Hello everyone,
I have a dyadic unbalanced panel dataset where every US listed firm is being compared to all other US listed firms in that same industry, done for every industry. For every observation, firm A is being compared to firm B over a certain duration (ranging between 1985 and 2010, however, the period is not of the same duration or covering the same years for every pair of firms). So, for example: Figgie Intl is paired with MDT Corp between 1988 and 1995. But Figgie Intl is paired with Respironics between 1990 and 2000. For each year we have for both firms the number of employees, the revenue, the unique identifier, and the name of the firm. Additionally, we have the number of inventors during the previous 5 years, for both firms, as well as the mobility count (from firm A to B as wel af from firm B to A) in that year and the number of US states in which both firms are collocated.
In my research, I want to analyze the effect of firm colocation (independent variable) on the mobility of inventors (dependent variable).
I am currently wondering how I analyse this data properly. I am encountering a challenge due to the inconsistency in the years covered by the data across different dyads. As the dataset had over 150.000 observations, I am a little bit lost at how to start. I thought about creating dummy variables for each period, but because the periods are so inconsistent, I don't know how to handle this.
I am thankful for any advise and support.
Kind regards,
Sana
I have a dyadic unbalanced panel dataset where every US listed firm is being compared to all other US listed firms in that same industry, done for every industry. For every observation, firm A is being compared to firm B over a certain duration (ranging between 1985 and 2010, however, the period is not of the same duration or covering the same years for every pair of firms). So, for example: Figgie Intl is paired with MDT Corp between 1988 and 1995. But Figgie Intl is paired with Respironics between 1990 and 2000. For each year we have for both firms the number of employees, the revenue, the unique identifier, and the name of the firm. Additionally, we have the number of inventors during the previous 5 years, for both firms, as well as the mobility count (from firm A to B as wel af from firm B to A) in that year and the number of US states in which both firms are collocated.
In my research, I want to analyze the effect of firm colocation (independent variable) on the mobility of inventors (dependent variable).
I am currently wondering how I analyse this data properly. I am encountering a challenge due to the inconsistency in the years covered by the data across different dyads. As the dataset had over 150.000 observations, I am a little bit lost at how to start. I thought about creating dummy variables for each period, but because the periods are so inconsistent, I don't know how to handle this.
I am thankful for any advise and support.
Kind regards,
Sana
