the last two decades, identifying the ebb and flow of thematic emphases, and any noteworthy gaps apparent in research.

Purpose

We sought to determine what content was covered in UK ‘home international’ education research literature published between 2000 and 2022. When we planned our study, we aimed to explore, via a high-level review of article abstracts: (a) the coverage of research; (b) the status afforded to different educational issues and (c) any gaps in content. This overview would help us to better understand which issues were considered important in recent educational research, what appeared to have less coverage and whether there were any particularly significant gaps in the content coverage that could inform future research activity. Our project was guided by seven research questions relating to coverage during the time period: (1) What are the characteristics of examined studies? (2) What is the content of the studies? (3) Which phases of education are the focus? (4) Which nations are the focus? (5) How do content and nation interact? (6) Does content differ across time in published studies? (7) Do content, time, and nation interact in published studies?

Method

Search approach and criteria

We carried out a meta-synthesis (Temple University Libraries, n.d.) of cross-national research involving the four UK nations (England, Northern Ireland, Scotland and Wales) between 2000 and 2022. This meant capturing the main content of research literature in this area of education study. We chose to examine abstracts rather than whole articles because our aim was to build a high-level picture of the range and coverage of this field of study. The use of abstracts would, thus, facilitate an overview of study content. This type of approach has been employed elsewhere; for example, in design-based learning (Anderson and Shattuck 2012; Mangiatordi and Serenelli 2013).

Increasingly used in automated systematic reviewing processes, the benefits of abstract searching include the relative rapidity with which information may be gathered (Choong et al. 2014). Abstracts are useful for establishing a valid overview of a field of study, with some experiments suggesting that abstract-only searching does not consistently yield lower effectiveness compared with entire-article searching (Lin 2009). The speed of abstract-only searching means that it is particularly suited to describing the evolution of research approaches over substantial periods of time and has been utilised in medical research contexts (e.g. Stang et al. 2017). Additionally, abstract-only searching may be regarded as more inclusive, allowing review studies to include evidence where only abstracts are available, creating a level playing field with other papers so that comparisons are based on standard units. Abstract searching requires that researchers have a clear sense of the dimensions of interest that drive the search process (e.g. for us, this included the four UK nations and educational phases). However, despite the benefits, it is acknowledged that abstract-only searching inevitably has limitations, too, particularly in relation to concerns about abstract quality. Abstracts may not contain all data covered (Saric et al. 2020): for example, the abstract might not mention all the education sectors that are expanded on in the full study. It has been noted, too, that the quality of abstracts can be inconsistent (Mangiatordi and Serenelli 2013), and that users searching full texts are more likely to find relevant articles than users searching abstracts (Lin 2009). We discuss the advantages and limitations of abstract-only searching more fully later in this paper.

Boell and Cecez-Kecmanovic (2015, 49) suggest the importance of ‘protocols that prescribe how to identify, select, assess and synthesise evidence from the literature’ in systematic literature reviews, thus highlighting the need to consider whether the collected literature meets the criteria set out to address the research question. We inspected several academic databases and search resources, including ERIC.ed, Science Direct, Google Scholar and ResearchGate. The following keywords were used to search for literature: united+kingdom+education; united+kingdom+curriculum; united+kingdom+curriculum+comparison; curriculum+in+uk+nations; comparison+of+uk+nations; uk+nations+educational+comparison; uk+nations+education; Ireland+England+Scotland+Wales+education; England+Northern Ireland+Scotland+Wales+comparison; England+Scotland+Wales+Northern Ireland+Education+comparison. We adopted strict criteria for the selection of literature. To be included in the selection, the literature outputs had to be journal/academic articles, book (chapters), reports, or grey literature² relevant to the UK home international comparison field, authored by professionals or academics. The literature had to be published between 2000 and 2022 and had to compare at least two of the four UK nations (e.g. Northern Ireland and Wales; England and Scotland; and Northern Ireland, Scotland, Wales). Literature had to be written in the English language. Each item of literature meeting the criteria was added to an electronic spreadsheet, along with the following metadata: the author(s), year, title, abstract (the unit of analysis adopted), content codes (codes) (explained below), notes (important information about the research), and source (which academic database or search resource the article came from).

