for instance bank data which can be represented in panel and macroeconomic data which is time series?
You can use both types of data in one panel data model. I don't know what exactly is your research plan but I assume that you are estimating a model in which the dependent variable is a bank-specific indicator observed in a given year and explanatory variables are either other bank characteristics or macroeconomic indicators. In this case your model may be of the form
bank_ind1it=alfai+beta1*bank_ind2it+ beta2*macro_indt+epsilonit
There are of course some issues while using specification of this type (for example you should use time-series unit root test for macro_ind and panel unit root tests for remaining variables) but they don't prevent from using this specification. I estimate this type of model in the attached paper (published also in Applied Economics).
Piotr Cizkowiczi is totally correct.
However I just would like to add for the sake of general knowledge, that more complex modern techniques regarding to panel data series in multiple variable analysis are cointegration and sctochastic techniques, usually used for fixing problems in data time series and also convenient in forecasting trough panel data.
I recommend this paper as a good example for cointegration.
Article Finance-Growth-Crisis Nexus in India: Evidence from Cointegr...
- Badges
- Science topic
- Similar topics
- Statistics
- Time Series