A panel data set (also longitudinal data) has both a cross-sectional and a time series dimension, where all cross section units are observed during the whole time period. Perform all the diagonistic tests required in case of a panel data. do panel staionarity test . check for cross section dependence. Can perform panel ARDL.

Pooled OLS seems an appropriate option since some data lack the time dimension. Hope this helps.

I think that by applying regression and correlation analysis you will achieve good enough results for the purposes of your research. Perform tests for the adequacy of the model and for the significance of parameters.

If you do not get the desired results, then apply other methods.

Success!

The following models can be applied:

1. Pooled panel model,

2. Dynamic panal model,

3. Fixed effects model, and

4. Random effects model.

With the following regression equation;

Yit = a + bXit + eit , where

i = individuals, and t = over the years. For example, in fixed effects, eit 'error term' is non-stochastic, whereas in random effects it is stochastic.

You can do cross section, and run another pseudo panel data as robustness check

Try to use more advance system dynamic model on your data after testing for edogeneity and auto correlation.

Tell us more what you want to evaluate/consider so that we can advise because cross-section versus panel data tell different stories as follows:

(i) Cross-sectional data will help you derive the impacts of independent variables on the dependent variable(s) in one time period (in one year, one month, one week, one month, or one day) across units of observations. (ii) A panel data set (also longitudinal data), on the other hand, maintains observations for cross-sectional data over several different time-periods.

A simple answer to your question, since you have both cross-sectional data and their panels over several periods is, therefore, as follows:

Otherwise, I agree with many of the comments you have already received😊

The above question has already been answered. Now, I have a question out of curiosity, and the question is:

If you have a panel and cross-sectional but unbalanced data, then what models will be applied?

Xtreme GLS with homogeneity and heterogeneity, I guess. Is this a right answer?