0
$\begingroup$

i am a beginner for panel data econometrics. Need help of experts of the field deciding if it is panel or pooled data or should i use any other methodology.

I have data from wind mill and we use wind energy for industrial purpose. At different height we have different material(approximately at each 10 feet difference, material is a different alloy) to handle to weight and wind movement.

To predict maintenance requirement, at different height i have for dependent variable as fracture crack in milimeter at the specific height. This is taken randomly and i have it for <35% of wind mills. Its done randomly for the mills annually and not a time series data because i don't have values for mills for which survey was done, year by year basis.

below are independent variables

  • height bracket
  • average wind speed
  • environmental variables: precipitation moisture in air, air pressure(This is summarized at annual level, so i only take extreme values, such as maximum air pressure, minimum air pressure and so on for other environmental variables.
  • material type
  • rotatin per minute of the wind mill.
  • age of the wind mill

There are hundreds of such wind mill in the field. I am building model to predict the damage crack in milimeter at the specific height given all the windmills of the field.

Can anyone suggest if it is panel data? It doesn't look like cross-sectional data though, because i am modelling all the mills of the field. Central tendency as well as dispersion of independent variables of the wind mill are different. It doesn't look like pool data because i don't have data for all the years for the mills for which data is available. Is it panel data?

So can anyone help me understand what type of data it is and more importantly how to model this data to normalize the variance between the mills throught panel data or otherwise?

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.