Data Management

Very often the impetus behind the introduction of a new system is to make information accessible to a greater number of users - where this is not explicitly the case it is often a by-product of the change. For example linking the registration data from student records to the creation of accounts and module registrations within learning environments suddenly exposes the data in the administrative systems to the learner at a much earlier stage, or perhaps even for the first time. Similarly the increase in availability of web 'self-service' functionality for staff and students alike to carry out basic administrative tasks reveals data that had previously been for the most part hidden from the subject. Good information and data management has always been an essential part of any business process and of making effective use of IT systems. The difference now is that any errors in the data are more highly visible and the dissatisfaction of these 'new' users often manifests itself in blaming the system as a poor product rather than identifying the business processes and/or the data preparation and processing as the root cause. Some of these issues are examined in more detail in a short paper that accompanied a JISC infoNet presentation to an ILT web special interest group.

All of this means that when an organisation decides to implement a new system, one of the tasks that has to be undertaken with great care and with precise timing is the process of moving data from one or more old systems to the new. This is usually referred to as Data Migration. The importance, complexity and duration of the data migration process must not be underestimated, as it almost always goes wrong at some stage, and work should start as early as possible. Typically, data migration is performed by a set of customised programs or scripts that are written by IT Services staff and/or external technical consultants to automatically transfer the data from the old system(s) to the new. Time and resources must be allocated for writing the customised programs when planning data migration activities.

Data migration comprises four fundamental processes:

• Data Cleansing
• Data Mapping
• Data Conversion and
• Data Population

It ought to go without saying that an essential part of managing the data and ensuring data quality is to understand what information you are holding and why. There is a section in the Process Review infoKit on Information and Data. You may also wish to look at the infoKit on Records Management.