Table 1.
Behavioral results
| Pain | Aversive tone | |||
|---|---|---|---|---|
| Valence | Contingencya | Valence | Contingencya | |
| Δ(ΔCS+pain, ΔCS−) (95% CI) | CS+pain (95% CI) | Δ(ΔCS+tone, ΔCS−) (95%CI) | CS+tone (95% CI) | |
| Discovery study (n = 25) | 25 (16–34) | 56 (40–70) | 15 (7–25) | 22 (1–42) |
| Validation study (n = 49) | ||||
| Sub-sample 1 (n = 26) | 21 (12–31) | 48 (29–65) | 15 (4–25) | 19 (1–37) |
| Sub-sample 2 (n = 23) | 29 (18–41) | 66 (50–80) | 27 (16–38) | 70 (55–84) |
Results show evidence for threat learning in both the pain and the unpleasant tone conditions, in all three studies. Values represent the mean and bootstrapped confidence intervals value in parenthesis. ΔCS+pain – valence change between the end of acquisition and habituation for the conditioned stimulus associated with pain, ΔCS+tone – valence change between the end of acquisition and habituation for the conditioned stimulus associated with aversive tone, ΔCS− – valence change between the end of acquisition and habituation for the conditioned stimulus associated with no aversive stimulus. See Eq. 1 for more details. CS+pain – contingency ratings of the conditioned stimulus associated with pain, CS+tone – contingency ratings of the conditioned stimulus associated with aversive tone. A similar table with all the participants (without exclusion) can be found in the Supplementary Material (Supplementary Table 4).
aContingency ratings were measured with different scales in the discovery and validation study 1 as compared to validation study 2 (see the section “Methods”, subsection “Behavioral paradigm”).