Coding and analysis

Both authors (here, referred to as Researcher 1 and Researcher 2) were involved in the analysis. Researcher 1 searched through the identified databases and search resources to locate literature that appeared to meet our selection criteria, and Researcher 2 checked the selected abstracts for relevance. We initially recorded 57 abstracts and their metadata on the spreadsheet. Upon closer inspection, though, we found that four abstracts did not quite meet the criteria, as the content went beyond our focus on education (e.g. covering the financial situation of universities). We therefore discounted these four abstracts, which brought the total number of abstracts included in the analysis down to 53. The 53 studies which were included in our analysis of abstracts are listed in Appendix A and indicated in the references.

Using an electronic spreadsheet as our data organisation tool allowed us to view the abstracts and associated data in chronological order, when needed, and compare the codes assigned to each abstract. Once all the abstracts were represented on the spreadsheet, we read them through to get a sense of what each paper was about. An approach informed by the principles of thematic analysis (Braun and Clarke 2006) underpinned our investigation of the abstracts. We applied content codes to communicate the themes of the abstracts: these were phrases or words which broadly encapsulated the topics (e.g. ‘Covid’; ‘curriculum document’; ‘standards’). During this process, the abstracts were viewed in random rather than chronological order, so as to reduce any possible impact of bias or prior knowledge; we attempted not to look at the year of each abstract.

In total, the 53 abstracts went through four rounds of coding. Some content codes came from previous work in this space (Mouthaan et al. 2021), but, as more abstracts were analysed for themes, new content codes were added to the database if it seemed that existing codes did not capture the newly arising themes adequately. In addition, as the process of coding progressed, some of the initial content codes were revised. For instance, some content codes were combined if they expressed a similar theme (e.g. ‘educational policy’ and ‘policy review’ were collapsed into ‘national review’), whilst other codes were divided into two separate codes to better express the specific emphases of abstracts, where needed (e.g. ‘devolution’ and ‘devolution impact’). This procedure resulted in a total of 97 revised content codes that could be applied to abstracts in the final round of coding. We also kept a separate note detailing which abstracts had seen new content codes introduced, as a way of recording our processes and thoughts whilst we added new codes. This enabled us to be transparent about what we did and why, and to record the active analysis in which we had engaged (Braun and Clarke 2006).

During the first round of content code application, we read the abstracts carefully and assigned codes to the key themes. As noted above, if any theme was not adequately captured in the existing content codes, we added a new code. In the second round, Researcher 1 applied codes based on the International Standard Classification of Education (ISCED) framework³ to the abstracts and checked the content codes that had been applied in the first round. This standardisation process ensured reliability and allowed identification of any codes that may have been missed in the first round. In the third round, Researcher 2 cross-checked 25% of the coded abstracts and ISCED codes. The two researchers met between the first and second rounds of coding to discuss any instances where there was uncertainty over codes. After the third round, the researchers met again to discuss whether the codes were consistent with those applied by Researcher 1 in the earlier phases. In round 4, Researcher 1 applied the final list of content codes to the abstracts one more time and referred to the codes that Researcher 2 assigned in the third round, in cases of uncertainty. Once all abstracts were coded with the final content codes and ISCED codes, we were able to analyse the frequency with which codes occurred in the data and thereby identify emphases and prevalent themes in the abstracts.

Findings

Through the analysis of abstracts outlined above, we were able to address our research questions. This allowed us to gain broad insight into the characteristics of the included studies, based on the information in the abstracts. It enabled us to build a picture of the UK cross-national education research in terms of content and themes, and the education phases and nations covered, as well as combinations of these aspects. In the following paragraphs, we present the main findings.

In terms of output type, the majority of the 53 included studies were journal articles (68%). There were also reports (17%), forms of grey literature (6%), book chapters (6%), blog posts (2%) and seminar papers (2%) (see Table 1).

Table 1. Output type across the 53 included studies.

Output type	N studies (total = 53)	Proportion of studies (%)
Journal article	36	68
Report	9	17
Grey literature	3	6
Book chapter	3	6
Blog post	1	2
Seminar paper	1	2

Percentages do not sum to 100 due to rounding.

During our exploration of the themes of the included studies, we reviewed all content codes that had been assigned to the abstracts in the final coding round. To gather a sense of which areas of content were most salient, and thus could be considered themes, we identified content codes that had appeared at least five times across the set of 53 abstracts. In total, nine content codes passed this analytical threshold and were considered to have qualified as themes (see Table 2).

Table 2. Most prevalent themes across the set of 53 abstracts.

