Time-series Correlation

V

Vertje

New member
Joined
Jun 5, 2012
Messages
1
Hi all,

I was wondering if anyone could help me out with the following problem;

For my study I’m investigating if a change in GDP of a country is correlated to the number of Patents granted in the same country. I’ve collected 27 years data of GDP & granted Patents (data collection per year).

I’ve read about correlation of time-series, but I’m not sure how to handle this problem. Both the GDP and Patent numbers tend to grow over time, so I suspect a correlation simply because of the increasing trend? Is it possible to “filter” out this trend? Or this is not needed at all?

I hope someone can explain this is a simple manner, since I’m not quite that experienced with statistics
icon_rolleyes.gif
. It is possible for me to use Excel or SPSS.

Thank you in advance.
 
Sounds to me like you need to start with:

1) # of Patents

vs

2) 1st Difference in GDP


Quick Example

GDP: 1 3 4 5 7 7
1st Difference in GDP: 3-1 = 2, 4-3 = 1, 5-4 = 1, etc
Patents: 2 4 2 3 5
 
How about looking at the "change in patent granted per year" vs. "change in GDP per year".

I suppose TK was hinting at similar "rate" correlation.
 
You must log in or register to reply here.
Top