How do students experience inclusive assessment? A critical review of contemporary literature

ABSTRACT

Assessment drives learning and determines success in higher education. In a robust and defensible system, assessment should not exclude based on extraneous student characteristics, particularly as the student body becomes more diverse. This research sought to examine classroom assessment designs that might make assessment inclusive. A critical literature review was conducted identifying 13 research papers where outcomes of inclusive assessment were reported. Included studies focussed on students with disabilities, international and linguistically diverse students. Only one study examined the effects of inclusive assessment design on student learning. Efforts to make assessment more inclusive were as follows: offering students choice, programmatic approaches to assessment and co-design of assessment and policies that promote inclusion. Universal design for assessment has not been widely implemented within the sector. This is likely due to limited theorisation and operationalisation of inclusive assessment and assessment design processes that favour tradition and taken-for-granted assumptions about how assessment should be. Assessment designers should consider the ways in which assessment might exclude and to foster wider scholarship towards assessment for inclusion.

KEYWORDS:

• Assessment design
• higher education
• social inclusion
• disability inclusion
• universal design

Introduction

Efforts to widen participation in higher education range from outreach programs in schools, to admissions schemes, and support during studies to improve success and retention of students from diverse backgrounds. However, few efforts focus specifically on assessment, which is likely a barrier to achievement and success in higher education (Moriña 2017). Though Universal Design for Learning (UDL) provides clear pathways to improved instructional design within higher education, including within classroom assessment (Ketterlin-Geller, Johnstone, and Thurlow 2015; Burgstahler 2020; CAST 2018) students from diverse backgrounds still report assessment as a significant challenge within higher education (Grimes et al. 2019b). This might be due to assessment strategies that do not account for the increasingly diverse student population. Inclusion is also a subjective matter: what one student perceives as inclusion, might not be for another; there is unlikely to be one solution. In the present resource-constrained environment of higher education, it is imperative that we understand what types of assessment designs can promote the success of diverse students. This is important to guide practice and further research. This paper aims to address this need through a critical review of the research literature, focussing on peer-reviewed, empirical studies of assessment methods that students experience as inclusive in higher education.

Widening participation in higher education

The move towards universal higher education has significantly broadened the targets and thus diversity of higher education, with an increase in students from non-traditional backgrounds (Trow 2017). Inclusion has been a goal of higher education in relation to a number of equity groups. In Australia, equity student groups include the following: Aboriginal and Torres Strait Islander peoples, people from low socio-economic status backgrounds, people with disability, people from remote, rural or isolated areas, people who are the first in their family to attend a university or other higher educational institution, people from non-English speaking backgrounds and women in non-traditional areas (Willems 2010). Similarly in the UK, widening participation seeks to improve HE participation for those from lower socio-economic groups, people with disabilities, ethnic minorities, mature students and part-time students (Lewis and Johnston 2002). Beyond government policy identified groups, inclusion might also encompass international students, students from multicultural backgrounds (Blasco 2015), students with low digital literacy (Buzzetto-Hollywood et al. 2018), first-in-family (O’Shea et al. 2016) and under-represented students in general (DiBartolo et al. 2016). Preparatory programs and admission are meaningful initial steps on the path to widening participation (Dodd, Ellis, and Singh 2020). While anti-discrimination laws have required reasonable adjustments and changes to be specifically made to assessment to improve equity for certain groups of students, i.e. those with disabilities, a systematic approach to widening participation through inclusive assessment practices has not yet been realised.

Assessments practices in higher education

The design and implementation of inclusive assessment are challenging, not least because the multiple purposes of assessment are already in tension. That assessment needs to serve certification or accreditation functions, foster students’ learning, and also ideally prepare students for a future beyond university where can no longer rely on the advice and judgement of teachers (Boud and Soler 2016) – already places substantial demands on what assessment should do.

The concept of inclusive assessment has existed for some time, defined as ‘the design and use of fair and effective assessment methods and practices that enable all students to demonstrate to their full potential what they know, understand and can do’ (Hockings 2010, 34). Several frameworks incorporate the term ‘Universal Design’ for application in higher education, which have been influenced by the ‘Principles for Universal Design’ developed in design and architectural practice (Burgstahler 2015). Universal Design seeks to design built environments to be as accessible as possible from the outset to as many people as possible (Hamraie 2013). While higher education inclusion frameworks ultimately advocate for proactive teaching strategies which ensure that education is accessible to diverse students, few principles relate to assessment specifically (Burgstahler 2015).

