Hello!
I am researching on the actual turnover behavior of the employees of non-profit organization, through applying the 'discrete-time hazard model'.
In order to do this, I changed the person-data to person-period data structure (non-profit organization employees between 2012-2016). From the following data, for me to find the actual turnover behavior, I have to see if employees from the previous year (t-1) have changed their occupation in the next years (t). My question in this is: in applying the discrete-time hazard model, am i supposed to remove the turnover value (ex. shown as '.' in the data set) from the very first year (t-1) or code it as '0', since this variable (turnover) isn't available for the very first year.
I assigned '0' for people who stayed in non-profit sector, '1' for people who have moved on, '2' for people who moved on to for-profit, '3' to government and '4' for unemployment.
Thank you in advance!
BHL
I am researching on the actual turnover behavior of the employees of non-profit organization, through applying the 'discrete-time hazard model'.
In order to do this, I changed the person-data to person-period data structure (non-profit organization employees between 2012-2016). From the following data, for me to find the actual turnover behavior, I have to see if employees from the previous year (t-1) have changed their occupation in the next years (t). My question in this is: in applying the discrete-time hazard model, am i supposed to remove the turnover value (ex. shown as '.' in the data set) from the very first year (t-1) or code it as '0', since this variable (turnover) isn't available for the very first year.
I assigned '0' for people who stayed in non-profit sector, '1' for people who have moved on, '2' for people who moved on to for-profit, '3' to government and '4' for unemployment.
Thank you in advance!
BHL
Attached Files