Theme (content code)	Definition	Number of abstracts assigned content code (total abstracts = 53)	Proportion of abstracts (%)
National review	Stated or implied review of educational policy.	21	40
Inequality/inequalities	Reference to any type of inequality (including but not limited to equality, diversity and inclusion, social class, deprivation, disadvantage, vulnerable learners).	10	19
Devolution impact	Reference to the effects of devolution on areas such as education or policy.	10	19
Post Compulsory Education & Training (PCET)	Reference to PCET, including Further Education, Higher Education, and post-compulsory vocational training and apprenticeships.	9	17
Devolution	Reference to devolution.	9	17
Methodology	Reference to study methodology (e.g. novel method uses).	6	11
Education system overview	Reference to education system review and/or comparison (e.g. structural comparison of types of educational institutions).	5	9
Covid	Reference to the COVID-19 pandemic.	5	9
Teacher education	Reference to Initial Teacher Education (ITE) programmes, teacher training, school-based teacher training and teacher educators.	5	9
	Total number of content codes assigned	80

Abstracts may contain more than one theme. To qualify as a theme, a content code had to appear at least 5 times across the set of 53 abstracts. Percentages do not sum to 100 due to multiple content codes being assigned to individual abstracts.

Overall, national review was the most popular theme (appearing in around 40% of the abstracts), whilst around 19% of the abstracts mentioned inequalities, devolution impact, and 17% mentioned Post Compulsory Education and Training, and devolution directly. Approximately 11% of the abstracts mentioned methodology. This suggests that a substantial proportion of home international comparisons literature published over the last 20 years was interested in national policies. Considering the high profile of assessment and accountability in education over the past two decades, it seemed surprising that content codes relating to these areas did not overtly feature in the analysis (see section ‘Discussion’).

The analysis allowed us to determine which educational phases were the focus of the included studies, based on the coding of abstracts.

As indicated in Table 3, the vast majority of the analysed abstracts (68%) tended to mention at least two phases of education (cross-phase). Post-secondary education was the most commonly mentioned education phase (23% of abstracts), with primary education and upper secondary education each accounting for around 5% of abstracts. According to the analysis of abstracts, none of the included studies took early childhood education, pre-primary education, or lower secondary education as a focus.

Table 3. Educational phases referred to across the 53 abstracts.

Educational phases (codes)	Number of abstracts (total = 53)	Proportion of abstracts (%)
01: early childhood	0	0
02: pre-primary	0	0
1: primary e	3	6
2: lower secondary	0	0
3: upper secondary	2	4
4: post-secondary	12	23
5: cross-phase	36	68

The educational phase codes were based on the ISCED framework and adapted for the purposes of the project. Cross-phase refers to at least two phases of education. Percentages do not sum to 100 due to rounding.

It was also important to explore how these educational phases related to the four UK nations, by establishing the associations between each studied educational phase and the combinations of nations on which they focused (see Table 4). The analysis indicated that the studies across each represented education phase (i.e. primary, upper secondary and post-secondary) included examination of all four UK nations. Cross-phase comparison across all four nations was a feature of 68% of abstracts and there was, too, an evident interest in post-secondary education – again, across all four nations (19% of abstracts). This suggests that literature in the cross-national comparative space since 2000 tended to focus on all four nations across multiple phases of education, with some additional and specific interest in post-secondary education.

Table 4. The interaction between educational phases and nations across the 53 abstracts.

Educational phases (codes)	Nations	Number of abstracts (total = 53)	Proportion of abstracts (%)
1: primary	England, Northern Ireland, Scotland, Wales	2	4
	England and Scotland	1	2
3: upper secondary	England, Northern Ireland, Scotland, Wales,	2	4
4: post-secondary	England, Northern Ireland,Scotland, Wales	10	19
	England and Scotland	2	4
5: cross-phase	England, Northern Ireland,Scotland, Wales	36	68

The educational phase codes were based on the ISCED framework and adapted for the purposes of the project. Cross-phase refers to at least two phases of education. Percentages do not sum to 100 due to rounding.

To investigate which nation(s) were being investigated with which nation(s) we looked at all 53 abstracts ordered by year of publication, and counted which nations co-occurred in each abstract. As shown in Table 5, a large proportion of the included studies (over 70%) investigated all four nations, with 9% investigating England, Scotland and Wales; or England and Scotland. Other combinations were also apparent: for example, one study examined England and Northern Ireland; another study examined England and Wales. However, not all nations co-occurred in all combinations: for instance, no studies investigated Scotland and Northern Ireland together. It is noticeable that England is, by and large, an ever-present comparator nation across the studies, with the exception of two studies that compared Scotland and Wales.

