💙Code available github.com/luiantav/vp_... & github.com/luiantav/vp_...

8/8

Finally, we show that trait anxiety is associated with stronger generalisation and that this overgeneralization is associated with value rather than perceptual mechanisms.
7/n

Using computational modelling we show that we can dissociate perceptual and value-based generalisation. Response distributions (here for generalisation stimuli closest to the CS+) reflect predicted distributions.
6/n

We predict that trial-wise rating distributions differ for perceptual & value-based generalisation (binary vs. gradual). To generalise value, participants can draw on different functions (Gaussian-like, monotonic).
5/n

We show that outcome expectancy ratings learned for a CS+ generalise to perceptually similar stimuli - although showing different patterns.
4/n

👻Using a pretty task involving dangerous (&scary) Spaceflowers we collect data on participants’ perceptual discriminability, aversive learning and generalisation behaviour.
3/n

Generalisation of learned outcome expectancies to similar stimuli can result from confusing two stimuli (perceptual mechanism) or from the transfer of learned value (value mechanism).

We use a modelling approach to dissociate these two mechanisms.
2/n