Universal Design approaches take a cognitive and materialist approach to inclusion, which are likely to draw focus to making changes to the configuration rather than the substance of the assessment task. UDL principles are to provide multiple means of engagement, representation, and action and expression. When applied to assessment, this means assessment tasks should be communicated through multiple media, and there should be options for students to demonstrate their capabilities (CAST 2018). Though this might particularly help students with learning or sensory disabilities, it is unlikely to fully account for broader student diversity, and continue to inadvertently disadvantage or exclude. McArthur (2016) interrogates this through her conceptualisation of assessment for social justice, where she suggests that a focus on the procedural justice within assessment has distracted from broader opportunities to achieve justice of outcomes in higher education, both for students and society.

Practical guidelines for the implementation of inclusive assessment exist within teaching and learning guidance/resources rather than as peer-reviewed publications, commonly at the university level. The value of this body of work has not yet been fully realised, since there is a relative dearth of studies on the design and implementation of inclusive assessment. A literature review of peer-reviewed empirical articles on Universal Design (Roberts, Satlykgylyjova, and Park 2015) encompassing 2000–2014 only identified one journal article which specifically focussed on assessment (Ragpot 2011).

Accommodations or adjustments are another common approach to inclusion in assessment, tailored to individual students. Universities generally have processes where students need to declare a disability and submit medical reports on their condition to develop an individualised access plan and therefore accommodations for specific assessments. Common accommodations include the following: altering the setting, timing, scheduling, presentation or allowed responses (Lombardi, McGuire, and Tarconish 2018). Research examining accommodations shows mixed results, limited by a small number of studies and complexity that is difficult to reduce (Jain 2020). An individual accommodation approach risks perpetuating a deficit discourse that focuses on the individual as the source of access concerns rather than the inaccessible system – and the burden of proof rests with the student (Gabel and Miskovic 2014). This results in a hidden population of students, who may not wish to disclose disability for risk of stigma, feeling different or perceptions of unfairness (Kendall 2016; Grimes et al. 2019a). Staff with a deficit view of disability may be reluctant to offer adjustments due to perceptions of unfairness or they do not know how (except for giving extra time on tests or providing a quieter location) (Becker and Palladino 2016) and where they do, this is often on the periphery.

Equity group students’ assessment experiences and outcomes

Though equity group students’ experiences of university have been well explored, few have focussed specifically on experiences of assessment. Madriaga et al. (2010) surveyed several hundred disabled and non-disabled students in the UK, finding students had similar assessment-related challenges overall, but that disabled students reported more difficulties in completing work in the allocated time. Hanafin et al. (2007) interviewed 16 students with physical disabilities and dyslexia and reported that while some barriers within assessment were addressed, there remained challenges to access including physical accessibility, the structure of teaching and staff attitudes to disability. International students’ experiences of assessment have also been challenging, due to differences in the ways that assessment is done in a particular country or institution which are usually not made explicit. This has been the case for students in different European countries (Blasco 2015), as well as Chinese international students studying in Western countries (Dai, Matthews, and Reyes 2019). However, all of these studies have explored the status quo, rather than specific efforts towards inclusive assessment.

Several studies have explored how changes to assessment can improve inclusivity. Flexibility and choice have been frequently argued to be an inclusive assessment option, where students do not have to take up a task type that might disadvantage them (Lawrie et al. 2017). Keating et al. (2012) implemented assessment choice, where students could choose an existing option or propose an alternative format of their own. A workshop and learning activities were offered to establish students’ understanding of the different task formats. In focus groups, students did prefer to have more of a say in how they were assessed. However, in this work, it was unclear if students involved in the project were indeed differently abled, so while promising, this work did not discern if there was a variable impact on equity group students. In contrast, Ragpot’s (2011) study did not provide a choice but instead modified task types according to UDL. Weekly tasks relating to class discussion, a group poster, a dramatic play and an end of term essay were all used. In focus groups, students reported these tasks fostered deeper learning and engagement with the material. However, it was unclear if a particular equity group was targeted within this work, and so while the implementation was successful, again it is unknown if students from equity groups fared better than in previous iterations of assessment.