Table 5. Co-occurrence of nations across the 53 abstracts.

Combinations	Number of abstracts (total = 53)	Percentage of abstracts
England, Northern Ireland, Scotland, Wales	38	72
England, Scotland, Wales	5	9
England, Scotland	5	9
Scotland, Wales	2	4
England, Wales	1	2
England, Northern Ireland	1	2
England, Northern Ireland, Wales	1	2

As part of the analysis, we considered which themes were most frequent in relation to the different types of possible nation combinations (i.e. England and Scotland and Wales; or Scotland and Wales). Our analysis showed that it was usual for studies examining the most common themes of interest to focus on all four home nations (accounting for 80% of all codes assigned). It is noticeable, as well, that some themes (Post Compulsory Education and Training, and Covid) were only looked at in a four nations context.

We wanted to ascertain which content codes (themes) were most prevalent in the studies between 2000 and 2022. To this end, we grouped the abstracts by year and analysed content code frequency; the findings are displayed in Table 7.

The content codes in Table 7 are those that were frequently occurring in specific years. If only one study was conducted in a given year, it was not included in the analysis. For instance, there were two studies included in the table for the year 2001 and both covered education system overview and inequality. If multiple studies were conducted in a given year but the content codes applied for each year were different, we noted this as no specific, frequent code for that year. As Table 7 indicates, according to our analysis, national reviews were a focus for four different years (2007, 2011, 2012 and 2021). For two years (2001 and 2021), there was a focus on inequalities. It is unsurprising that Covid emerged as a focus for cross-national study in the 2020–2021 period.

Finally, to explore the main content themes for each year and determine to which nation(s) they related, we analysed the content codes for each year and identified which countries were covered in the studies. Figure 1 shows that most of the studies with the most frequently occurring codes tended to investigate all four UK nations (83%). This is not surprising, given that 38 of the 53 sources we looked at compared all four UK nations (see Table 6). Three other patterns emerged from this analysis: England appeared in all studies, Scotland was a near continual focus of research (appearing in 97% of studies), whilst Northern Ireland was relatively overlooked (appearing in only 86% of the studies).

Figure 1. The most prevalent themes for each year, linked to nation.

EN = England; NI = Northern Ireland; SC = Scotland; WA = Wales.

Table 6. Number and proportion of assigned content codes in relation to nation combinations.

	Themes (content codes)
	Education system overview	Inequality/inequalities	Post Compulsory Education & Training	National review	Devolution	Devolution impact	Teacher education	Methodology	Covid	Total content codes assigned	Proportion of content codes assigned
England, Northern Ireland, Scotland, Wales	4	9	9	16	6	9	3	3	5	64	80
England, Scotland, Wales	1	1		2	1					5	6
England, Scotland				2	1		2	2		7	9
Scotland, Wales						1		1		2	3
England, Wales				1	1					2	3
Total content codes assigned	5	10	9	21	9	10	5	6	5	80

This table presents the number and proportion of themes (content codes) assigned to the 53 abstracts in relation to the combination of nations explored by the abstracts. Abstracts may contain more than one theme. To qualify as a theme, a content code had to appear at least 5 times across the set of 53 abstracts. The percentages represent the proportion of content codes, showing that most content codes (80%) were related to the four nation combination. Percentages do not sum to 100 due to rounding.

Table 7. Most prevalent themes in relation to year of study publication.

	Year
	2001	2002	2003	2004	2005	2006	2007	2008	2009	2010	2011	2012	2013	2014	2015	2016	2017	2018	2019	2020	2021	2022
Theme(content code)	a f					b	c				c	c		d						e	e c f

To qualify as a theme, a content code had to appear at least 5 times across the set of 53 abstracts. a: education system overview; b: devolution impact; c: national review; d: PCET; e: Covid; f: inequality/ies.

