Message-ID: <10471972.1075853136867.JavaMail.evans@thyme>
Date: Tue, 5 Jun 2001 14:15:00 -0700 (PDT)
From: felecia.acevedo@enron.com
To: andrea.yowman@enron.com, elspeth.inglis@enron.com
Subject: RE: PRC Demographics
Cc: michelle.cash@enron.com, gina.corteselli@enron.com
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: quoted-printable
Bcc: michelle.cash@enron.com, gina.corteselli@enron.com
X-From: Felecia Acevedo <Felecia Acevedo/ENRON@enronXgate@ENRON>
X-To: Andrea Yowman <Andrea Yowman/ENRON@enronXgate>, Elspeth Inglis <Elspeth Inglis/ENRON@enronXgate>
X-cc: Michelle Cash <Michelle Cash/HOU/ECT@ECT>, Gina Corteselli <Gina Corteselli/ENRON@enronXgate>
X-bcc: 
X-Folder: \MCASH (Non-Privileged)\Cash, Michelle\PRC
X-Origin: Cash-M
X-FileName: MCASH (Non-Privileged).pst


When looking at representation and determining if there is advserse impact =
you need to look at the percentages as below:


                 Total              1        =09            2         =09  =
    =09 3                      4                     5                 NR =
=20
Males          26=09        4=09 (15.4%)      16 (61.7%)           6=09(23.=
1%)         0  (0.0%)         0 (0.0%)       0 (0.0%) =20

Females       6           0   ( 0.0%)        3 (50.0%)            2 (33.3%)=
         1 (16.7%)        0 (0.0%)       0 (0.0%)


You want to determine if women (or minorities, or over 40), based on their =
representation, are being adversely effected.  You look at the total women =
and how their distribution over the ratings equates, then you look at the t=
otal men and how their distribution equates--you are looking at them indivi=
dually and not overall.  After this is calculated you determine by rating c=
ategory if women are within 80% of the ratings men received-- if they are w=
ithin 80% there is no adverse impact.  If it is less than 80% there is adve=
rse impact.  For example, let's look at the "2" category.  There are 16 of =
26 men that received "2's" or 61.5% of the male representation.  There are =
3 of 6 females that received "2's"  or 50% of the female representation.  T=
o determine if the females are being adversely effected you take 50% and di=
vide it by 61.5% and get 81.3%--as this is more than 80% there is no advers=
e impact.  Now, let me throw a curve ball.  In calculating adverse impact i=
t is assumed that if women (or minorites) are being effected more it is a b=
ad thing.  In the case of looking at ratings this may not be the case so a =
standard formula cannot be applied.  For example,  if more women are gettin=
g "1's" than the males this is not a bad thing but if more women are gettin=
g "5"s this is a bad thing!  You have to use some judgement on this because=
 if more female's based on their representation are getting 3's maybe that =
is bad and maybe that is good--you have to look at the overall picture.  Wh=
en you are calculating adverse impact on employment actions like hires or t=
erminations it is very straight forward but when looking at it from a ratin=
g perspective it can get tricky because you are making comparisions over mu=
tiply categories which mean different things--the 1's being good and the 5'=
s being not so good.=20

I would recommend that you proceed with the chart as I typed it above and n=
ot try to build in the adverse impact piece.  The HR person looking at the =
data in the PRC meeting will need to do some quick calculations based on ho=
w the distribution is for determining adverse impact.  If you have any ques=
tions or if I can help please let me know.


=20

 -----Original Message-----
From: =09Yowman, Andrea =20
Sent:=09Tuesday, June 05, 2001 10:05 AM
To:=09Acevedo, Felecia
Subject:=09FW: PRC Demographics

Please give me your thoughts

 -----Original Message-----
From: =09Inglis, Elspeth =20
Sent:=09Tuesday, June 05, 2001 9:28 AM
To:=09Yowman, Andrea; Cash, Michelle
Cc:=09Corteselli, Gina
Subject:=09FW: PRC Demographics

FYI

this is the revised demographic screen. Please review and advise. thanks
Elspeth

 << OLE Object: Picture (Device Independent Bitmap) >>=20



<Embedded Picture (Metafile)>