Table 1. Average parameter estimates for cumulative prospect theory obtained in the affect-poor and affect-rich lottery problems and results of significance testing.
Lottery Problems | Significance Test for Differences Between Affect-Rich and Affect-Poor Lottery Problems | ||||
---|---|---|---|---|---|
Parameters | Affect-Poor | Affect-Rich | t(22) | p | |
γ | 0.77 (0.24) | 0.43 (0.33) | 4.88 | <. 001* | |
δ | 0.99 (1.63) | 2.47 (3.10) | 2.56 | = .018** | |
α | 0.73 (0.21) | 0.79 (0.26) | 0.71 | = .480 | |
g | 0.07 (0.11) | 0.12 (0.19) | 1.15 | = .261 | |
G 2 | 40.60 (23.86) | 59.74 (32.81) | 2.07 | = .051 |
Note. Standard deviations are in parentheses. γ and α model the sensitivity to probabilities and outcomes, respectively, with higher values indicating higher sensitivity; δ models the elevation, with higher values indicating higher risk aversion; g indicates the probability of random guessing. The G 2 expected under chance is 116.45.
* Significant tests after adopting a Bonferroni-Holm correction [50]. With m = 5 tests, the observed p values are first ordered in ascending order and are then tested with α1 = 0.05/m, α2 = 0.05/(m−1), …, αj = 0.05/(m−(j−1)).
** One-tailed.