Discussion

Drawing inspiration from the work of Raffe et al. (1999), our exploratory review offers insight into the nature of research comparing education across the four UK nations (England, Northern Ireland, Scotland and Wales) since 2000. We felt that the reasons for cross-national comparisons provided by Raffe et al. (1999) were still highly relevant, and that this is supported by the research attention paid to ‘home internationals’ analysis that has prevailed since devolution (e.g. Brisard, Menter, and Smith 2007; M. Donnelly and Evans 2019; Gray, Hooper, and Sandford 2022; Hodgson et al. 2019; Phillips 2003). Through a careful analysis of the abstracts of 53 studies, we were able to gain an understanding of which educational issues had been discussed and whether there were any notable gaps in the research.

Our findings reflect the continued interest in UK cross-national study, with evidence of papers in this area being published throughout the 22-year period covered by our review (with a few noticeable exceptions, discussed in the following paragraphs). The sustained use of this approach may reflect the relative ‘closeness’ of these national cases. The broad, system-level similarities of the home nations outlined by Raffe et al. (1999) still appear relevant (see Hodgson et al. 2019) and means that comparisons between them can result in highly focused analyses that can contribute to a deeper understanding (Krause 2018). This point reinforces the idea that the choice of valid units is central to comparative study and the utility of the method (e.g. see Farrell 1979), and that contextual similarity can allow researchers to observe how different nations deal with the same or similar issues (Croxford and Raffe 2014; Phillips 2003).

It was evident from our analysis of abstracts that much research in this area within our chosen time period was characterised by breadth of scope, underscoring how the approach can serve research with a range of purposes (e.g. Auld, Rappleye, and Morris 2019; Bray 2023). Most studies within our review time period (2000–2022) looked across educational phases rather than concentrating on a specific phase and tended to make comparisons across England, Northern Ireland, Scotland and Wales. This tendency to compare situations across all four UK nations contrasts with the observation relating to an earlier time period that ‘very few [descriptions] compare all four systems’ (Raffe et al. 1999, 9), suggesting that the cross-national UK research field has evolved in this respect in the years since 1999. In our review, it was interesting to note, though, that some nations were represented more than others. Both Scotland and England were heavily represented across the 53 studies, whilst Northern Ireland appeared in the fewest studies. Reasons for representation emphases and gaps are complex, may include factors such as system size, and would need further research to investigate.

Close analysis of the content codes in our study suggested that a key area of interest for researchers between 2000 and 2022 was national education policy review. This is understandable, since the post-devolution period was a time of intense curriculum change. For example, during this time, Scotland implemented its Curriculum for Excellence and Wales was developing a new Curriculum for Wales. However, it is noteworthy that we identified coverage of a range of other issues in the included studies. Whilst some of these could be relatively easily anticipated, such as ‘impact of devolution’, or immediately relatable to specific curriculum developments in a given nation, other issues are more difficult to explain. A case in point is the degree of interest in Post-Compulsory Education and Training, which likely links to the observation that post-secondary education was the single most popular category of focus. Further and deeper analysis would be needed to explain what was driving interest in this area of research during the review period. We need to recognise that our findings are, inevitably, only a snapshot of research interests over a specific period. For example, we would anticipate that cross-national Covid-related research might feature with even more emphasis in future reviews of research outputs, as more data sets relevant to education in the pandemic and post-pandemic years become available for comparative analysis.

We were keen to uncover whether there were any gaps in UK cross-national research, and our findings highlight some of these. Interestingly, some content codes (themes) which we brought into the study from previous work (e.g. studies that contributed to Mouthaan et al. 2021) did not feature as heavily across the 53 studies included in our review. For example, assessment appeared in only two studies, in 2020 and 2021: hence, assessment was not on the table showing the nine most prevalent themes across the set of 53 abstracts (Table 2). Given the importance and impact of assessment, particularly high stake assessments such as GCSE examinations (Jerrim 2022), it seemed surprising that assessment did not feature as a frequently investigated area of interest. Of course, this does not mean that research has not investigated assessment during our chosen time period. All we can conclude, on the basis of our analysis of abstracts, is that a cross-national comparison approach has not tended to be adopted for exploring assessment in the UK over the last two decades. Our inclusion criteria meant that we focused on research that compared at least two of the four UK nations, which automatically excluded any literature involving individual UK nations. Another feature evident from our analysis is that, in certain time periods, there was no published research in this space that met our criteria. This may be a sign that UK cross-national comparison was less popular during these times, or it could simply be due to inevitable lags in publication cycles (i.e. the time it takes from study completion to publication).

