Help on correlations for multiple observation

[James Paul]

Hi all,

Appreciate your help in solving this problem using SPSS

I have a data set that looks similar to this table. Each firm can adopt 1
strategy. There are 100 such firms. Tobin’s Q for a firm changes every year.
On an average there are about 5 observations of Tobin’sQ for a firm.

Firm Strategy Year Tobin’sQ
ABC 1 2008 1.2
ABC 1 2007 1.1
Abc 1 2006 1.3
def 2 2008 1.1
def 2 2007 1.0
deF 2 2006 1.2

My goal is to find any correlations between Strategy and Tobin’s Q. If I run
correlations in the data set as is (N=5*100), it will project statistical
significance that is actually not there.
Supposedly there is some way to define clusters and run the procedure with
N=100. How do I do this in SPSS (PASW 18)

Another person who uses Stata uses the command
xi:reg Y x1 x2 x3, cluster(firm)

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (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

[Swank, Paul R]

Depends on whether some firms use different strategies across the years. In that case, you would have a multilevel model and should use mixed.

Dr. Paul R. Swank,
Professor and Director of Research
Children's Learning Institute
University of Texas Health Science Center-Houston


-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of James Paul
Sent: Wednesday, May 12, 2010 2:11 PM
To: [hidden email]
Subject: Help on correlations for multiple observation

Hi all,

Appreciate your help in solving this problem using SPSS

I have a data set that looks similar to this table. Each firm can adopt 1
strategy. There are 100 such firms. Tobin’s Q for a firm changes every year.
On an average there are about 5 observations of Tobin’sQ for a firm.

Firm Strategy Year Tobin’sQ
ABC 1 2008 1.2
ABC 1 2007 1.1
Abc 1 2006 1.3
def 2 2008 1.1
def 2 2007 1.0
deF 2 2006 1.2

My goal is to find any correlations between Strategy and Tobin’s Q. If I run
correlations in the data set as is (N=5*100), it will project statistical
significance that is actually not there.
Supposedly there is some way to define clusters and run the procedure with
N=100. How do I do this in SPSS (PASW 18)

Another person who uses Stata uses the command
xi:reg Y x1 x2 x3, cluster(firm)

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (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

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (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

[Kylie]

Hi James,

Take a look at the these two articles:

Bland, MJ & Altman, DG. Calculating correlation coefficients with repeated
observations: Part 1 - correlation within subjects. BMJ 1995; 310: 446
(http://www.bmj.com/cgi/content/full/310/6977/446)

Bland, MJ & Altman, DG. Calculating correlation coefficients with repeated
observations: Part 2 - correlation between subjects. BMJ 1995; 310: 663
(http://www.bmj.com/cgi/content/full/310/6980/633)

Cheers,
Kylie.


Quoting James Paul <[hidden email]>:

> Hi all,
>
> Appreciate your help in solving this problem using SPSS
>
> I have a data set that looks similar to this table. Each firm can adopt 1
> strategy. There are 100 such firms. Tobin’s Q for a firm changes every
> year.
> On an average there are about 5 observations of Tobin’sQ for a firm.
>
> Firm Strategy Year Tobin’sQ
> ABC 1 2008 1.2
> ABC 1 2007 1.1
> Abc 1 2006 1.3
> def 2 2008 1.1
> def 2 2007 1.0
> deF 2 2006 1.2
>
> My goal is to find any correlations between Strategy and Tobin’s Q. If I
> run
> correlations in the data set as is (N=5*100), it will project statistical
> significance that is actually not there.
> Supposedly there is some way to define clusters and run the procedure with
> N=100. How do I do this in SPSS (PASW 18)
>
> Another person who uses Stata uses the command
> xi:reg Y x1 x2 x3, cluster(firm)
>
> =====================
> To manage your subscription to SPSSX-L, send a message to
> [hidden email] (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
>

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (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