We released v0.1 of the Digital Assessment Maturity Model in March 2023, and spent the rest of the month talking to people from the sector and refining our initial work. This blog post lays out the version 0.2 of the model and outlines how it may be used.
Current assessment landscape
Assessment processes are complex [Citation needed]. Different universities do things differently and use differing terminology, however, the general process is somewhat similar. Our first step in developing a maturity model was to attempt to define the assessment process, shown below:
Note, there will be considerable institutional nuance, but the process chart aims to capture most of the activities that take place to ensure assessments run effectively.
Our second step was then to identify and define areas of activity within the assessment process. We have called these:
Assessment design
Assessment creation and resourcing
Assessment completion and submission
Marking and feedback
Quality assurance and enhancement
These were chosen because they often require the work of different groups of people within the institution. For example, timetabling and estates teams may only focus on assessment creation and resourcing activities in relation to scheduling on-campus digital exams. It also allows us to potentially tag relevant resources in the future.
We are aware that this is an incredibly simplified process map compared to sector practice, and that each of the five activity areas is inextricably linked to each other – even if that is not immediately apparent from the process map.
Growing digital assessment
Digital assessment has increased in UK Higher Education over the last five years, driven by underlying strategy and greatly accelerated due to the COVID-19 pandemic. Universities and Colleges may have found themselves un- or under-prepared for a mass move to online exams, for example, and innovation practice happened at pace.
In these circumstances, enhancements to digital assessment may not have occurred evenly across all aspects of the assessment process. You may have concentrated efforts on creation and resourcing, or completion and submission, as these are the most visible to students, and considered digitising the quality assurance and enhancement element less so.
It is worth noting that while even enhancements across the five activity areas is not entirely necessary, as you progress through the digitisation of assessments, the links between stages can cause bottlenecks in your overall processes where they are at different points in their maturity journey.
Creating the maturity model
We started from the point that accessibility is something that is baked-in to all aspects of the assessment process, and that digital assessment meets all the legal and moral requirements currently in place. While this assumption may be fair, where this is not the case, accessibility should be approached before any other changes as per this maturity model.
We have split the maturity model into four stages, with each stage denoting a significant shift in what you may observe. The stages are:
Approaching and understanding,
Experimenting and exploring,
Operating and embedding,
Transforming and evolving.
The progression from one stage to the next is cumulative; each stage builds from the one before. Within each stage, we have attempted to identify a non-exhaustive list of activities from across the wide spectrum of assessment practice as examples of what you might see at that stage.
The model is not intended to be a pathway you progress along, but rather a self-assessment checkpoint and to give you suggestions for the future of digital assessment in your institution. For many institutions, moving along the Digital Assessment Maturity Model may not be appropriate or favourable. Indeed, achieving ‘transforming and evolving’ may not be compatible with your internal practices or strategies.
Please don’t use this as a checklist to determine your stage; use it as a holistic tool to give a potential direction of travel. There will be some relation between your institutional assessment strategy, however, you should ensure an alignment between your strategy and practices over this model. The most value may come from comparing the maturity stages across all five areas of activity of the assessment process, especially if there are wide discrepancies, and to help update and renew institutional strategies.
Remember that Jisc has a range of experts and experience in supporting the development of digital assessment, purchasing frameworks and services, so please reach out to your Relationship Manager for more information.
Considerations and constraints
We have purposely avoided assessment methods and methodologies, as they are often subjective and pedagogically-driven (don’t worry, we are looking at that elsewhere in Jisc). We have tried to focus on the digital assessment process and the underlying technologies as enablers.
Introducing the model v0.2
We have structured the Digital Assessment Maturity Model so you can scan across each area of activity of the assessment process and get a quick understanding of the differences between each stage of maturity. Each element has a summary title that gives a very basic overview, followed by some examples of what you might see.
In addition, we have created an abridged version.
Key changes
Based on sector feedback, we made a number of notable changes to the model.
The most obvious change from v0.1 is the move away from a graph to a more tabular representation. While it is anticipated that institutions will move through the model from left-to-right, the graph gave the appearance that ‘rightmost was better’. Some institutions will not find that to be the case.
The second biggest change was reduction of the model from five to four stages through the amalgamation of ‘operational’ and ‘embedding’. We also took the opportunity to tweak some of terminology and provide more context for what each stage means. Although this does take us away from existing Jisc maturity model nomenclature, we felt it made the model more relevant and usable.
The third biggest change is the restructuring of information to provide a better coverage of activities and to layer content detail. While we previously had the model and blog posts for two levels of detail, we now have summary headline, what you might see, and the blog posts. This should ensure the model is useful as an in-depth guide and as an ‘at-a-glance’ tool. This may necessitate the rewriting of the blog posts, but that was outside the scope of the current work.