Another possible gap in the research relates to the educational phases that did not feature as a sole focus for UK cross-national research in our studied time period. Our analysis suggested a lack of ‘home international’ research that looked solely at early childhood education, or pre-primary education, or lower secondary education across the UK nations. However, given our methodology and the study scope, we cannot definitively ascertain that research in these areas did not exist. Although we used familiar terms to identify educational phases (e.g. ‘early childhood education’), we could have inadvertently missed research using alternative terms. It may also be that some research did not distinguish between lower and upper secondary education and/or included lower secondary phase as part of secondary education in general. Additionally, as the majority of studies dealt with multiple educational phases, the cross-phase code did not identify specifically which phases were investigated by these studies. Sources that investigated multiple phases may have included early childhood education, but the cross-phase code would not allow us to identify the specific education phases.

Implications

Given the study’s findings and context, it would be interesting to explore why certain content themes were investigated more than others (e.g. national review) using a cross-national comparison approach. As mentioned earlier, this approach appears particularly well suited to system-level policy analysis; however, it does not explain why some policies rather than others were of interest during the selected time period. Is it simply because there was interest in those areas – or were other drivers present? Our analysis of data indicated that certain themes were more prevalent in certain years (specifically, education system overview, and inequality/inequalities (2001); national review (2007, 2011 and 2012); Post Compulsory Education and Training (2014); Covid (2020); Covid, national review, and inequality/inequalities (2021)). Future research could investigate why certain content appeared more frequently in these years, with particular focus given to any policy changes or key educational movements and initiatives that may have occurred around those times. Future research could usefully investigate, too, some of the issues identified in this paper in the context of early childhood education and pre-primary education across the UK, which may be under-researched using a cross-national approach. Considering that early childhood is reported as an important time in children’s development (e.g. Barnett Steven and Nores 2012; Britto et al. 2017), the research gaps identified here may benefit from more attention, especially with regards to the key themes identified (e.g. national review of policy and inequality/inequalities). Lastly, it would be useful to establish more confidence in our exploratory findings through triangulating the findings with other methods. For example, an automated high-level analysis of research in this space could be carried out using software tools for bibliometric analyses (e.g. van Eck and Waltman 2013) to see if this produces similar findings to our manual analysis.

Limitations

There are several limitations that need to be acknowledged. The first set of limitations refers to the method of review employed – that is, the analysis of abstracts. As abstracts may not always adequately communicate the full meaning and content of the article, we sometimes had to skim-read the body of an article in addition to the abstract, in order to apply the relevant content codes. Therefore, despite our chosen approach providing, overall, a useful and informative overview of the field, it is possible that some themes could have been missed. Variation in abstract quality is recognised in other research (e.g. Mangiatordi and Serenelli 2013; Saric et al. 2020), and must be borne in mind when considering our findings. Another methodological issue was that some of the output types (e.g. items of grey literature) did not have abstracts. In these cases, we had to read the full article and summarise it before we could apply content codes. This inconsistency is a limitation since it means that we were unable to treat all studies in our sample in the same way.

Furthermore, during the coding process, we added new content codes and collated others. It could be argued that some content codes were too similar or overlapping, and therefore that they could have been collated or deleted (e.g. devolution and devolution impact). However, methodological issues are to be expected in exploratory studies and can be used to inform the design of future research. Additionally, the approach we employed was tied to the purpose of our study – to build an overview of UK cross-national education research in our chosen time period. It was, therefore, appropriate to focus on the abstracts, with this larger, broad-brush picture in mind. Although we had started by reading the articles in full, we felt this may, in fact, distract from the ability to build a general sense of what each research output was focusing on. Therefore, we considered the nature of the units that were compared and felt that the abstract-only approach helped us fulfill the aims of this study.

We are mindful, too, that our definition of frequently occurring content codes influenced outcomes. As reported earlier, we defined codes which came up at least five times throughout the 53 abstracts as frequently occurring. This raises some interesting points, such as who decides what the baseline for ‘frequency’ should be, and what it says about the status of themes that were only found in four documents. More broadly, it highlights the often subjective and sometimes arbitrary aspects of research design and, importantly, reinforces the need for researchers to make their methodological choices and decisions transparent so that these can be taken into account and discussed in the interpretation of findings. Taken together, these limitations mean that we need to be tentative about the implications that we draw from the study. At the same time, we would argue that the study’s value lies in the way it draws attention to areas for further exploration and flags potential changes over time – both important elements in this field of inquiry.

