topic Re: Different characters definition throughout the years in Statistical Procedures

Different characters definition throughout the years

yael — Mon, 28 Jan 2019 14:02:24 GMT

Hello to you,

Thanks for your attention,

I have a question regarding sample characters definition.

For example, I have yearly observations of 300 firms for 10 years (2004-2013), total observations are 3000.

30 firms are defined as partnership ownership for the whole period (2004-2013). So, if I want to check the difference between partnership ownership to other firms, I check the gap between the 30 firms to 270 firms.

But, part of the "other firms" (270) holds also partnership ownership in specific years but not throughout all the years.

Will it be correct to add also these observations to the group of partnership ownership, or should I use only

the 300 observations that I mentioned above.

I hope that I clarify my question

Thanks a lot

Yael

Re: Different characters definition throughout the years

ballardw — Mon, 28 Jan 2019 15:34:36 GMT

That sort of membership may well depend on the particular analysis you are considering.

A number of options that come to mind are:

1) creating a separate category for those that switch

2) assigning membership in a category if a majority of the years are in that category

3) assigning membership based on the status at the start.

4) assigning membership based on the status at the end .

A particularly thorough analysis might consider all of these 4 and see if any of the particular rules make a difference.

Re: Different characters definition throughout the years

yael — Mon, 28 Jan 2019 17:40:46 GMT

@ballardw Thanks a lot!