I have data consisting of two trends over time.
For example, the GDP and net immigration figures of the UK, from 1950-2020 (not the real data).
I can plot the 2 trend lines on the same graph and visually examine if they look like they correlate over time. However, is there any way to examine if they statistically correlate over time?
Note, this need to be a simple analysis that one can do in MS excel of other free software (e.g. JASP), as it is for a student.
Perhaps, a sliding-window correlation can be plotted (added) to everything that you described.
E.g., for 70 years quarterly data, i.e., for 280 data points, you can plot 80-dataPoints sliding window for correlation.
... or, just have a look at previous years similar discussions, like from 2016:
https://www.researchgate.net/post/How-can-I-analyze-the-correlation-between-two-variables-over-time
Thank you Steftcho P. Dokov
I think that would do - just to clarify is a 'sliding window correlation' the same as a 'rolling correlation'?
I found this useful tutorial - https://www.statology.org/rolling-correlation-excel/
Yes, the same meaning: "rolling the window" = like = sliding the window.
Good luck Joseph Whittaker.
- Badges
- Science topic
- Similar topics
- Computer Communications (Networks)
- Simulators