JISC

MUSIC (Measuring and Understanding the Systems Integration Challenge in Higherand Further Education) was a study undertaken by the Centre for Social and BusinessInformatics at NewcastleUniversity, following a successful tender to the JointInformation Systems Committee of the Higher Education Funding Councils.

Executive Summary

Introduction

The JISCOrganisational Support Committee commissioned the project in order to feed intomanagement guidance and development activities on the issue of effective integrationof ICT-based information systems. The remit of MUSIC was to provide grounded empirical evidence for the extent andnature of systems integration in the sector; to uncover organisational and managerial factors that encourage or inhibit integration; and to explore benefits and risks. This Executive Summary provides an overview of the principal aspects of the challenges of systems integration and how institutions for higher and further education are responding to them. It offers recommendations to support the JISC in developing advice and guidance for institutions.

Background: The Systems Integration Challenge

The challenge of systems integration has emerged in higher and further education as institutions have become more reliant on computer-based information systems to support both administrative functions (Human Resources, Finance, Student Records) and functions in teaching, learning and research (Virtual Learning Environments, electronic resource discovery tools, etc.). Over the last 10-15 years, most institutions have moved from using information systems which were predominantly built and maintained in-house to the purchase of products from external vendors. Anecdotal evidence has suggested that making these systems work together, or inter-operate, is an increasingly important agenda for institutions. Demands for information systems integration come from a number of sources: the need of institutional management for coherent management information; increasing expectations that systems can ‘seamlessly’ support ‘the student experience’; efforts to eradicate duplication or expensive re-keying of data. Previous work for the JISC suggested that institutions across the sector have made progress but much needs to be done to ensure that any potential benefits will be realised in practice. Moreover, benefits claimed for integration are not always clear and unambiguous.

Overall, very little systematic assessment exists of how prevalent certain approaches to integration are, or how successful. As a result institutions do not have information to benchmark or position themselves within the sector, and they lack materials to help them to weigh up the costs and benefits of particular courses of action (or inaction). The MUSIC project was designed to plug these gaps in knowledge for the JISC and the wider HE/FE Community.

Methodology

This study was based on original research evidence using a variety of social science techniques. These were: a web-based survey of Universities and Colleges; telephone interviews in 29 institutions; site visits to five institutions.

Measuring Systems Integration

The web-based survey element of the study facilitated a mapping of the pattern and extent of systems integration and inter-operability in UK higher and further education. As well as providing a description of the variety of systems used in the sector for key functions, we give an account of those functions that are most typically integrated and the means by which integration is achieved. The data indicate that the key back office functions – Finance, Human Resources (HR) and Estates – are supported by a wide range of products from major enterprise suppliers as well as more specialist suppliers to the sector. Institutions looking to replace systems to support these functions are significantly more likely to replace their HR system.

The most common linkages between systems were between:

• Finance and Student Management (100% reporting some form of integration)
• Library and Student Management (100%)
• Finance and Human Resources (95%)
• Student Management and Time Table (91%)
• Student Management and VLE/MLE (90%)
• Finance and Estates (83%)
• Human Resources and Research Support (75%)

The weakest linkages (all <10% reporting some form of integration) were between:

• Finance and Timetable or E-portfolio
• Estates and Library, VLE/MLE and e-portfolio
• Timetable and Library and Research Support
• Library and Research support

The linkages between systems were much stronger in universities compared with colleges. The general pattern for all institutions who responded was comprised of two broad groups of systems.

• The ‘administrative’ systems, including systems for Finance, HR, Estates and Student Management, were likely to be integrated and to be more ‘tightly’ integrated with each other
• The second grouping of ‘teaching and learning’ systems including VLE/MLE, Library, timetabling, e-portfolio and, again, Student Management systems were also fairly highly integrated, although less so than the administrative systems, in a hub and spoke configuration centred on the Student Management systems.

The common element to these two groups was the Student Management system which appears to form the centre or core of the systems integration landscape in higher and further education. The technical means by which systems interoperate are varied. By far the most typical method is the use of periodic ‘data dumps’ from one system to another. We also explored the attitudes of senior IT managers towards systems integration. We found overwhelming consensus among this community that integration is a priority for institutions and that it is an issue of increasing importance. There is also substantial agreement that the current situation is not satisfactory, and that end-users are pressing for change.

Understanding System Integration in Practice

The key drivers of systems integration were generally reported as the following issues:

• More accurate, timely and consistent management information
• Minimising inefficient duplication of effort
• Enhancing the end user (student and staff) experience
• Generating new information to support business development activity

Only a minority of institutions had formal systems integration strategies. Many institutions indicated that systems integration was subsumed within wider informationor IT strategies and that it featured as a (significant) element of specific project plans. Integration was almost always described as incomplete and in some universities ambitious plans for further integration are in place. A common concern was that new implementations and developments did not ‘undo’ achieved integration.

The main reported examples of integration projects focused on:

• various portals (for student, staff, alumni or applicants)
• single login schemes
• on-line student registration and other self-service administrative tasks.

