I’m writing something about the importance to collecting monitoring and evaluation data in a way that makes it possible to then link to other datasets. I can’t find anything that does this in a simple way, but I’m sure such a thing must exist. Can anyone help?
This is what I’ve written so far. Any comments on it would be appreciated.
The critical enabler in using existing data sources is the quality, nature and organisation of the data that are held by the institution. If these data are held by the institution in a format that facilitates linking with other data then they can be extremely useful for evaluation. The features of a database that support the linking with other datasets are as follows.
- Absolute values. Are the data recorded as absolute values rather than percentages?
- Granularity. Is it possible to extract very fine grained and detailed information from the database? For example, can you find out how many people attended a course, how many people from a particular cohort and whether John Smith attended this course.
- Unique indicators. Does the data held have unique indicators that can be used to identify them? Many people share a surname or a data of birth but numbers like the Unique Pupil Number (UPN) or the UCAS Personal ID should be genuinely unique. Similarly for areas many areas share a town name but postcode is unique.
- Standard taxonomies. Are the data recorded using standard taxonomies? For example is the National Statistics Socio-economic Classification (NS-SEC) being used to classify the socio-economic background of learners.
Ensuring that these features underpin database design and data collection will mean that it is possible to link the data held with other data sources and to ensure that it is useful for evaluation.