Jail population finding is misleading

[eads]

It seems to be changing quite a bit day-to-day despite the large sample size. We need to figure out if there's a problem with our averaging calculation or just the concept.

[nwinklareth]

David can you explain a bit more in detail what you mean by this?

Also which population counts are you using?

[eads]

This is the first data finding we worked out on Saturday, more a tech demo than anything. I'm trying to build out the site into a worthy demo with some docs explaining how to plug in and develop for it.

Here's the text "Since 2013-07-22, the average number of bookings per day is 193.4 and the average number of departures per day is 191.5. This means that the jail population is changing by 1.9 persons per day."

I just look at the difference between the averages. I'm not sure that's telling me what I think it is.

[nwinklareth]

You are drawing a straight line from the beginning of a squiggly curve to its current end and all you are calculating is the difference between those two points. We are dealing with a system that sort of balances itself. Each day there are people who enter the system and people who leave the system. The numbers for each vary by quite a bit, with some days there are more people entering the system then leaving and other days the reverse. We also see trends, with a increase in population from the middle of July peaking in the middle of September and then falling off. What you could report is the average for each with a standard deviation but just how understandable would that be to people.

Let me back off and asking you this question: what is it you would like to communicate with this? What change to you want to have happen in the reader when the read this?

[eads]

This whole section is up for grabs at the moment -- I'm not entirely sure
what we want in this section. I'll work up a slightly refined alternative
and get your feedback.

On Tue, Dec 3, 2013 at 11:18 PM, nwinklareth [email protected]:

You are drawing a straight line from the beginning of a squiggly curve to
its current end and all you are calculating is the difference between those
two points. We are dealing with a system that sort of balances itself. Each
day there are people who enter the system and people who leave the system.
The numbers for each vary by quite a bit, with some days there are more
people entering the system then leaving and other days the reverse. We also
see trends, with a increase in population from the middle of July peaking
in the middle of September and then falling off. What you could report is
the average for each with a standard deviation but just how understandable
would that be to people.

Let me back off and asking you this question: what is it you would like to
communicate with this? What change to you want to have happen in the reader
when the read this?

—
Reply to this email directly or view it on GitHubhttps://github.com//issues/12#issuecomment-29780304
.

David Eads - 773.354.2285
News applications developer, Chicago Tribune (
http://www.chicagotribune.com/news/data/)
Founder, FreeGeek Chicago (http://freegeekchicago.org)

"Medical statistics will be our standard of measurement: we will weigh life
for life and see where the dead lie thicker, among the workers or among the
privileged." -- Rudolf Virchow

[bepetersn]

I find Norbert's explanation plausible, personally. As for the purpose of the finding itself, ... I think it was just an experiment anyway.

[eads]

Yeah, I don't think we do either. I would like to put something interesting
there...

On Tue, Dec 3, 2013 at 11:50 PM, Brian Everett Peterson <
[email protected]> wrote:

I find Norbert's explanation plausible, personally. Maybe we don't need
the population trend "finding".

—
Reply to this email directly or view it on GitHubhttps://github.com//issues/12#issuecomment-29781330
.

David Eads - 773.354.2285
News applications developer, Chicago Tribune (
http://www.chicagotribune.com/news/data/)
Founder, FreeGeek Chicago (http://freegeekchicago.org)

"Medical statistics will be our standard of measurement: we will weigh life
for life and see where the dead lie thicker, among the workers or among the
privileged." -- Rudolf Virchow