Dependance between categorical variables in a whole population


I have a population of 507 documents (which represent the entire phenomena studied at a give time period) from which a series of variables -most of which are nominal- where extracted. The end goal was to describe said phenomena and establish the possibility that the outcome is influenced by the factors that caused the phenomena to happen.

The initial part of describing was effectively achieved with uni variate analysis with distribution and central tendency measures. The second part is where is where i began having trouble... I do understand that significance is not applicable as there is no sample and no extrapolation to a population, also, even though this data is only for now it does not make much sense to conceive a larger population through time or assume it is a process as I understand is a valid option (as I've read in earlier posts).

So in the end what I've done is begin by crosstabulating the variables that I believe are related and then running a Chi square (through SPSS), also to complement this I obtained the values for Kruskall and Goodman tau, uncertainty coefficient and Cramers V. To interpret these I did not pay attention to the significance but only the calculated value, using it to simply state that there is dependance or independence.

Being aware that I'm not extrapolating to a population simply establishing the existence of a relationship (not even specifying directionality), is it valid to use the above mentioned tests?

As it is my first time posting, after reading quite a few, please excuse any mistakes and thank you for considering this and your time,