EViews.com

Dear All,

I want to estimate a panel regression model using the World Values Survey.
The WVS is collected in 5 yearly waves so the data which is available (which I am using) is from 2005-2009 and 2010-2014.
My Independent variable are GDP, Trade Openness and Gross Capital formation which are available yearly. I have a dataset for 64 countries.

How should I manage the different frequency of data i.e. yearly independent data and the five yearly wave data?
How should I input the five yearly wave data against the annual data for my panel regression?


Thank you.

Dear mbelgie,

Could I see these data? I do not know what is the World Values Survey. Maybe I can help you.

Best Regards, Jmagomez.

thank you for your response.

I am posting a link to WVS website.

http://www.worldvaluessurvey.org/WVSOnline.jsp

From this survey, I have used data on certain questions.

Dear mbelgie,

I think you should transform the wave data into annual data. The problem is: how could the interpolation be done?

You may use more waves to make a better interpolation and analysis.

Best Regards.

Dear Jmagomez,

Thank you for your response.

Do you think I can use a mean value for the years for the wave data? Could that be a way of managing interpolation?

Dear mbelgie,

Yes, I believe you would be taking a good technique to adapt the data available. But as I said, I think you should use a longer period than two waves.

Regards.

Dear jmagomez,

Thank you for the reply. I have used the same value of the WVS for the rest of the reported wave years. So if it says 25 % for 1999 to 2004 wave, I have used the 25% for all these years.

Now I am trying to build the model.

Thank you again for the help.