2
$\begingroup$

Context: Zoology undergrad dissertation. The format of my results will be % time individual birds spent doing certain activities (and possibly number of times an activity such as pecking was performed). There is no external way to tell whether the individual bird is male or female, so my hypothesis is that males and females will have different activity budgets. Neither myself and my dissertation supervisor know what statistical test to perform - what test can be used for this?

$\endgroup$
  • $\begingroup$ Can you edit your answer to clarify exactly what data you have? e.g. discrete observations at fixed time intervals, or continuous observations, or something else? $\endgroup$ – R Greg Stacey Mar 27 at 17:02
  • $\begingroup$ The data is continuous observations, sorry I didn't include that! $\endgroup$ – Oliver Mar 29 at 13:04
  • $\begingroup$ It will help if you edit your answer to include that! The more specific you can be about your data the better. Also, Bjorn's suggestion of clustering sounds good to me. If you don't know which birds are male and female, I don't see how you can test if males and females have different behaviours! So the other option is to cluster your data (in some way) and see if you get two clear groups. You might be able to infer the two groups are male and female. If you can get a male/female label for any of the data, though, that would be much better! $\endgroup$ – R Greg Stacey Mar 29 at 18:21
1
$\begingroup$

You could try clustering the behaviour patterns in an unsupervised way using e.g. UMAP (available in several softwares: python had umap-learn, R has a package accessing the python package - watch e.g. this YouTube video for a nice explanation or the article for more details) or t-SNE. You can then see whether you get two clusters that are sized in about the right ratio for the ZZ vs. ZW chromosome ratio, characterize what the clusters are like and then speculate whether they might be clusters by ZZ vs. ZW.

Can you get a feather or other DNA sample in some way? Getting a read on that on a at least a few individuals that way would help a lot to see whether that matches up with the theory. I believe DNA tests on feathers are pretty cheap nowadays.

$\endgroup$
  • $\begingroup$ I'll look into that, thank you! I can't get any DNA samples unfortunately. $\endgroup$ – Oliver Mar 29 at 12:59

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.