hello, I work on export competitiveness. I have heterogeneous variables (percentages, indices and values), the objective of the research is to explain export competitiveness from these variables in order to make forecasts. from my readings I have noticed that the appropriate model for this issue is the ARDL model, whereas I have only 24 observations; I wonder if I have the right to use the ARDL model
You will likely get poor results. 30+ is the standard. Other researchers asked a similar question in RG (for instance https://www.researchgate.net/post/What_are_the_minimum_number_of_the_observation_that_are_necessary_for_ARDL_model). It seems you are working with macro data on exports. Often these are available on Q or monthly basis. This would be a way through to increase the number of observations, provided you adjust for seasonality beforehand.
kind regards
Andrea
Time Series Regression with Stationary Variables: An Introduction to the ARDL Model
Czar Yobero
1/12/2017
https://rpubs.com/cyobero/ardl
- Badges
- Science topic
- Similar topics
- Biological Science
- Bioecology
- Systems Ecology
- Landscape Ecology
- Connectivity