Table 2.
Mixed effects exponential growth curve model, predicting rated pain from number of prior stimuli.
| Fixed effects | Estimate | SE | p-value |
|---|---|---|---|
| Decline (g) | −0.51 | 0.04 | 0.000 |
| Alarm response (r) | |||
| Naïvehi | 15.18 | 0.97 | 0.000 |
| Naïvelo | 16.62 | 1.03 | 0.000 |
| Sleep | 6.86 | 1.65 | 0.000 |
| Awake | −1.38 | 1.11 | 0.216 |
| Asymptote (a) | |||
| Naïvehi | 22.88 | 0.46 | 0.000 |
| Naïvelo | 11.46 | 0.50 | 0.000 |
| Sleep | 13.08 | 0.70 | 0.000 |
| Awake | 10.89 | 0.48 | 0.000 |
| Random effects | SD | Correlation | |
| Asymptote (η) | 13.04 | ||
| Alarm response (ξ) | 16.27 | −0.48 | |
| Residual (ε) | 5.37 | ||
Note. The model is based on equation 1, but with fixed effects of the asymptote (a) and alarm response (r) calculated separately for each group (Naïvehi (n = 32), Naïvelo (n = 28), Sleep (n = 11) and Awake (n = 24). Random effects describe the standard deviation of subject variation around the asymptote (η) and the alarm response (ξ) in addition to the residual (ε). Overall there is a significant alarm response in naïvelo (p < 0,001), naïve high (p < 0,001) and sleep condition (p < 0.001). No significant pain alarm response in the awake condition (p = 0.216).