comparing means

For questions regarding the import, export and manipulation of data in EViews, including graphing and basic statistics.

Moderators: EViews Gareth, EViews Steve, EViews Moderator, EViews Jason

michield
Posts: 2
Joined: Wed Dec 17, 2014 10:44 am

comparing means

Postby michield » Wed Dec 17, 2014 10:51 am

Hello,

I'm investigating the influence of investment on the ownershipstructure of public companies. Now I would like to compare the means of investment and ownership concentration. It's a time series from 2006 to 2012. I would like to know if there is a difference between the means of investment between the companies that have a ownership concentration lower than 50% and higher than 50%. So how do I make this split in ownership concentration and how do I compare those means?

thanks in advance!

EViews Glenn
EViews Developer
Posts: 2682
Joined: Wed Oct 15, 2008 9:17 am

Re: comparing means

Postby EViews Glenn » Wed Dec 17, 2014 10:59 am

Suppose the series for investment is INVEST and the concentration series is CONCENT (which has the percentage concentration).

Open the INVEST series.

Click on View/Descriptive Statistics & Tests/Equality Tests By Classification...

Enter the expression:

CONCENT>50

in the Series/Group for classify edit field. You should be able to leave everything else as is since you only have two categories. Click on OK.

michield
Posts: 2
Joined: Wed Dec 17, 2014 10:44 am

Re: comparing means

Postby michield » Thu Dec 18, 2014 4:53 am

But how do I interpreted this? CONCENT 0 has a mean of 0.51 and concent1 has a mean of 0.54. Does the zero stands for <50% en the one for >50%?

thanks

trubador
Did you use forum search?
Posts: 1520
Joined: Thu Nov 20, 2008 12:04 pm

Re: comparing means

Postby trubador » Thu Dec 18, 2014 6:22 am

Code: Select all

equation eq.ls invest c (concent>0.5)
Constant term (c) represents the cases of (concent<0.5) and therefore will be equal to 0.51. Coefficient of (concent>0.5), however, will give you the difference (i.e. 0.54-0.51=0.03). If the impact of owner concentration on investment differs between <50% and >50%, then you expect this coefficient to be statistically significant (e.g. p-value < 0.05).


Return to “Data Manipulation”

Who is online

Users browsing this forum: No registered users and 2 guests