Merging datasets: HH or roaster which file should be master?

[ashar awan]

Hi,

My analysis will be like poverty headcount index, for fuel poverty using a panel data-set.

What do you suggest me to follow;
1. using roster as master file and add many other household related files (e.g electrical assests of HH) using m:1

or

2. using household level files and merging with roster (roster as master) using 1:m

or it does not matter at all?

[Marcos Almeida]

Both shall work accordingly.

[ashar awan]

Marcos Almeida Thank you so much