Building a model

[Alina Sheyman-3]

I'm trying to choose between two ways of building a regression model,
namely between two ways of representing my dependent variable. Here is the
model that I am building - I'm looking for a relationship between the
number of people that registered for a specific program (y-value) and how
they found out about a program - tv, radio, newspaper (x-values). My two
choices here are to use the actual number of registrants or to use a new
measure that I've created - a ratio between new registrants and total
number of ads they were exposed to. My intuition tells me that since I am
looking at several months worth of data, where the number of available ads
varied significantly, it makes more sense to look at the ratio rather than
the number of registants. Is this correct or am I forcing x and y to be too
correlated?

thank you,
Alina Sheyman

[Ornelas, Fermin]

Both approaches are worth the try. Here the exploratory analysis should
help you decide how well each predictor is correlated with the response
variable and with each other. Moreover modeling diagnostics
(satisfaction of basic assumptions, F-test, t-statistics, confidence
intervals) will help you decide which model is more statistically
robust. Having said that, it seems that a third approach could be done.
If you define the outcome as binary for participants and non
participants as function of the different add exposures then you will
have a probability value of participation which is also interesting.
Then you could sell the idea of who is more likely to enroll in the
service. For this you could use logistic regression.

Fermin Ornelas, Ph.D.
Management Analyst III, AZ DES
Tel: [phone]
E-mail: [hidden email]


-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
Alina Sheyman
Sent: Friday, April 06, 2007 8:45 AM
To: [hidden email]
Subject: Building a model

I'm trying to choose between two ways of building a regression model,
namely between two ways of representing my dependent variable. Here is
the
model that I am building - I'm looking for a relationship between the
number of people that registered for a specific program (y-value) and
how
they found out about a program - tv, radio, newspaper (x-values). My two
choices here are to use the actual number of registrants or to use a new
measure that I've created - a ratio between new registrants and total
number of ads they were exposed to. My intuition tells me that since I
am
looking at several months worth of data, where the number of available
ads
varied significantly, it makes more sense to look at the ratio rather
than
the number of registants. Is this correct or am I forcing x and y to be
too
correlated?

thank you,
Alina Sheyman

NOTICE: This e-mail (and any attachments) may contain PRIVILEGED OR
CONFIDENTIAL information and is intended only for the use of the
specific
individual(s) to whom it is addressed.  It may contain information that
is
privileged and confidential under state and federal law.  This
information
may be used or disclosed only in accordance with law, and you may be
subject to penalties under law for improper use or further disclosure of

the information in this e-mail and its attachments. If you have received

this e-mail in error, please immediately notify the person named above
by
reply e-mail, and then delete the original e-mail.  Thank you.

[Luo, Peter-2]

How many months are we talking here? If there are more than 20, then you
may need to apply time series analysis, i.e. to link monthly registrants
to monthly ad volume, or lagged monthly ad volume. That would deal your
concern that ad quantity varied significantly over time.

What is your analysis unit, e.g. individual, mall, week, or month?


-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
Alina Sheyman
Sent: Friday, April 06, 2007 11:45 AM
To: [hidden email]
Subject: Building a model

I'm trying to choose between two ways of building a regression model,
namely between two ways of representing my dependent variable. Here is
the
model that I am building - I'm looking for a relationship between the
number of people that registered for a specific program (y-value) and
how
they found out about a program - tv, radio, newspaper (x-values). My two
choices here are to use the actual number of registrants or to use a new
measure that I've created - a ratio between new registrants and total
number of ads they were exposed to. My intuition tells me that since I
am
looking at several months worth of data, where the number of available
ads
varied significantly, it makes more sense to look at the ratio rather
than
the number of registants. Is this correct or am I forcing x and y to be
too
correlated?

thank you,
Alina Sheyman


This message is the property of Draftfcb and contains information which may
be privileged or confidential. It is meant only for the intended recipients
and/or their authorized agents.  If you believe you have received this
message in error, please notify us immediately by return e-mail and destroy
any printed or electronic copies of the message.  Any unauthorized use,
dissemination, disclosure, or copying of this message or the information
contained in it, is strictly prohibited and may be unlawful.
Thank you for your cooperation.
(A)