Table 1 PSIS-LOO scores for the whole sample.

From: Self-beneficial belief updating as a coping mechanism for stress-induced negative affect

Model

PSIS-LOO

LOO-SE

LOO-Diff

(SE-Diff)

% of \(\widehat{k}\) > 0.7

No. est. parameters

Unity model (M1)

 − 2028.5

257.0

267.1 (52.0)

0.09

3

Ability model (M2)

 − 1884.4

247.4

123.0 (95.9)

0.53

4

Valence model (M3)

 − 1761.4

280.4

 

0.17

4

Mean model (M0)

 − 2531.9

219.2

770.5 (93.5)

0.00

2

  1. LOO = sum PSIS-LOO, approximate leave-one-out cross-validation (LOO) using Pareto-smoothed importance sampling (PSIS); LOO-SE = Standard error of PSIS-LOO; LOO-Diff (SE-Diff) = Difference in expected predictive accuracy (PSIS-LOO) for all models from the model with the highest PSIS-LOO (Valence Model) and standard errors of differences; percentage of \(\widehat{k}\)—estimated shape parameters of the generalized Pareto distribution—exceeding 0.7 (all according to Vehtari et al.60); No. Est. Parameters = number of estimated parameters in the model.
Back to article page