Studies that have reported on student outcomes in terms of marks or grades in relation to equity groups – and particularly students with disabilities (SWD) – appear to be few. Most recently, O’Neill (2017) reported no significant difference in student grades when provided with a choice of assessment across nine modules, and overall grades exceeded those of previous cohorts. This aligns with research by Craddock and Mathias (2009), who provided an assignment as an alternative to a closed book unseen examination to address the concern that previous cohorts of students with dyslexia were disproportionately failing the examination. They also established there was no significant difference between groups in performance and that having a non-exam option allayed stress amongst all students. Stegers-Jager et al. (2016) examined differences in the performance of students in a Dutch medical education context based on ethnicity and socio-demographic characteristics. With no specific inclusive assessment approach, they found that while there were no differences based on first-generation university students and non-first-generation students, local Dutch students fared better on written, practical, and language exams. Taken together, this work suggests that inclusive assessment is likely to benefit equity group students, but is far from conclusive.

Whilst UDL, principles for inclusion, and assessment accommodations have been discussed for many years, there is a need for a more thorough and critical review of contemporary research on inclusive assessment outcomes in higher education. Such a review will help to elucidate firstly what kind of inclusive assessment approaches have been tested (rather than hypothetical), and secondly, which equity groups (including and beyond SWD) might benefit from such approaches. Our research questions are therefore:

• Who are the students targeted in inclusive assessment?

• What are the effects/outcomes for students of inclusive assessment?

• What recommendations does the literature make about designing inclusive assessment?

Methods

Review approach

In this review, we take a critical perspective. A critical review takes a purposeful and critical approach to literature synthesis (Sutton et al. 2019) and allows us to ‘evaluate what is of value from the previous body of work’ (Grant and Booth 2009, 93). We do not make assumptions about the value of particular interventions and seek to understand the differential effects of modifications of assessment processes and assessment task design for different groups of students (Maxwell 2012). To inform further research, it is important to know what has been implemented, and what works for whom, since context and background (of a student, the discipline and the setting) will play a large part in what is considered appropriately inclusive (Pawson 2013).

Literature scoping

To identify an appropriate timeframe and search terms, we conducted an initial scoping search for relevant papers via three core education databases: ERIC, Scopus, Education+. The aim was to identify how widely the topic of inclusion is covered within the academic literature and which terms were most common. The time frame was set from 2005 to 2020. 2005 was chosen because in August 2005, The Disability Standards for Education, Australia, came into effect, predating The Equality Act from the UK (2010). Initial scoping returned many irrelevant publications, with many not focussing on higher education, let alone equity groups.

Furthermore, the search in a critical review is not meant to be exhaustive; instead, it seeks to identify the most significant items in the field (Grant and Booth 2009). We therefore searched only using the ERIC (via EBSCOhost) database to target education-related papers. The time limit for publications was specified from 2015 to 2020, to capture more recent attention to the practice of inclusive assessment in higher education. This dovetailed with a previous systematic review of empirical research on UDL in higher education, which contained work up until the prior year (Roberts, Satlykgylyjova, and Park 2015). We revised our search terms to capture the relevant literature. The final search string was as follows: AB assessment AND AB (‘higher education’ or ‘university’) AND AB (‘minority’ OR ‘equit*’ OR ‘inclusi*’ or ‘social’ or ‘disab*’), where ‘AB’ refers to searching in the abstract text.

Searches were limited to peer-reviewed articles, book chapters and conference proceedings. The search was conducted on 17 September 2020.

Title and abstract screening

572 publications were subject to title and abstract screening in Covidence (www.covidence.org), web-based software for managing literature reviews. We chose to focus on the peer-reviewed research literature because we intended to understand the landscape of educational evidence of inclusive assessment strategies. In Table 1, the main inclusion and exclusion criteria that were used at this stage are listed.

Table 1. Inclusion and exclusion criteria.

