Hi, I would like to test if daily temperatures in May differ between three years (2014, 2015 and 2022)?
I was thinking of using a Kruskall Wallis test but I am not sure if my data is independent since I use data from one station. The idea is that I would like to compare weather conditions between the three years for a certain region.
Statistical independence states that the probability of one should not impact the probability of another. When it comes to weather it is difficult since it is cyclic, however, the temperature of May 1st in 2014 should not impact the temperature of May 1st in 2015, above the general assumption that the temperature in May between years should be roughly more similar than between May and June. Therefore, I would argue that the independence is preserved in your particular case.
Tess Blok,
Apart from the insightful comment of Carl Alexander Frisk, you may want to take a look at the article below.
Article A Note on the Flowering of Ajuga orientalis L. in Relation t...