Further changes were made to :
‘Quality management’ was changed to ‘quality assurance and enhancement’ to better match sector terminology
The assessment process was added, and colour-coded against the model for ease of reference
Additional emphasis that the examples in each stage are indicative of what you might see, rather than a checklist of things to be achieved.
Dropping ‘phases’ in favour of ‘areas of activity’ when describing the assessment process.
Working alongside colleagues from Jisc and the sector, we will be further refining the model, particularly the terminology and examples at each stage. We especially hope to engage with colleagues in quality enhancement, timetabling and estates to assist with these refinements.
As the blog posts are largely redundant now that a lot more information has been moved into the model, we intend to rewrite the accompanying blog posts. We may repurpose them to provide additional context or next-steps.
If you are interested in being part of this review, please contact us. If you have any comments or questions, please add them below.
When the COVID-19 pandemic hit in early 2020, it forced UK education institutions to shift teaching online. Many of us assumed that it might only be for a few weeks, but then as spring gave way to summer, it was clear that assessments would have to change markedly and quickly. Three years on, we have reached a point where we can reflect on these rapid changes in practice. Maybe you introduced a number of new software services and are struggling to embed their use, or maybe you feel you failed to capitalise on some of the benefits of digitalisation throughout your digital assessment process.
At Jisc, we have begun the process of classifying digital assessment maturity model to allow institutions the chance to assess their current level of maturity and use it to plan their next steps.
We have split the maturity model into five stages, with each stage denoting a significant shift in practice. The stages are:
Approaching and understanding;
Experimenting and exploring;
Operational;
Embedded;
Transformational.
The progression from one stage to the next is cumulative; each stage builds from the one before. Within each stage, we have attempted to identify activities from across the wide spectrum of assessment practice within an institution. After introducing the generalised model, you will find more detailed sections for each phase of the assessment process below.
We didn’t intend for the model to be a linear pathway you progress along, but rather a self-assessment checkpoint for your institution (previously I did an individual digital literacies self-assessment tool) and to give you suggestions for the future of digital assessment in your institution. For many institutions, moving along the digital assessment maturity may not be appropriate or favourable. Indeed, achieving ‘transformational’ may not be compatible with your internal practices or strategies.
Please don’t use this as a checklist to determine your stage; use it as a holistic tool to give a potential direction of travel. Remember that Jisc has a range of experts and experience in supporting the development of digital assessment, so please reach out to your Relationship Manager for more information.
Considerations and constraints
We have purposely avoided assessment methods and methodologies, as they are often subjective and pedagogically-driven (Jisc are looking at that elsewhere). Therefore, it was important to focus on the digital assessment process and the underlying technologies as enablers.
We identified five phases of the assessment process: design, creation and resourcing, completion and submission, marking and feedback, and quality management. Although institutions may have their own unique way of approaching digital assessment, we anticipate these areas are suitable for all. However, these phases may appear linearly or concurrently, and different teams in your institution may out the related activities. We will be following up this post with a series that unpacks each area in more detail, giving specific examples of some of the activities you might see for each phase.
Introducing the digital assessment maturity model
Digital Assessment Maturity Model. Emma Beatson, 2023
Digital assessments and the digitalisation of the processes surrounding assessment may be rare or non-existent at the approaching and understanding stage. Many activities will be paper-based and manual. Where technology is used to support assessment, it is primarily around administration or data collection. Assessments will be designed to be accessible to all students, which may include the use of digital alternatives, assistive technologies or interventions, such as amanuensis.
Institutions that are actively experimenting and exploring may be using digital assessments in small-scale trials, or looking at alternative assessment modes and methods. This may include automated knowledge-check MCQs (multiple choice questions) and formative assessment. They are beginning to understand the benefits of using technology to assess student learning, and to build the skills and knowledge required to effectively implement digital assessment tools. Institutions may be looking at BYOD (bring your own device) or other policies to ensure students have access to suitable learning technologies.
When digital assessment is operational, it may be used to enhance traditional assessment process and to add optimisations into the assessment process. Essays are submitted via a VLE, exams may have moved online, and the use of anti-plagiarism software may be commonplace. There may be processes and policies designed to ensure digital assessments are consistent, reliable and valid. Discrete systems to support assessment may start being integrated and specified in the curriculum. There is suitable infrastructure and resource to support digital assessment, including Wi-Fi, hardware, and software. Staff and students are trained and supported in the use of new technologies for assessment.