Inclusion criteria

Papers published after 2015 Publications about higher education Focus on assessment relating to learning – whether formative or summative (with at least a section on assessment) Undergraduate or postgraduate higher education Relates inclusion/equity to assessment practices, focus on equity groups Empirical studies Foregrounds students’ experiences or outcomes Peer-reviewed journal articles, book chapters

Exclusion criteria

Publications without a research component Publications without any relation to student assessment Focus on selection into the course/admission process Focus on other forms of assessment not related to student learning (e.g. focus on assessment of a course; assessment of pathways to institutions of higher education, prior learning assessment, assessment relating to bridging programs, assessment of medical conditions) Focus on diagnosing/assessing the disability/impairment Inclusion/equity is considered separate to assessment design or linked to course completion Focus on assessment concepts not students Focus on curriculum refinement/change Reports, books, theses Publications about workplace learning Focus on vocational education (VE), further education (FE), cooperative learning, community-based learning, language learning (not as a part of a course at institutions of HE), assessment in MOOCs Focus on assessment policies Focus on bridging programs (pathways to universities)

Full-text review

A total of 24 publications were included for full-text review. This closer examination revealed that 12 papers did not meet the inclusion criteria (not about assessment, not research papers, not about student experiences or not about equity) and were excluded. Snowballing did not identify any additional papers within the relevant timeframe. As a secondary measure, we also reviewed all papers that were deemed at the title and abstract stage to be conceptually about inclusive assessment and identified one additional paper which contained empirical data. Therefore, 13 papers in total were included for data extraction.

Data extraction and analysis

The research team initially read three papers and met to discuss and develop a data extraction template. Key context data were extracted including country of research, research approach (qualitative, quantitative and mixed methods), sample size, inclusion focus, data collection and analysis approaches, assessment types/design, outcomes, recommendations and a holistic commentary on the article quality. For each article, data were extracted by at least two researchers, and all extractions were reviewed side-by-side in Covidence to resolve any discrepancies in extraction. All papers were discussed among the researchers through regular meetings to develop a combined understanding of papers’ scope, which groups of students they focussed on, the types of inclusive assessment implemented and how student outcomes were considered, in response to the research questions.

Results

Included study characteristics

The included studies (asterisked in the reference list) were conducted in a variety of countries and included SWD, international students and linguistically diverse students. Most of the papers included in the review used a qualitative research design. Full study details are available in Appendix, and Table 2 contains a summary. Many studies contained mixed subject disciplines, but those mentioned were Arts, Computer Science, Education, English as a Second Language, Health, Humanities, Mathematics, Nursing, Science and Social science. One study (O’Neill 2017) did not report on the specific characteristics of included students.

Table 2. Included study details.

Study characteristics		Number of studies
Country of the study	Belgium Finland Greece Ireland Japan Malaysia South Africa United Kingdom USA Zimbabwe	1 1 1 1 1 1 1 3 2 1
Study design	Qualitative Quantitative Mixed methods	8 1 4
Types of students	International students SWDs Linguistically diverse students No grouping	4 6 2 1

In the studies that involved SWD, participants with various learning disabilities/cognitive disorders (e.g. dyslexia, autism spectrum and Asperger’s syndrome, ADHD, dyscalculia, developmental co-ordination disorder) made up the largest group. Fewer participating students reported medical conditions or mental health conditions. Sensory disabilities and physical disabilities comprised the smallest group of participants. Many students declared having multiple disabilities. Another group of studies (Dickinson 2018; Hurst and Mona 2017; Kaur, Noman, and Nordin 2017; Matheson and Sutcliffe 2017) explored the experiences of students from different cultural and language groups. All studies involved students, ranging from a case study of 3 (Nieminen and Pesonen 2020) to a survey of 229 (O’Neill 2017). Three studies also involved a small number of staff (Hurst and Mona 2017; Morris, Milton, and Goldstone 2019; O’Neill 2017).

Inclusive assessment designs

Six studies reported on students’ perspectives on or previous experiences of assessment, and seven papers focussed on the outcomes of implemented inclusive assessments. Of these papers, most developed a range of assessments or provided a choice of assessments, in contrast to their previous practices where there was no choice and frequently a final written exam. While many talked explicitly about implementing Universal Design Principles for assessment (Dickinson 2018; Nieminen and Tuohilampi 2020; O’Neill 2017), others came to the same position on choice and variety of assessment without referencing UDL (Hurst and Mona 2017; Kaur, Noman, and Nordin 2017; Sedghi and Rushworth 2017).

