The Cycles Institute has published an article on Bartels Test for cycle significance. The article refers to data from an even sample space (12 segments of 41 months each). Bartels Test of Cycle Significance
For my challenge, I am using the Lomb-Scargle method to determine spectrum based on periodic data with missing data points (uneven sampling). I would like to perform a Bartels-like test to ensure underlying cycles are in fact "statistically significant".
As I am dealing with an periodic data with gaps, should I used a fixed period (such as 41 months) and use a varying number of samples within each segment or should I just use 41 consecutive samples?
Alternatively - is there a better approach available? There is limited data available on Bartels, however there are multiple claims this is still the best method available (but I am unsure about suitability of uneven sampling)
It has been 35 years since I learned maths - but I am enjoying the re-acquainting knowledge. I would really appreciate it if someone could help me make sense of this and point me in the right direction.
Thank you in advance.