Gah! Just Gah! Data access on our campus is somewhat of a sticky question. It’s not that we don’t want people to be able to get into data they need and create their own reports, it’s that our database is truly not relational or normalized in *any* way. It has.. personality. *charming smile* It’s a mainframe database originally coded in COBOL that got wrapped up in enough SQL and bubble gum to dump into an Oracle database structure some four or five years ago.
But even though it’s in a relational database structure, it’s *not* relational. There are no common primary keys. There are no automatic joins between tables. Some tables that should be able to talk to each other, can’t. Fields can be named identically in different tables and mean entirely different things, and fields that are named entirely differently in different tables might contain the same data. Codes that might have meant one thing ten years ago have been co-opted and used to mean different things now, often without documentation (or consistency, but that’s a data-entry problem, not a data-structure problem). To add to the confusion, it’s an operational database designed to support operational business functions, not a datamart or datawarehouse designed to support analytical functions.
This is no secret. I was warned when I interviewed that the database was only so loosely speaking. I’ve been here slightly over a year and I’m only just scratching the surface of learning the quirks and ins-and-outs of our database. And I work *daily* with our data and our database, which ought to say something about the level of complexity we’re dealing with.
So when people on campus want to get direct query access to our database so they don’t have to go through one of the gatekeepers, it causes all kinds of hullaballoo. Trust me, I’d *love* to be able to set up access for common requesters of data and teach them to fetch their own data. It would save me quite a bit of time and make them happier to not have to wait (even though the average wait time on a routine data request tends to be less than 48 hours, but that’s an issue for another day). But given the complexity of our data structure, it’s not feasible.
The support required to train them on even the basics would be beyond our current training capacity, to say nothing of trying to support them on more complex queries, which they’d inevitably want to dive into right away. And trying to explain to people that while they might be getting *an* answer, because of the intricacies of our database it may not the *right* answer just results in unnecessary confusion and an even larger expenditure of time to untangle, compounded exponentially if they’ve gone crying to the four winds that they’ve got *the* answer.
(Data use, in general, is also something that the majority of campus could use some education on. For instance, when a form asks for the number of first-time, first-degree seeking new freshmen for a given year, the answer will *never* be over 7,000 for our campus. Never.)
So the answer when someone – even someone who could otherwise legitimately have data access – requests carte blanche data access is almost invariably “no”. Which usually results in a backlash of conspiracy theories about how the administration is hiding large pools of money (it’s all always about money, isn’t it?) and they won’t let people access the data because someone will find them out. *sigh*
This has been my week. Happy New Year..?
(Two work-related posts in as many days, with no new knitting/spinner/fiber content in either. Sorry about that!)