[anbende]

In general, you're right of course.

The article in question used a lagged design, using temporally prior markers to predict later change. This is not fully causal, but it's not fully correlative either. In the field, it's consider to provide preliminary causal evidence. Still needs additional research to be sure.

[JamesBarney]

It's better than straight correlation, but given the historical track records of these studies turning into a reproducible intervention of clinical significance I'd say the chances are low. The article event mentions this with

> Unlike in previous research, Zainal and Newman did not find that self-mastery, or feeling in control of one’s fate, had an effect on the mental health of participants across the 18-year period.