In economic activities, if we want to estimate the value of certain variables, in a regression equation, Does the data used for regression analysis tend to be in the form of panel data? If the data is in the category, can it be used for the estimation process?
The data used for regression analysis can be in the form of time series data, cross-sectional data or panel data. If the nature of categorical data is simlier to cross-sectional data then you can used those. Further please go through following attached ppt which gives very simple and clear explanation about above three types of data.
thank you very much Dilesha, and
have been willing to send PPT to me,
Best regards,
i putu artaya
Agree with Dilesha. But does the use of regression with economic data "tend" to mean that panel data are used? I don't know if anyone has actually done a study of that, but I would not be surprised either way - i.e., the answer might easily be "yes" or "no."
At many institutions studying economic activity, I expect panel data is the norm. However, a variety of other activities may occur.
For example, regarding official energy statistics from repeated establishment surveys, you might be surprised to know that periodic sampling of finite populations at the US Energy Information Administration, to provide information on energy markets, uses regression for some key surveys to 'predict' (estimate not forecast) for missing (not-in-sample and nonresponse) data. Here is a URL for a presentation on that: https://www.researchgate.net/publication/319914742_Quasi-Cutoff_Sampling_and_the_Classical_Ratio_Estimator_-_Application_to_Establishment_Surveys_for_Official_Statistics_at_the_US_Energy_Information_Administration_-_Historical_Development
The variety of activities involving inference from economic data would also be hard to quantify since some efforts (surveys or studies) are much bigger than others. How do you count them?
Perhaps it would be best to summarize by saying that for economic activities, regression is used with panel data, and in a variety of other ways. Dilesha noted cross-sectional data and time series data. Whether or not panel data are most likely is debatable.
(Dilesha: A pdf file would be nice. My phone won't open a ppt file.)
Marcus Hellwig
using regression equations, with variables:
Y= sales growth
X1= advertising costs
X2 = The company's profit
X3 = Industrial Growth
use common efeect model.
..and this is for next year's research plan..
thank you very much, sir Marcus Hellwig,
Best regards,
Putu Artaya
I agree with Dilesha. Furthermore, i would advice you that if the dependent variable is ordered variable (ordered-category) then the best techniques to be used are Probit and/Logit analysis.
Probit and Logit Models Example
https://www.youtube.com/watch?v=66Mn9GAWrAE
Probit and Logit Models in Stata
https://www.youtube.com/watch?v=wU1DVbpD9SY
mister Yarob Kullab,
thank you very much for the information, very helpful for me..
best regard,
I Putu Artaya
I think your main-question was:
If the data is in the category, can it be used for the estimation process?
Yes, you can .... but! before you should clarify whether they are independent - from each other. If not - use Bayes - Theorem for carefull prediction.
for those:
Y= sales growth
X1= advertising costs
X2 = The company's profit
X3 = Industrial Growth
I recommend a minimumdatset number of 20 values - or even more - for each one as you must know that distributions/ frequencies may have "long tails"
Yours Marcus
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Regression