Digital assessments become fully integrated and embedded into the curriculum, when technology begins to provide adaptive and personalised learning and assessment experiences for students, and data analytics for staff use. Assessment technologies are integrated with various institutional systems, such as the VLE, student record system, and curriculum management system, to allow the creation of a holistic picture of student performance. Analytics may be used to trigger interventions for students at risk of failing. There may be purpose-designed spaces on campus for student assessment.
At the transformational stage, digital assessments are used to support and enable different forms of pedagogic practice, such as problem-based learning, competency-based education and personalised learning. Data generated by assessment is used for continuous improvement of process and performance. Institutions have fully integrated digital assessment tools into their teaching and learning. They are actively and critically exploring new technologies and approaches for assessment, and are beginning to use predictive analytics to better understand student learning.
Digital assessment design
When approaching and understanding assessment design, most assessment will be simple, and often paper-based. The design process is often manual. Digital assessment may be sometimes used to supplement traditional assessment methods.
Experimenting with digital assessment may introduce additional modes of assessment, such as MCQs, and may use existing technologies, such as the VLE tools to introduce elements of question randomization and auto grading of basic assessments. The assessment process may involve increased digitalization. The use of digital assessment may be focused on low-/no-stakes formative assessment initially. Technology use is primarily aimed at replicating existing methods of assessing.
As things become operational, elements of personalization may become apparent, especially to support additional needs. There may be some level of automation of the design process through the use of rubric builders or shared resources. Disciplines may receive additional software to support subject-specific requirements, such as mathematical notation. Assessment methods or approaches may be augmented through the use of technology. Specific technologies may be used beyond the VLE, such as in-class MCQ services.
The shift to embedded digital assessment, will see the design process informed by data analytics and may include elements of machine learning to determine the optimum assessment strategy. There will be increased personalization designed into assessments. The use of automated knowledge and application checks as formative assessment will be extensive, giving students real-time feedback on their performance. The assessment methods or approaches may be significantly modified through the use of technology.
In a transformational scenario, assessments are fully dynamic, constantly updated, and personalised to each student. Machine learning algorithms may refine the assessment design to optimise the assessment experience. Assessment design will be integrated into a larger digital learning ecosystem, made of a range of specific services to enable variety and flexibility for staff and students. Assessments will assess a wide-range of skills and applications of knowledge. Assessment design is likely significantly reimagined through the use of technology.
Digital assessment creation and resourcing
At the level of approaching and understanding, assessments are often limited to a small number of assessment types and are manually created. The timetabling and administration of assessments is a manual, paper-based process, relying on the labour of academics and administrators. There may be a lack of consistency in process and practice, potentially leading to errors and scheduling conflicts.
When experimenting and exploring digital assessment creation and resourcing, you may begin to see the use of standardized templates and question banks. There may be some level of automation of the creation process, and the integration of multimedia or interactive elements. Spreadsheets or calendars may be used to record and plan exam and assignment timetables. There is little integration of these systems with others relating to the assessment process.
As digital assessment creation is operationalised, the creation process is more automated and may offer rule-based differentiation or personalisation of assessments, such as the use of adaptive release in VLEs. There will begin to be a level of integration between timetabling, assessment management, VLE, curriculum management, or Student Information systems. Timetabling is fully-automated but may rely on some human ‘tidying’. Assessment load and bunching can be identified and acted upon. Additional resources will be available to students to support them to develop their assessment skills, often within the assessment system itself.
When digital assessment is embedded, integrated timetabling, assessment and student information systems will automatically generate a personalised, flexible and adaptive assessment. Students may have the opportunity to choose the mode of assessment to evidence meeting learning outcomes. The creation process will use data from historical assessments, student performance and demographic data to optimise the creation and resourcing of assessment.
Digital assessment creation and resourcing may be transformational when assessments are fully dynamic. This means the modes, scheduling, and resourcing of assessments may be variable and flexible. Students may be able to complete assessments according to their own schedules, rather than follow the institution’s calendar. Timetabling may be fully automated and actively avoids bunching, and may allow personalisation by booking of assessment windows. Curriculum data from other systems may be used to automatically generate assessment points.
Digital assessment completion and submission
There may be no digital element of assessment at the approaching and understanding the completion and submission process. While students may use word processors or other software to generate their work, submission may be paper- or storage media-based. There is a limited use of technology to manage and track submissions; this is likely paper-based or a standalone system, such as a spreadsheet. In some instances, assessments are printed out for marking, feedback, and archival purposes.
