dummy variables in regression

[Carole Ann Rollins]

I am finishing my statistical analysis for my PhD. I searched the archives and could not find the answer to my question.

I just started using SPSS.  I have an anaysis involving a data set of
500 homes. 250 homes have xeriscape landscaping and 250 homes have
non-xeriscape landscaping. I have the house sales prices, house attributes (15 variables). I want to predict the value of xeriscape landscaping.

Questions:

1.  Do I put the xeriscape as a variable in my data set (with xeri=1 and
non-xeri=0)? Do I also put non-xeriscape as a variable in my data set (with
non-xeri = 1 and xeri = 0?

2.  When I go to enter data for the linear regression analysis do I put sales
price as my dependent variable?

3.  Do I put all the other house attributes (excluding xeriscape and non-xeriscape
as the independent variables?

4.  Where do I put the dummy variable xeri = 1 and non-xeri = 0?

Thanks.

Carole Rollins
415-898-5895

Nature Technologies International LLC
Carole Ann Rollins
P.O.Box 1519
Novato, CA 94948
415-898-5895
707-225-5762
fax: [phone]
http://www.nature-technologies.com
[hidden email]

[Björn Türoque]

The answers to your question depend on the theoretical model you are
testing, but your xeriscape variable would be considered one of your
independent variables. You might want to think about running your regression
once with all of your other independent not including xeriscape landscaping
variable, and then running it again with all the other independent variables
and the xeriscape landscaping variable. The results could be quite
interesting.

Also if you enter your dependent variable (Home sale price) in whole
dollars, your results under B will give you the results in dollars. I am
assuming you think xeriscape landscaping will increase the value of a
home... if this is the case, it might be easier for interperitation if you
coded the variable as 1 being has landscaping and 0 being does not have
landscaping. that way the B that comes up will indicate the increse in value
associated with having landscaping.

Björn Türoque

Some people are just born to rock!


On 8/14/07, Carole Ann Rollins <[hidden email]> wrote:

>
> I am finishing my statistical analysis for my PhD. I searched the archives
> and could not find the answer to my question.
>
> I just started using SPSS.  I have an anaysis involving a data set of
> 500 homes. 250 homes have xeriscape landscaping and 250 homes have
> non-xeriscape landscaping. I have the house sales prices, house attributes
> (15 variables). I want to predict the value of xeriscape landscaping.
>
> Questions:
>
> 1.  Do I put the xeriscape as a variable in my data set (with xeri=1 and
> non-xeri=0)? Do I also put non-xeriscape as a variable in my data set
> (with
> non-xeri = 1 and xeri = 0?
>
> 2.  When I go to enter data for the linear regression analysis do I put
> sales
> price as my dependent variable?
>
> 3.  Do I put all the other house attributes (excluding xeriscape and
> non-xeriscape
> as the independent variables?
>
> 4.  Where do I put the dummy variable xeri = 1 and non-xeri = 0?
>
> Thanks.
>
> Carole Rollins
> 415-898-5895
>
> Nature Technologies International LLC
> Carole Ann Rollins
> P.O.Box 1519
> Novato, CA 94948
> 415-898-5895
> 707-225-5762
> fax: [phone]
> http://www.nature-technologies.com
> [hidden email]
>

--

[Manmit Shrimali-2]

Team,

I have 11 attributes measured on 1-7 scale. Attributes measure the importance of feature consumer consider when choosing product x. I also have DV i.e. performance of product X. Eventually I will perform factor analysis followed by cluster analysis and discriminant analysis. Then also linear regression.

My question is that in scale data should we center the means as I have in goole search that in few examples centering is done before performing analysis.
I request all experts to please share their experience and help me understand why centering is required and in what situation we should use.

Any input is highly appreciated.

[Richard Ristow]

There's a lot to be said, but one or two points ...

At 02:13 PM 8/14/2007, Carole Ann Rollins wrote:

>1.  Do I put the xeriscape as a variable in my data set (with xeri=1
>and non-xeri=0)? Do I also put non-xeriscape as a variable in my data
>set (with non-xeri = 1 and xeri = 0?

Customarily, you'd put the first one ('xeri') in your data set. The two
variables are redundant with each other, so it's of no use, and
potentially confusing, to include both.

>2.  When I go to enter data for the linear regression analysis do I
>put sales price as my dependent variable?

You wrote "I want to predict the value of xeriscape landscaping." If
your operational definition of 'value' is 'house sales price', then,
yes, sales price is the dependent variable. (Does the term 'operational
definition' make sense to you? It means choosing a measurable quantity
as a measure of a conceptual one. 'Sales price' is measurable; there's
a record of it. 'Value' is an abstract quantity. But it seems
reasonable to use sales price as a measure of value. But you MUST state
this, and your reasons for thinking it justified, in the Methods
section of your dissertation. Ask your advisor.)

>3.  Do I put all the other house attributes (excluding xeriscape and
>non-xeriscape as the independent variables?

Yes, you do; but here you need to be careful. Talk this over with your
advisor before running.

You write that you have 15 house attributes. That's a lot of
independent-variable coefficients to interpret. Before putting any one
in, check
. That it's at interval level of measurement
. That you believe it should have an effect on house price.
Here, again, you should have a discussion with your advisor.

>4.  Where do I put the dummy variable xeri = 1 and non-xeri = 0?

It's an independent variable.

-Good luck,
  Richard

[Ornelas, Fermin]

You may want to do a research on hedonic estimation. These studies deal
with real estate market studies and could help you come up with a
relevant hypothesis as well as model specification to be tested for your
study.

I am not sure that having the variable of interest on the right hand
side of the estimation equation can render much relevance to your study,
especially for a Ph.D. dissertation.

Fermin Ornelas, Ph.D.
Management Analyst III, AZ DES
1789 W. Jefferson Street
Phoenix, AZ 85032
Tel: [phone]
E-mail: [hidden email]


-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
Carole Ann Rollins
Sent: Tuesday, August 14, 2007 11:13 AM
To: [hidden email]
Subject: dummy variables in regression

I am finishing my statistical analysis for my PhD. I searched the
archives and could not find the answer to my question.

I just started using SPSS.  I have an anaysis involving a data set of
500 homes. 250 homes have xeriscape landscaping and 250 homes have
non-xeriscape landscaping. I have the house sales prices, house
attributes (15 variables). I want to predict the value of xeriscape
landscaping.

Questions:

1.  Do I put the xeriscape as a variable in my data set (with xeri=1 and
non-xeri=0)? Do I also put non-xeriscape as a variable in my data set
(with
non-xeri = 1 and xeri = 0?

2.  When I go to enter data for the linear regression analysis do I put
sales
price as my dependent variable?

3.  Do I put all the other house attributes (excluding xeriscape and
non-xeriscape
as the independent variables?

4.  Where do I put the dummy variable xeri = 1 and non-xeri = 0?

Thanks.

Carole Rollins
415-898-5895

Nature Technologies International LLC
Carole Ann Rollins
P.O.Box 1519
Novato, CA 94948
415-898-5895
707-225-5762
fax: [phone]
http://www.nature-technologies.com
[hidden email]

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.