Choice was implemented in a number of ways. Dickinson (2018) drew on UDL to design an English writing course for EAL learners in Japan, incorporating choices of topics, tools and means of expression for writing tasks, and including peer assessment and feedback to foster collaboration and community. Hurst and Mona (2017) instead implemented translanguaging pedagogies in a first-year writing course, where students are supported to ‘cross between all known languages in multilingual classroom contexts’ (132). Students were provided with an opportunity to choose the language to produce assessment artefacts, and this was modelled by the lecturers incorporating (and explaining) terminology from across the languages they spoke, though the main body of work was still in English. O’Neill (2017) investigated choice over nine different modules in three different configurations: a choice between two different individual assessments, a choice between two different group assessments and a choice between the group and an individual assessment. Staff were asked to share a completed template with students, which explained how they had considered equity across the assessment designs.

Two studies investigated how choice could be implemented in groupwork assessments. One focussed on international students allocated diversely within the groups to encourage local and international students to work together, with a series of activities related to the group work undertaken in the class (Sedghi and Rushworth 2017). In contrast, Kaur et al. (2017) commenced with a groupwork co-design process, identifying parameters for group composition, choices of topics and formats and developing a rubric for the task. Students were able to choose a homogenous language group with similar language proficiency and choose the language of task presentation (English or Malay), since all students and assessors could understand both languages.

Matheson and Sutcliffe (2017) drew on the concept of flexible pedagogy to develop a range of assessment tasks that supported international students to tackle complex problems in creative ways, which aligned with the objectives of the course. This included group and individual tasks, a ‘Patchwork Portfolio’ including podcasts and group work, and reflective tasks.

Most radically, Nieminen and Tuohilampi (2020) drew on UDL for a mathematics course to replace a final exam with a summative self-assessment. This was supported through extensive formative assessment and feedback, including self-assessment and peer assessment. Furthermore, rubrics and exemplars were provided at the outset of the course.

Student perceptions and outcomes of assessment

Most of the outcomes were based on student perceptions rather than learning or achievement outcomes. Across all 13 studies included – and particularly in the studies that did not attempt to implement an inclusive assessment – there was no situation where assessment was reported to be entirely inclusive. Instead, there were varying perceptions of how satisfactory assessment was.

Student perceptions

Choice was valued across several studies (Black, Weinberg, and Brodwin 2015). Students with additional learning need valued choice more (38%) than those who did not (21%) (Morris, Milton, and Goldstone 2019) but rates of preferences varied; Dickinson (2018) reported 80% preferred a choice of topics and 88% preferred a choice of formats. Hurst and Mona (2017) offered a choice in language (in South Africa there are 11 official languages) which students reported made them feel their language diversity was valued, however noted that this was restricted to that particular subject and thus had limited broader impact. Griful-Freixenet et al. (2017) also reported that one student preferred fewer rather than greater numbers of assessment tasks.

Many students noted they had difficulty in obtaining accommodations (Black, Weinberg, and Brodwin 2015; Majoko 2018; Stampoltzis et al. 2015) or came up against inflexibility of assessments (Majoko 2018; Stampoltzis et al. 2015), although Nieminen and Tuohilampi (2020) pointed out that removing the exam meant that no accommodations were required. Some students found that in a UDL-revised environment that accessing accommodations was also more difficult (Griful-Freixenet et al. 2017).

In the group task where international and local students were asked to work together, international students preferred this process more than local students (Sedghi and Rushworth 2017), but everyone agreed that assigning groups was more inclusive than leaving group member selection up to individual students. International students liked that it helped them to improve their cognitive skills, whilst local students focussed on their emotional and interpersonal skill improvement.

Socio-emotional effects

The need to disclose a disability to access more inclusive versions of assessment through accommodations made students feel uncomfortable (Black, Weinberg, and Brodwin 2015; Stampoltzis et al. 2015). In contrast, for linguistically diverse students, there was improved self-esteem, self-efficacy and motivation, and reduced anxiety when there was an opportunity for choice (Kaur, Noman, and Nordin 2017) or a variety of assessment tasks across the term (Shi 2018). Interacting through group tasks helped to broaden students’ perspectives and increased their sense of belonging (Matheson and Sutcliffe 2017). Hurst and Mona (2017) also remarked that the translanguaging approach appeared to be most beneficial for those who were ‘usually the most disadvantaged, students from rural or township backgrounds who find the (Eurocentric) institutional culture and language most alienating’ (144). These students reported being more engaged in learning and felt that their diverse backgrounds were valued rather than seen as a deficit.

