topic Re: Difference between data columns in Statistical Procedures

Difference between data columns

yael — Tue, 12 Feb 2019 08:06:58 GMT

Hello

I am asking for your help,

I have 5 continuous variables while each of it has 5 categories.

For example, a variable of the "firm age" has five existing categories such as an observation can be shown only in one category.

Or a variable of "firm size" has also 5 categories and a single observation can only be in one of the existing categories.

My question - what are the statistical ways for checking the difference between the categories (columns)? F test is the only way?

Thanks for your help

Re: Difference between data columns

Kurt_Bremser — Tue, 12 Feb 2019 12:48:04 GMT

Please supply some example data in usable form (data step with datalines, see my footnotes) and what kind of "difference" you want to get out of it.

Re: Difference between data columns

PaigeMiller — Tue, 12 Feb 2019 13:15:09 GMT

It's really unclear what you are saying, as as suggested above by @Kurt_Bremser, providing a portion of your actual data, and an example of the desired output, would make this whole situation clear. Specifically:

I have 5 continuous variables while each of it has 5 categories.

This sentence really is hard to decipher. Continuous variables cannot have 5 categories.

My question - what are the statistical ways for checking the difference between the categories (columns)? F test is the only way?

What difference are you referring to? What statistic? Difference in means? Difference of standard deviations? Difference in distributions?

What variable are we looking at to find a difference?

Re: Difference between data columns

yael — Tue, 12 Feb 2019 13:23:20 GMT

@Kurt_Bremser thanks for your reply.

I hope that my following sample will be clearer.

I have the following data:

			Firm stages
variables / stages	1	2	3	4	5
Sales Growth (%)	6	11	4	2	-3
Firm Age	30	33	34	34	30
Debts / T. assets (%)	60	62	55	57	51

These figures are based on 5000 observations.

The variables figures are means in each firm stage.

There are criterions for each firm stage and the variables figures are divided into the firm stages by these criteria.

I want to check if there is a statistical significance between the firm stages means.

I believe I can check every 2 stages with the t-test, but I wonder if there is any other way to check it?

Thanks a lot for your remarks

Have a nice day

Yael

Re: Difference between data columns

PaigeMiller — Tue, 12 Feb 2019 13:25:43 GMT

I want to check if there is a statistical significance between the firm stages means.

You can't do this with the data as presented. You need some estimate of the standard deviations in order to perform statistical tests.

Re: Difference between data columns

Kurt_Bremser — Tue, 12 Feb 2019 13:26:42 GMT

DATA

STEP

WITH

DATALINES.

Please.

What you posted tells us NOTHING about the structure and attributes of your dataset.

Re: Difference between data columns

yael — Tue, 12 Feb 2019 13:29:01 GMT

Thanks a lot! Yael