Pointless squiggles
Mood:
annoyed
Read/Post Comments (6)
I am going to do something which may be somewhat cliche and use my blog to gripe about my job.
Actually, it's not the job. I like my job and the people I work with. It's the County that's driving me nuts.
See, I was hired to do data entry. I am supremely overqualified to just type in numbers, so in order to not be bored, I've also modified our database so we can slice'n'dice those numbers in all sorts of ways, helping us identify who we're helping and how.
County Flub #1:
We need to identify the number of people we've seen for the first time in a given month AND, separately, the number of people who were seen this month that had been seen previously.
Sounds easy:
Count number of people who were seen for the first time this month, and count number of people who were seen this month that had also been seen previously. QED.
Directions from the County:
The first time you see someone, count them as New. If you see them a second time that month, count them the second time as Duplicate.
What?
We're not counting the number of times we see people; we're counting number of people seen. If every new person were seen at least twice in a given month, then you could say that N is a subset of D. However, this is not always the case. Therefore if N is only a partial subset of D, you never know how many people you've seen.
After explaining this to the County rep, she said she'd have to get back to me. She never did, but I was told by my boss that, at the next meeting, they had revised the requirements to coincide with my definition of counting each person once as either New or Duplicate within a given month.
County Flub #2
We gather information of how many people received which of our services (we provide several different types of services.) We also gather demographic data.
My boss had been told that our demographic data must exactly match our service data. Sounds great, except that it doesn't.
See, if a kid goes to one of our early literacy programs AND receives speech therapy, that kid is counted once in the Early Literacy row and once in the Speech Therapy row. That's fine, because it tells us which services are being used.
The problem arises because, if our demographics are supposed to exactly match the number of services provided, then that one kid will be counted twice in the demographics, and our numbers are small enough that double-counting even a couple of kids can significantly skew our data.
When I e'd the County, I was told someone would get back to me.
I can very clearly picture Karen Bowman (statistics guru, among other talents) looking blank, then saying, "What?! Are they insane?!?!!"
Numbers, like words, are merely symbols. They have little value if they have no consistent, underlying purpose.
[/rant]
QotD
Words differently arranged have a different meaning, and meanings differently arranged have different effects.
-- Blaise Pascal (1623–1662)
Actually, it's not the job. I like my job and the people I work with. It's the County that's driving me nuts.
See, I was hired to do data entry. I am supremely overqualified to just type in numbers, so in order to not be bored, I've also modified our database so we can slice'n'dice those numbers in all sorts of ways, helping us identify who we're helping and how.
County Flub #1:
We need to identify the number of people we've seen for the first time in a given month AND, separately, the number of people who were seen this month that had been seen previously.
Sounds easy:
Count number of people who were seen for the first time this month, and count number of people who were seen this month that had also been seen previously. QED.
Directions from the County:
The first time you see someone, count them as New. If you see them a second time that month, count them the second time as Duplicate.
What?
We're not counting the number of times we see people; we're counting number of people seen. If every new person were seen at least twice in a given month, then you could say that N is a subset of D. However, this is not always the case. Therefore if N is only a partial subset of D, you never know how many people you've seen.
After explaining this to the County rep, she said she'd have to get back to me. She never did, but I was told by my boss that, at the next meeting, they had revised the requirements to coincide with my definition of counting each person once as either New or Duplicate within a given month.
County Flub #2
We gather information of how many people received which of our services (we provide several different types of services.) We also gather demographic data.
My boss had been told that our demographic data must exactly match our service data. Sounds great, except that it doesn't.
See, if a kid goes to one of our early literacy programs AND receives speech therapy, that kid is counted once in the Early Literacy row and once in the Speech Therapy row. That's fine, because it tells us which services are being used.
The problem arises because, if our demographics are supposed to exactly match the number of services provided, then that one kid will be counted twice in the demographics, and our numbers are small enough that double-counting even a couple of kids can significantly skew our data.
When I e'd the County, I was told someone would get back to me.
I can very clearly picture Karen Bowman (statistics guru, among other talents) looking blank, then saying, "What?! Are they insane?!?!!"
Numbers, like words, are merely symbols. They have little value if they have no consistent, underlying purpose.
[/rant]
QotD
Words differently arranged have a different meaning, and meanings differently arranged have different effects.
-- Blaise Pascal (1623–1662)
Read/Post Comments (6)
Previous Entry :: Next Entry
Back to Top
