We study a technique that allows an experimentalist to quantify separately the degree to which an observed anomalous diffusion occurs due to:
i. non-stationarity (the ``Moses effect"), ii. extreme rare events (``Noah effect") or iii. temporal correlations (``Joseph effect").
This decomposition method offers a way to obtain a better understanding about the underlying dynamics of the system, without making prior assumptions on the model that describes it.
In addition, it facilitates the classification of fitting and non-fitting models that might be proposed for this purpose.
We apply the approach to data from the recent multi-group collaborative effort, to create an objective comparison of methods to decode anomalous diffusion, known as the ANDI challenge. Our hope is to promote the use of this technique for more systems in the future.