At the experimenting and exploring stage, digital platforms are used for submission, such as email, Google Forms, OneDrive, or basic VLE tools — but this is not consistent. Managing and tracking of submissions is carried out within the system, and may provide students with a digital receipt or record of submission.
When digital assessment is operational, there is use of digital submission of assessments for most modes. Often there will be specialist assessment tools, such as Turnitin or Wiseflow. There may be a level of automated plagiarism checking or grading present, as well as the ability to provide instant feedback (especially in the case of MCQs). Late submissions are automatically flagged, and there may be a notification system for staff and students. There may be some level of automation of reasonable adjustments, but this may rely on significant manual intervention.
The embedded level is similar to operational, but practice is more consistent. The digitalisation of non-traditional assessments, such as videoing presentations or scanning artwork, may be more commonplace, allowing digital marking and feedback. There may be more robust policy and processes to govern submitted media and its archival.
Transformational digital assessment may include the use of analytics and data visualisation to explore long-term trends and personalise student support. Reminders or hints could be released based on student activity, e.g. flagging students who haven’t accessed content or module handbooks. The process for providing additional support or reasonable adjustments may be fully automated, based on student needs, assessment type, and support available. There may be some level of integration between assessment systems and the tool used to complete the assignment.
Digital marking and feedback
The marking and feedback for assessments is manual at the approaching and understanding stage. Staff will grade assignments and provide hand-written feedback to students. The feedback process is often time-consuming and may not be consistent across teachers or assessments.
At the experimenting and exploring stage, there may be introduction of some level of automation in the process. Multiple choice questions may have pre-determined grading and automated per-question or generalised feedback. Feedback begins to be delivered in digital formats. Academic staff have access to suitable equipment to support digital marking, such as two monitors or input devices.
At an operational stage, personalised, digital feedback is provided for each student. There may be some element of aggregation and analysis of marking and feedback to provide both longitudinal development for individual students, and cohort analysis. Academic staff have access to multimedia creation resources in order to provide marking and feedback in the most appropriate manner. For group assessments, it may be possible to mark and feedback based on whole-group and individual contributions within the system. Rubrics will be used, but may not be marked against digitally. To support staff, assessment content may be converted to different media types to best support their working practices or additional needs. There may be provision for second-marking, multiple markers, team marking or batch marking within cohorts.
When digital assessment is embedded, there may be the incorporation of analytics and predictive models to provide adaptive, automated feedback. Feedback may be aggregated for individual students to allow them to note patterns. Rubric use is systematic and wide-spread, with grading and associated feedback digitally captured, and shared with other systems. Group assessments may allow peer assessment or weighting. The use of standard feedback phrases may be partially automated or suggested. Second-marking, multiple markers, team marking, or batch marking may be fully automated based on pre-set criteria.
To be transformational, the marking and feedback process is dynamic and adaptive. Aggregated feedback will prompt skill development for students, and will highlight suitable resources and support opportunities automatically. The use of AI or machine learning may automate some of the marking and feedback process. Feedback may be constantly updated and optimised based on real-time student data. Academic integrity detection may encompass new technologies and methods, such AI authoring. This may include automated checks, or the ability to determine document histories or analysis of snapshots during the production of the assessment submission.
Digital quality management of assessment
When approaching and understanding digital assessment, student assessment data may be stored in ad-hoc, basic tools, such as Microsoft Excel. Sharing data may only take place via email or on-premises network storage. Exam boards may be paper-based, in-person meetings, with manual processes to support them.
You will see the inclusion of digital assessment into the process and management of assessments at the experimenting and exploring stage. Rubrics and standards may be evident, but there may be limited use of technology to support quality control.
At an operational stage, robust quality measures are evident, including peer review or testing. There are likely to be specific software packages through which to record, distribute and analyse academic outcomes. Assessment data will be passed between relevant systems to allow the automation of grade return, feedback collation and management, and exam board processes. Where not explicitly designed out, the use of plagiarism and academic integrity tools may be used.
The embedded stage involves more sophisticated quality control measures, such as digital processes to support external validation, sampling, or accreditation with disciplinary bodies. Assessment data and validated results will be synchronised across multiple systems, and may be used to provide insight into student performance and predictive analytics.
Where digital assessment is transformational, the advanced use of data analytics and machine learning to monitor and improve assessment quality. Assessment data management from a variety of sources will be fully integrated and seamless.
Conclusions
This work is at its very early stages, and we are looking to road test it with the sector. We have plans to work with colleagues from various universities to further refine this model, with the hope that it provides value to senior leaders, timetabling managers, IT directors and Quality Management leaders.
Please contact me if you are interested in being part of this review. If you have any comments or questions, please add them below.
