Date: Mon, 20 Sep 2010 21:54:49 -0400
Reply-To: Richard Ristow <wrristow@mindspring.com>
Sender: "SPSSX(r) Discussion" <SPSSX-L@LISTSERV.UGA.EDU>
From: Richard Ristow <wrristow@mindspring.com>
Subject: Re: Failure to Comply Rate?
In-Reply-To: <201009171817.o8HHN60g007021@willow.cc.uga.edu>
Content-Type: text/plain; charset="us-ascii"; format=flowed
At 02:17 PM 9/17/2010, Damir wrote:
>I have a large dataset that contains individual charges for jail
>bookings. I need to identify jail pre-trial releases and whether
>their next admission is because they "failed to appear" for court.
>
>Some offenders are held at the jail until their trial and others are
>released from jail on various forms of release prior to their trial:
>1) own recognizance, bond, and something we call supervised
>release (VAR Release Reason). What I am trying to determine is
>the "failure to appear" rate for those offenders released for the
>above noted reasons.
>
>In some cases offenders may be admitted into the jail on new
>charges, and that this type of admission is not necessarily a
>"failure to appear" admission. However individuals may also be
>admitted as a result of new charges along with a failure to appear
>offense. Also, in some cases, individuals may be admitted into the
>county jail for failure to appear; however, the failure to appear
>resulted from a court hearing (called an arraignment) rather than
>after being released from the jail.
>
>My variables include:
>
>Inmateid (unique identification for inmates that follow them over
>time) (numeric)
>Admission id (unique identification for admissions with the same
>date and time) (numeric)
>BookingDateTime (Adatetime22)
>ReleaseDateTime (Adatetime22)
>ChargeCode (this includes the "failure to appear" charge) (numeric)
>ReleaseReason (this includes the code for the types of release
>reasons I noted above) (numeric)
>
>The unit of analysis for the dataset is charge based, so that a
>unique inmateid and/or admissionid may contain one or multiple records.
So you want to tabulate admissions that include a "failure to appear"
(possibly with other charges) by the release codes from the previous admission.
I'm not going to try code yet, but it wouldn't be hard to summarize
to a single record per admission, with
Inmateid
BookingDateTime
ReleaseDateTime
ReleaseReason
and whether the admission includes a "failure to appear" charge
It would be nice to have a "LEAD" function (analogous to "LAG") to
attach to each admission, whether the *following* admission includes
a "failure to appear" charge, but it'll probably take a little
juggling. Maybe it can be done with LEAD in CREATE -- anybody help with that?
Then, you need to identify the releases that are NOT followed by a
"failure to appear". I don't know how you do that. I suppose, if at
the inmate's next admission there's no "failure to appear", then that
settles it. What if there's no subsequent admission?
Anyhow, there are some moderately tricky data-manipulation issues
here; but the conceptual issue, how you determine whether an inmate
failed to appear after a release, needs to be resolved first.
=====================
To manage your subscription to SPSSX-L, send a message to
LISTSERV@LISTSERV.UGA.EDU (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD
| 
