Creating labels and getting the right output

Moderators: statman, Analyst Techy, andris, Fierce, GerineL, Smash

kjellkristoffer
Posts: 4
Joined: Fri Mar 23, 2012 9:48 am

Creating labels and getting the right output

Postby kjellkristoffer » Wed Apr 18, 2012 12:08 pm

Hi.

Im doing a correlation test between price and quality on consumer products, and have collected data on 50 productgroups.
The values on price range from 0 - 20 000, and the score on quality range from 1 - 100.

What I want is to do a correlation (one tailed) test between these two variables, using spearmans rho.

The problem I got is that I want to separate these data by type of products (Frequent and non-Frequent).

How can I create a dummy variable or label these data to be seperated into different products and productgroups on the output?

An example:
Non-Frequent products:
Washingmachines: N (20) - Correlation .64 (between price and quality)
Blenders: N (8) - Correlation .34
......

Frequent products:
Toilettpaper: N (19) - Correlation .50
Dog food: N (12) - Correlation .43
.....


Please help me:)

Regards Kjell.
apeape
Posts: 56
Joined: Mon May 02, 2011 6:07 pm

Re: Creating labels and getting the right output

Postby apeape » Wed Apr 18, 2012 1:43 pm

Once you have a variable that designates products as frequent or non-frequent purchases, it is a simple matter of using the Data -> Split File -> Compare Groups (based on frequency variable) option.
kjellkristoffer
Posts: 4
Joined: Fri Mar 23, 2012 9:48 am

Re: Creating labels and getting the right output

Postby kjellkristoffer » Tue Apr 24, 2012 9:40 am

apeape wrote:Once you have a variable that designates products as frequent or non-frequent purchases, it is a simple matter of using the Data -> Split File -> Compare Groups (based on frequency variable) option.
Thanks for answer:)
Another thing im having problem with is how to compute my variables, if I need to.. ?

As said before I got a dataset with 2 variables, quality and price. Problem is this: quality only range from 1-100, while price range from 1-35 000 NOK, because of the different productgroups (a HDtv is more expensive than toilettpaper..)

How can I do a correlation study (average correlation) for all the productgroups? Wouldnt it be wrong if I took all and runned a bivariate onetailed spearmans rho, on all these, since it then would corrrelate a steady score agianst different price "ranges"?

The only way I can think of is doing all productgroups seperated, and then add all the correlations and then devide on number of productgroups?

Last question: some of the tests i got in the dataset is using a different quality rank, where 5 is max and 1 is the lowest, how can I compute these so that 5 = 100 and 1=1 like my other quality rank? 4,9 = ? 4,8 = ?

Who is online

Users browsing this forum: No registered users and 1 guest

cron