Follow

How does one find out the interaction data associated with toots? E.g. likes, boosts, no of comments.

I have a strong suspicion that the time of posting significantly affects visibility and therefore interaction, and I'd like to do some statistical analysis!

@douginamug I cannot answer your question, but I can say from experience that (at least in my Fedi garden) users are more active Monday-Thursday: those would be the best days for posting. On weekends the timeline is much slower, and Friday's too close to weekends and seems to be among the days when people have a rest from the interwebz.

@douginamug just don't go around hoarding people's toots without consent to perform your analysis. Don't.

@steko although I think I agree with you, I feel your tone is a bit preachy? Maybe it's just my mood.

In any case, what are the ethical concerns of an individual download toots to perform local analysis (or not downloading and analyzing 'in situ') so long as no personal details of others are published?

What harm is there in publishing anonymized stats?

@douginamug
I've a bit of trouble understanding the problem, but maybe it's just me...
Isn't it possible to access only public posts for the analysis? If so, then I must say that if only date and boost/fav are kept, I can't see any issue.
Anyway, people may have become more sensitive due to a more than borderline academic study about Fediverse not long ago: mastodon.nzoss.nz/users/strype
But I think we should be able to differenciate between ethical and unethical practices :)
@steko

@tfardet @douginamug there are several issues.
1) there is no central server, and every instance has its own privacy policy. Being able to access public toots by manual browsing is not the same as mass download.
2) again with no central server, there is not a single endpoint to get the content you're looking for. Not all instances may be happy with mass download of content (bandwidth is not free)
...

@tfardet @douginamug
3) as mentioned (I should have brought it up in the first place) there's been a previous study that was riddled with naive assumptions so any attempt at new studies that doesn't start with an analysis of that failure is doomed yo repeat the same mistakes
4) personally I understand the proposed analysis as "how to maximize the impact of your toots by choosing the right time of the day" that I find at odds with the ethos of Mastodon

@steko @douginamug Thanks for the answers!
I must say I'm not sure I understand point 1... could you elaborate on what difference you would see?
I did not think about point 2 but I'd need to research this more. I must say I'm not 100% convince this would put much additional load on servers, but worth investigating!
3) yes
4) I hope trying to reach out to more people on a theme you're interested in can be understood and distinguished from seeking "ego-boosting" ^^
Plus there's just curiosity

@tfardet @douginamug think of server X having terms of use where mass downloading toots is forbidden. Then server Y, where terms of use require approval from administrators, etc. Not all instances have the same rules and not being explicitly forbidden doesn't mean it's welcome. E.g. there is one instance (can't remember which one) that allows/allowed full text search by indexing toots from other instances and that was not OK.

@steko @douginamug ah, I see what you mean now, interesting... these are configurations where federation must become quite the headache...
I'll think about it! Thanks :)

Sign in to participate in the conversation
Mastodon

The social network of the future: No ads, no corporate surveillance, ethical design, and decentralization! Own your data with Mastodon!