Page 1 of 1

estimation of data collected in separated periods(e.g.after 4 years)

Posted: Thu Apr 05, 2018 8:41 am
by mgorette
Hi,

Is there any way you can estimate annual data from surveys done every 4 year.
I want to see the impact of time use(on income generating activities, on rest,...) data on GDP. This means that I will be using time series of average number of hours spent on different kind of activities, but available data are collected every 4 years and up to now, only 4 surveys which contain data I need were done, though I want to estimate using yearly data and 4 observations are not enough. I wanted to know if I estimate time use data on years that are between years of surveys(e.g. 1st survey done in 2000, 2nd done in 2005.... and I also need data of 2001,2002,...).

What is the best method to use while estimating using this kind of data?

Here is proposed model specification: Log GDP per capita= α+ β1X1 + β2X2+ β3X3+ β4X4+ β5X5+ Ut (Xs being average number of hours spent on different kind of activities)

Thanks

Re: estimation of data collected in separated periods(e.g.after 4 years)

Posted: Thu Apr 05, 2018 8:57 am
by startz
Sometimes people interpolate between observed points. But in this case you only have four real observations, so you are pretty much out of luck.