Table 2 Cognitive modelling results showing the proportions of participants that relied on each learning mechanism in Experiments 2–4

From: Learning from outcomes shapes reliance on moral rules versus cost–benefit reasoning

Learning mechanism

Experiment 2

Experiment 3

Experiment 4

CBR

Rule

CBR

Rule

CBR

Rule

\({\mathbb{E}}({\boldsymbol{f}}| {\boldsymbol{Y}}\;)\;(\boldsymbol{ \%})\)

φ

\({\mathbb{E}}({\boldsymbol{f}}| {\boldsymbol{Y}}\;)\;(\boldsymbol{ \%})\)

φ

\({\mathbb{E}}({\boldsymbol{f}}| {\boldsymbol{Y}}\;)\;(\boldsymbol{ \%})\)

φ

\({\mathbb{E}}({\boldsymbol{f}}| {\boldsymbol{Y}}\;)\;(\boldsymbol{ \%})\)

φ

\({\mathbb{E}}({\boldsymbol{f}}| {\boldsymbol{Y}}\;)\;(\boldsymbol{ \%})\)

φ

\({\mathbb{E}}({\boldsymbol{f}}| {\boldsymbol{Y}}\;)\;(\boldsymbol{ \%})\)

φ

Metacognitive learning

89.4

1

27.9

0.05

92.5

1

35.8

0.06

85.5

1

41.9

0.26

Behavioural learning

5.2

0

67.3

0.95

5

0

60.8

0.94

11.1

0

53.5

0.74

No learning

5.4

0

4.8

0

2.5

0

3.4

0

3.3

0

4.6

0

  1. ‘CBR’ denotes the CBR Success condition, and ‘Rule’ denotes the Rule Success condition. The exceedance probability φ of a given model family is the probability that the proportion of participants best explained by a model from that family is greater than for any of the alternative model families. The largest proportions in each condition are highlighted in bold.