Conclusion

This study stemmed from an interest in exploring the scope and content of research comparing education systems across the four UK nations (England, Northern Ireland, Scotland and Wales) in the years following Raffe et al. (1999) influential work in this area. Our findings suggest that researchers have continued to recognise the worth of generating ‘home international’ comparisons and the value of treating the UK home nations as examples of close cases that can function as useful bases for exploring approaches to shared issues of concern. The analysis of 53 abstracts also uncovered some potential gaps in UK cross-national research activity (e.g. comparisons involving assessment; early years), which could help inform plans for future research. Raffe et al. (1999, 22) proposed that ‘home international research needs a historical perspective and a capacity for analysing social change if it is to understand these developments’. This connects with our view that it would be beneficial to explore the reasons behind the prominence of certain themes in ‘home international’ research in particular years between 2000 and 2022, thus contributing to a deeper understanding of the education policy landscape in the early twenty-first century. In all, we concur with Raffe et al.’s (1999, 22) statement that ‘there is a case for increasing attention to home international comparisons’, because ‘such comparison can serve a range of scientific and policy-related goals, and are of interest outside the UK as well as within it’.

Disclosure statement

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

Notes

1. ‘Home internationals’ (see Raffe et al. 1999) is a term used to refer to the four constituent nations of the UK. The term was synonymous with a series of football tournaments that took place involving England, Northern Ireland, Scotland and Wales between 1884 and 1984. The term has been adopted by many researchers interested in cross-national comparison work in the UK (e.g. Brisard, Menter, and Smith 2007; M. Donnelly and Evans 2019; Gray, Hooper, and Sandford 2022; Hodgson et al. 2019; Phillips 2003).

2. Grey literature typically includes (but is not limited to) the following: working papers, theses, white papers, policy documents (University of Leeds n.d.). In this study, we also classified sources as grey literature if we could not identify what the literature output type was with confidence.

3. ISCED (see further https://uis.unesco.org/en/topic/international-standard-classification-education-isced) is a well-known framework maintained by the UNSECO Institute for Statistics (UIS) that classifies educational activities and qualifications into internationally agreed groups to facilitate data comparison across national contexts. We adapted the framework slightly to make it relevant for our project and used the following codes. 01: early childhood education (below age 3); 02: pre-primary education (age 3 onwards); 1: primary education; 2: lower secondary education; 3: upper secondary education; 4: post-secondary education; 5: cross-phase education.

Appendix

A

Table A1. The 53 studies included in the analysis.

