An early-warning signal is a within-system change in dynamics (indicative of critical fluctuations and/or critical slowing down) that is predictive of a phase transition.
The system has rising instability which precedes a Tipping point.
- Stable state -> rising instability (increased variability) -> re-stabilization in a new attractor state
Early warning signs can only be studied within-person
We can’t do between-individual things but need within-individual measurements. Typically intensive time-series data. The system dynamics can vary between individuals.
Early warning signs should precede phase transitions
Can phase transitions be defined and predicted for the individual?
There should be periods of relative stability
In the figure above:
- Before transition: EWS increase
- After transition: EWS decrease
How long are the time periods? We have lots of timescales in psychology, from developmental time (onset of psychopathology) to real-time (panic attack) with something like sudden gains is somewhere in between.