Overall FE colleges reported much less integration effort than universities. There was, however, evidence that colleges were beginning to look to integration as an important future agenda. From the various forms of data, we have identified the following, largely implicit, approaches or patterns:

• Integrated in-house. These institutions have a DIY approach to systems and therefore to integration. This is seen to be beneficial in supporting rapid change and adaptation. However, there is a strong sense that this option is becoming unsustainable as the costs and complexity of systems increases.
• External systems from a limited number of vendors. This approach is akin to a classic ERP approach (but with the addition of some specialist educational systems to support teaching and learning). By minimising the number of vendors, integration issues are, it is argued, minimised or passed to the favoured vendor.
• Best of breed with ad hoc integration. In this approach each system is selected on the basis of its specific functionality and subsequent integration is handled in an essentially ad hoc manner through a mixture of bilateral periodic data dumps and live data adaptors.
• Best of breed with central co-ordination. In this approach some form of central co-ordination, variously referred to as a Single Point Of Truth (SPOT), a central hub or central bus, mediates between systems when they share data.
• Service Oriented Architecture (SOA). This approach, in which resources on a network are made available as independent services that can be accessed without knowledge of the underlying platform implementation, was generally noted as an aspiration or destination rather than a currently existing strategy.

Most institutions were far from ‘pure’ exponents of these approaches and many institutions described themselves as being in transition from one approach to another. Institutions were most likely to describe their approach as some form of ‘Best of Breed’.

The key barriers to further integration were noted as:

• resource issues, including the costs of internal staff and external services
• lack of necessary skills in-house (particularly acute in the FE sector)
• internal resistance from academic and service departments determined to ‘do their own thing’ and protect ‘their’ data
• lack of representation of the integration issue at a senior management level
• a lack of appreciation in parts of the organisation of the multiple uses to which data is put and therefore a tendency to be concerned with the adequacy of data for local purposes only.

The de jure and de-facto responsibility for data integration varied widely from being almost completely regarded as a matter for the IT department to approaches which incorporate multiple levels of stakeholder input and oversight. There appeared to be very limited formal project evaluation of integration projects within institutions. Institutions reported a mixed experience of systems vendor involvement in the integration aspect of implementation and development projects and initiatives. Some vendors were seen to have very limited understanding of the HE/FE environment.

The risks associated with systems integration were varied and included the following:

• tight integration could lead to the propagation of errors (but also meant that errors could be fixed quickly when found)
• successful integration could render the effort to provide that integration invisible and therefore underappreciated by end users
• talk of integration could raise end user expectations leading to disappointment and souring relations
• user engagement and ownership do not automatically follow from systems integration
• the failure of vendors to sustain their enthusiasm for integration projects could leave the institution with half completed projects or large bills to complete projects
• problems of scope creep and scope control on integration projects

The Dynamics of Systems Integration

Many larger institutions have been seeking to make much stronger external linkages in the domains of teaching, learning and research (e.g., links with other institutions, with partner colleges, with spin-off companies and research collaborators, with overseas institutions, or with regional development agencies). As these linkages develop, they are generating new demands for information systems to support such activities which go beyond the boundaries of the organisation. The largest and most sophisticated institutions are looking more and more to develop a Service Oriented Architecture (SOA) with a central ‘enterprise bus’ through which systems functionality, configured as specific services, would interact with each other to provide a basic information infrastructure. This, it is hoped, will ease sharing data between systems through the use of standardised interfaces. Coupled with the emphasis on SOA, many of the larger institutions have begun to show an interest in a ‘shared services’ model in which certain institutions take on the role of supplying information systems functionality, over the web, as a service to other institutions.

Taken together, the interest in SOA and shared services suggest the decomposition of the current, internally focused systems integration challenge and the emergence of a new set of issues concerned with standardisation and the sharing of data and services between institutions. These pressures are creating demands for new systems architectures. Current architectures are generally ‘Enterprise’ architectures based on a simple identification of ‘the enterprise’ as ‘the institution’. The new environment which further and higher education is moving towards needs to be able to envisage, and support, ‘enterprise’ at a range of scales (multi-institutional partnerships, discipline based networks, overseas alliances, links with schools, etc.). This represents a new stage in the systems integration challenge.

Conclusions and Recommendations

Conclusions

Systems integration represents a significant and growing challenge for higher and further education institutions. This challenge is felt particularly strongly, and has been mostly addressed in practice, in larger higher education institutions. The majority of institutions currently have an essentially ad hoc approach to systems integration, but many have plans and projects which take a more strategic approach to system integration. New pressures are emerging for information systems to support a wide range of interactions which extend across the formal boundaries of the institution and this constitutes a new stage in the systems integration challenge. There is much interest in new information systems architectures, in particular so called Service Oriented Architectures (SOA), and in new service-oriented models of systems provision, including shared services.

Recommendations

The JISC has provided some support for aspects of the integration agenda in the past (for example, through support services such as JISC InfoNet and through R&D projects such as IoNodes). The present situation, however, suggests that many institutions would benefit from support with making sense of this new stage in the systems integration challenge. We propose five specific activities which the JISC could usefully support.

1. The JISC should provide a focus for those struggling with the systems integration challenge. This could take the form of an annual conference or a strand of an existing conference which would provide an opportunity for sharing experiences and good practice
2. The JISC should provide support for work to explore new systems architectures and business models (including SOA and shared services) and their implications for higher and further education
3. The JISC should support work with the vendors of systems in the higher and further education sector to explore their approaches towards integration
4. The JISC should support research into the end-user expectations and experience of systems integration (students, academic, academic related and administrative staff and various external partners)
5. The JISC should support work on the better evaluation of integration projects and approaches.

Report available electronically only. Read the full report below.