Student learning and achievement

Only one study looked at how an inclusive assessment strategy impacted on learning outcomes. While O’Neill (2017) did not focus on a specific equity student group, she examined students’ and staff’s perceptions of fairness of assessment choice in general, where choices were designed to afford students different forms of expression while still assessing the same learning objectives. The study found that there was no impact of assessment choice on marks, and students performed better compared to the previous cohorts in the same unit and better than the university average.

Recommendations for inclusive assessment design

Across the included studies, recommendations covered specifics of assessment design and broader considerations. At the core was ensuring the task was a valid way of assessing particular learning outcomes (O’Neill 2017) and that assessment needed to be scaffolded through relevant learning activities (Sedghi and Rushworth 2017). Several studies recommended a programmatic approach to assessment design, to allow a variety of assessment types across a course to be used and to promote learning (Black, Weinberg, and Brodwin 2015; O’Neill 2017; Shi 2018; Stampoltzis et al. 2015). Using UDL to design assessment was important, regardless of student population makeup (Black, Weinberg, and Brodwin 2015; Dickinson 2018; Griful-Freixenet et al. 2017; Stampoltzis et al. 2015).

In terms of process, collaboration (between students, staff and disability support staff) and co-design were suggested to contribute to better student learning outcomes (Black, Weinberg, and Brodwin 2015; Nieminen and Tuohilampi 2020). Introducing and discussing students’ choices in assessment were considered one way to undertake this collaboration (Dickinson 2018; Kaur, Noman, and Nordin 2017) and were hypothesised to also facilitate belonging and increase student motivation (Kaur, Noman, and Nordin 2017; Shi 2018). The need to monitor adjustments after implementation was also important, to see if goals were met (Griful-Freixenet et al. 2017).

Beyond the immediate assessment design, articles also stressed the importance of staff awareness and education about diverse students, and how to support their needs (Black, Weinberg, and Brodwin 2015; Hurst and Mona 2017; Majoko 2018). Black et al. (2015) also argued that students needed support and a supportive environment to encourage help-seeking, since stigma and a fear of negative reactions from staff was a significant student concern. This was also recommended to be tackled at a policy level, reducing requirements for constant self-disclosure of conditions (Majoko 2018). Hurst and Mona (2017) also suggested that policy needed to better support new types of assessment practices, though Stampoltzis et al. (2015) noted that the disconnect between policy and practice should be addressed.

Discussion and conclusion

This review has focussed on the specifics of inclusive assessment design, in contrast to previous reviews which have considered assessment as a part of the broader inclusive education project (Lawrie et al. 2017; Roberts, Satlykgylyjova, and Park 2015). The main approach to inclusive assessment has been providing choices in assessment. Assessment choices could account for a range of diverse students (and indeed students were mostly satisfied with the provision of choice). However, offering choice means it is possible to avoid re-designing assessments that are still inherently not inclusive. This tinkering at the edges might again distract from considering more radical changes to assessment design and systems of assessment.

Notably, all inclusive assessment designs implemented were introduced and interwoven into teaching and learning activities, through discussion activities, the use of exemplars, role modelling and rubrics to orient students to the goals of the assessment. This aligns with Biggs’ (1996) concept of constructive alignment, in which learning activities should prepare learners to complete the assessment. This suggests inclusive assessment, like all assessment design, and requires a more holistic approach that considers the broader contexts as well as the specific mechanics of a task (Bearman et al. 2016).

Poor constructive alignment might undermine students’ perceptions of assessment validity (Ajjawi et al. 2020). This highlights a broader understanding of consequential validity where assessment should be judged not only on its capacity to produce good data but also on its individual and social consequences (intended and unintended) (Iliescu and Greiff 2021). Therefore, we suggest that students’ experiences of inclusion/exclusion in assessment might similarly undermine consequential validity. Furthermore, the consequential validity of each assessment design must be considered where choice is implemented (Sambell, McDowell, and Brown 1997).

