We Like Bad Data
Ian Bailey
Given such a development procedure (and yes, I’ve exaggerated but you all know it’s true) is there even the remotest chance that the system can meet the real user requirements ? No, of course not – and they never do, we simply move the goalposts as the project progresses so we can claim success at the end. And we do it time and time again. On multi-million dollar projects.
I’ve digressed into another favourite rant there, and all that was just to describe how a text field came into being. What about data quality ? Those users who have been stuffing the wrong type of information into text fields – has anyone thought that perhaps this is information they need to do their job, and they just had nowhere else to put it ? Maybe the system they were using was so shoddily designed that actually they need to abuse it just to be able to carry out their jobs ?
Much is made of re-factoring in software development, but what about re-factoring the information ? Surely, rather than go through the whole comic tragedy of process and data modelling one more time, we could simply look at the legacy data and see what the users actually do ? Oh no, sorry, that’s poor quality data, isn’t it ?
‹ prev | 1 2 3 4 | next ›
del.icio.us