Author	Date	Title
Allais	2011	The impact of implementation of national qualifications frameworks: A comparison of 16 countries
Andrews & Mycock	2007	Citizenship Education in the UK: Devolution, diversity and divergence
Andrews & Mycock	2008	Dilemmas of devolution: The ‘politics of Britishness’ and citizenship education
Andrews, Xenofontos & Sayers	2021	Estimation in the primary mathematics curricula of the United Kingdom: Ambivalent expectations of an essential competence
Arnott & Menter	2007	The same but different? Post-devolution regulation and control in education in Scotland and England
Atkins, Dalton, Phillips & Stojanovic	2021	Devolved public services: The NHS, schools and social care in the four nations
Bamber, Bullivant, Glover, King & McCann	2016	A comparative review of policy and practice for education for sustainable development/education for global citizenship (ESD/GC) in teacher education across the four nations of the UK
Beauchamp, Clarke, Hulme & Murray	2015	Teacher education in the United Kingdon post devolution: Convergences and divergences
Brisard, Menter, & Smith	2007	Researching trends in initial teacher education policy and practice in an era of globalisation and devolution: A rationale and a methodology for an Anglo‐Scottish ‘home international’ study
Burchardt & Holly	2009	Inequality and the devolved administrations: Scotland, Wales and NI
Chapman & Ainscow	2019	Using research to promote equity within education systems: Possibilities and barriers
Colwill & Gallagher	2007	Developing a curriculum for the twenty-first century: The experiences of England and NI
Croxford	2001	School differences and social segregation: Comparison between England, Wales and Scotland
Croxford & Raffe	2014	Social class, ethnicity and access to higher education in the four countries of the UK: 1996–2010
Croxford & Raffe	2015	The iron law of hierarchy? Institutional differentiation in UK higher education
Czerniawski, Gray, MacPhail, Bain, Conway & Guberman	2018	The professional learning needs and priorities of higher-education-based teacher educators in England, Ireland and Scotland
Donnelly & Brown	2022	‘Policy traction’ on social and emotional wellbeing: Comparing the education systems of England, Wales, Scotland and Northern Ireland
Gallacher & Raffe	2012	Higher education policy in post-devolution UK: More convergence than divergence?
Gunning & Rafe	2011	14–19 education across Great Britian: Convergence or divergence?
Healey, Jenkins & Lea	2014	Developing research-based curricula in college-based higher education
Hizli Alkan & Priestley	2019	Teacher mediation of curriculum making: The role of reflexivity
Hodgen, Pepper, Sturman & Ruddock	2010	Is the UK an outlier? An international comparison of upper secondary mathematics education
Hodgson, Spours & Waring	2011	Post-compulsory education and lifelong learning across the United Kingdom: Policy, organisation and governance
Jeffery & Wincott	2006	Devolution in the United Kingdom: Statehood and Citizenship in Transition
Kippin & Cairney	2021	The COVID-19 exams fiasco across the UK: Four nations and two windows of opportunity
Kisby & Sloam	2012	Citizenship, democracy and education in the UK: Towards a common framework for citizenship lessons in the four home nations
Kuhn, Bradshaw, Donkin, Fletcher, Liht & Wheater	2021	PISA 2018 additional analyses: What does PISA tell us about the wellbeing of 15 year olds?
Leat, Livingston & Priestly	2013	Curriculum deregulation in England and Scotland: Different directions of travel?
Longhurst, Stone, Dulohery, Scully, Campbell & Smith	2020	Strength, weakness, opportunity, threat (SWOT) analysis of the adaptations to anatomical education in the United Kingdom and Republic of Ireland in response to the COVID-19 pandemic
Machin, McNally & Wyness	2013	Educational attainment across the UK nations: Performance, inequality and evidence
Maguire	2021	Early leaving and the NEET agenda across the UK
Major, Eyles & Machin	2021	Learning loss since lockdown: Variation across the home nations
McCluskey, Cole, Daniels, Thompson & Tawell	2019	Exclusion from school in Scotland and across the UK: Contrasts and questions
McKie & Jackson	2006	Reflections on teacher education in the four nations of the United Kingdom
Mitchell	2006	Evolution and devolution: Citizenship, institutions and public policy
Mooney & Williams	2006	Forging new ‘ways of life’? Social policy and nation building in devolved Scotland and Wales
Munoz-Chereau & Ehren	2021	Inspection across the UK: How the four nations intend to contribute to school improvement
Murphy, Neil & Beggs	2007	Primary science teacher confidence revisited: Ten years on
Paterson & Iannelli	2007	Social class and educational attainment: A comparative study of England, Wales and Scotland
Phillips	2003	Education policy, comprehensive schooling and devolution in the disUnited Kingdon: An historical ‘home international’ analysis
Raffe	2005	Learning from ‘home international’ comparisons: 14–19 curriculum and qualifications reform in England, Scotland and Wales
Raffe, Brannen, Fairgrieve & Martin	2001	Participation, inclusiveness, academic drift and parity of esteem: A comparison of post- compulsory education and training in England, Wales, Scotland and NI
Reid	2010	Management of school attendance in the UK: A strategic analysis
Reynolds	2008	New labour, education and Wales: The devolution decade
Ross & Hutchings	2003	Attracting, developing and retaining effective teachers in the United Kingdom of Great Britain and Northern Ireland
Shattock & Horvath	2020	The decentralisation of the governance of UK higher education: The effects of devolution to Scotland, Wales and Northern Ireland, and on England
Sibieta & Cottell	2020	Education policy responses across the UK to the pandemic
Sibieta & Cottell	2021	Education re-opening and catch-up support across the UK
Sibieta & Fullard	2021	The evolution of cognitive skills during childhood across the UK
Simkins	2014	Governance of education in the United Kingdon: Convergence or divergence?
Wiggins & Tymms	2002	Dysfunctional effects of league tables: A comparison between English and Scottish primary schools
Woods, Torrance, Donnelly, Hamilton, Jones & Potter	2021	Constructs and purposes of school leadership in the UK
Wyse & Ferrari	2015	Creativity and education: Comparing the national curricula of the states of the European Union and the United Kingdom