The evidence base for contemporary inclusive assessment is quite small, with only 13 empirical peer-reviewed studies. A previous review on UDL by Roberts et al. (2015) also identified only one study that dealt with assessment outcomes up to 2015. Across all studies in this review, student and staff perceptions of inclusive assessment strategies were overwhelmingly positive, suggesting inclusive assessment is likely to make a difference to equity group students. Similar to Stentiford and Koutsouris’ (2020) scoping review on inclusive pedagogies, papers were relatively evenly split between disability inclusion, and social inclusion, which in this case was focussed on international and cultural and linguistically diverse students. Many other equity group students are unlikely to have been directly targeted within studies. Being essentialist about diversity is unlikely to be helpful: an intersectional approach (Cho, Crenshaw, and McCall 2013) to exploring the impacts of assessment on equity group students will be necessary to address the complexities of individual circumstances.

It is possible to surmise that there is no evidence of systematic implementation of inclusive assessment. This is unsurprising given literature highlighting that assessment practice is resistant to change. Deneen and Boud (2014) note that ‘resistance to assessment change is particularly resilient’ (577). Even when teachers’ assessment thinking became more sophisticated, their assessment practices remained the same (Offerdahl and Tomanek 2011). Bearman and colleagues (2016) note that the influence of disciplinary traditions (such as an established custom of essays or exams) on classroom assessment design is pervasive. Assessment also operates in resource-constrained environments, which may hamper innovation and learning-intensive approaches (Dawson et al. 2013). Navigating disciplinary traditions and environmental demands must be accounted for in assessment change alongside more theoretical concerns for inclusive assessment. There are substantial threats to validity where assessments are not inclusive, since they may inadvertently exclude students for the wrong reasons; equally validity is also at risk where the constructs, skills or capabilities of interest are insufficiently characterised or cannot be demonstrated through a particular assessment (Sambell, McDowell, and Brown 1997). The under-theorisation of assessment for inclusion at the level of classroom practice may also impact further operationalisation of inclusive assessment principles and successful practice.

Our review has some strengths and limitations. By restricting the search to 2015–2020, we focus on contemporary practice. Work referenced in the introduction suggests there has not been a significant shift in practice beyond the provision of choices, so a further review of the literature focussing on assessment with an extended timeframe may yield additional articles to strengthen the evidence presented, since previous more general reviews of inclusive education encompassed literature only as far back as 2005 (Roberts, Satlykgylyjova, and Park 2015) and 2010 (Lawrie et al. 2017). The majority of studies were also small-scale research projects taking place in a particular context, usually academics writing about their own practice, and focussing on students’ perceptions of assessments. This makes it hard for results to be generalised both due to context, the Ikea effect (Norton, Mochon, and Ariely 2012), where there is a perception of increased value where one has put effort into construction, and the Matthew effect (Perc 2014) where participants are more positively inclined towards the subject of the research since it has been brought to their attention. Furthermore, reporting bias exists within the field of education (Dawson and Dawson 2018), so reviewing published literature may be insufficient to capture instances where inclusive assessment implementation was unsuccessful.

What is needed in future studies therefore is research which seeks to examine the immediate and longer-term effects of inclusive assessment, which contributes to theory as well as practice. This could include the experiences of diverse students in assessment, an explicit focus on those still excluded, and indices of equality over time, for instance, the intersectionality of enrolled and graduating students, the nature and number of access plans or accommodations and how demands on academic and disability support change. Taking a broader and more holistic approach to consider inclusive assessment within disciplinary and institutional contexts may also afford helpful insights. A systematic conversation about transformation of the assessment towards inclusivity is encouraged.

Disclosure statement

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

Additional information

Notes on contributors

Joanna Tai

Joanna Tai is a Senior Research Fellow at the Centre for Research in Assessment and Digital Learning (CRADLE) at Deakin University. Her interests include student perspectives on learning and assessment from university to the workplace, peer-assisted learning, feedback and assessment literacy, developing capacity for evaluative judgement and research synthesis.

Rola Ajjawi

Rola Ajjawi is an Associate Professor Educational Research at CRADLE, Deakin University. Her research seeks to contribute new knowledge to promote student success at university and includes feedback, failure, emotions and belonging.

