Data validation and acceptance testingData validation and acceptance testing

Introduction

The earlier that data quality issues are found, the smaller their impact on the implementation project. It is therefore recommended to build validation into your work when populating your Pure with data - whether you use synchronizations, bulk imports, or the API. Migrating legacy data in particular often requires several iterations to refine the data mapping and conversion. Consider also how your data will be validated when your Pure system is live.

Ensure the configurations of your Pure work for the people who will be using it by carrying out acceptance testing with different groups.

Goals

• Ensure data quality
• Test your Pure from the perspectives of different users

Data validation

Who will be responsible for validation of the data in your Pure, and will it be automated or manual? Some of the work may be done by the IT specialist handling the conversion of any synchronised or imported data. HR or departmental staff may validate Person data, while researchers and library staff handle Research Outputs, and research management staff ensure the quality of funding data.

Synchronisations and imports of data produce logs showing errors that can be investigated and resolved, allowing some level of automated validation.

Workflows can be configured per content type so that new and/or modified records will be sent to users with specific roles for manual validation.

Acceptance testing

You might choose to define a set of acceptance tests to perform before going live with Pure, involving Pure administrators and users with various roles. For example, you could form a group of researchers and ask them to pilot the system by logging in to Pure and experimenting with adding their Research Outputs and updating their Portal profiles. Such a group could also help you launch your Pure by acting as ‘ambassadors’.