Anastasiya Umarova

Anastasiya Umarova is a PhD candidate at the Centre for Research in Assessment and Digital Learning (CRADLE) at Deakin University. She has a Bachelor of Teaching from Russia, Graduate Diploma in Commerce from Lincoln University, New Zealand, and Master of Education in Digital Learning from Monash University, Australia. Anastasiya has more than 5 years of experience in teaching German and English as foreign languages and one year as an instructional designer. Her research interests are in higher education, feedback and feedback literacy, digital learning and student identity.

Appendix. Details of included studies

Table

Study ID	Country	Study approach	Intervention	Total number of participants	Inclusion focus	Assessment context	Assessment type/design
Black, Weinberg, and Brodwin (2015)	United States	Qualitative study	No	15 total: 12 students with disabilities, 3 without disabilities.	Disability – learning disability (5); sensory (2), psychiatric (5) and mobility (4) impairment. Six students had more than one disability	Education major undergraduate students	Assessment in accordance with the principles of UDL/UDI
Dickinson (2018)	Japan	Mixed methods	Yes	40 students	English as Additional Language	First-year nursing students taking a 15-week English writing course	UDL applied to assessments – e.g. group assignment
Griful-Freixenet et al. (2017)	Belgium	Quantitative study	No	10 students	Mixed (students with and without disabilities)	Broad	Mixed (aligned with the principles of UDL)
Hurst and Mona (2017)	South Africa	Qualitative study	Yes	69 students	Ethnicity & language – acknowledging English as an additional language/Decolonisation of higher education	Undergraduate students in a first-year level Humanities course	Translanguaging assignment submission – mainly language biography essay, but there are two additional assignments not really mentioned in the article which can also be multilingual though fewer students took up that option.
Kaur, Noman, and Nordin (2017)	Malaysia	Qualitative study	Yes	114; 12 students completed an open-ended questionnaire	Linguistically diverse learners/international students	Master of Education students	Critical reflections, case studies, reports, presentations, group tasks, paper tests. Group work assessment: homogeneous groups (language, expertise, gender, number of members); topics and modes of presentation chosen by the groups; rubrics created by students in cooperation with the instructors; peer-assessment
Majoko (2018)	Zimbabwe	Qualitative study	No	17 students	Students with disabilities (SWDs)	Final year, final semester students (undergraduate or postgraduate) in education	Alternative assessment, special examination arrangements (extra time and separate exam venues), the extension of due dates
Matheson and Sutcliffe (2017)	UK	Qualitative study	Yes	18 in focus groups, 4 in-depth interviews, survey data and reflective essays from all 52 students in the cohort	International students	Master in International Business Management	Flexible pedagogy including wide range of assessment methods. Scrapbook task which was future-facing; individual evaluation of group created content; Patchwork portfolio including reflective essay, group based blog, podcast; alignment between assessment methods and goals of the subject
Morris, Milton, and Goldstone (2019)	UK	Mixed methods	No	111 undergraduate students completed the survey 30 students (25 volunteered, 5 were chosen) were interviewed	Students with disabilities	Broad	Choice of assessment
Nieminen and Tuohilampi (2020)	Finland	Qualitative study	Yes	3 students	Students with disabilities	Undergraduate mathematics	UDL approach, replaced the course exam with self-assessment
O’Neill (2017)	Ireland	Mixed methods	Yes	229 students	Fairness across all students – no particular group targeted.	Undergraduate courses in science, arts, humanities, social science, computer science and health science	Assessment choice with two options for each task. Individual and group options, choice of methods, topics, criteria, exam questions, submission times
Sedghi and Rushworth (2017)	UK	Quantitative study	Yes	168 students	International students	Science students – chemistry & environment	Multicultural group work – variable weighting depending on unit
Shi (2018)	United States	Mixed methods	No	198 students completed the questionnaire; 8 students were interviewed then brought together for a focus group	English as additional language, international students	Students undertaking an English as Second Language course in parallel to their other studies	Range of alternative assessments (performance tasks, e.g. oral presentations, multiple assignments instead of one final exam/test; group work; peer assessment)
Stampoltzis et al. (2015)	Greece	Qualitative study	No	10 students	SWDs (dyslexia)	Undergraduate and graduate students	Dependent on the professor's decision, an entitlement to oral rather than written exams