Table 1.
Variable name | Definition | |
---|---|---|
Variables described in main text | ||
1 | offer max value A | QA |
2 | offer value A | [pA]∧α * QA |
3 | offer risk A | sqrt[pA * (1 − pA)] * QA |
4 | offer max value B | QB |
5 | offer value B | [pB]∧α * QB |
6 | offer risk B | sqrt[pB * (1 − pB)] * QB |
7 | binary choice | 1 if A chosen, 0 if B chosen |
8 | weighted choice A | pA if A chosen, 0 if B chosen |
9 | weighted choice B | 0 if A chosen, pB if B chosen |
10 | risky choice | 1 if p of chosen offer is <1, 0 otherwise |
11 | chosen max value | offer max value of chosen offer |
12 | chosen value | ρ * offer value A if A chosen, offer value B if B chosen |
13 | chosen risk | risk of chosen offer |
14 | got juice | 1 if juice delivered, 0 if poor luck trial |
15 | taste A | 1 if A chosen and delivered, 0 otherwise |
16 | taste B | 1 if B chosen and delivered, 0 otherwise |
17 | received value | chosen max value * got juice |
18 | received value A | QA if A chosen and delivered, 0 otherwise |
19 | received value B | QB if B chosen and delivered, 0 otherwise |
20 | win bet | risky choice * got juice |
21 | risk outcome | chosen risk * got juice |
Other variablesb | ||
22 | offer value A (EV) | pA * QA |
23 | offer value A (MA) | [pA]∧mean(α) * QA |
24 | offer value A (dU) | pA * [QA]∧(1/α) |
25 | offer value B (EV) | pB * QB |
26 | offer value B (MA) | [pB]∧mean(α) * QB |
27 | offer value B (dU) | pB * [QB]∧(1/α) |
28 | chosen value (EV) | ρ * offer value A (EV) if A chosen, offer value B (EV) if B chosen |
29 | chosen value (MA) | ρ * offer value A (MA) if A chosen, offer value B (MA) if B chosen |
30 | chosen value (dU) | ρ∧(1/α) * offer value A (dU) if A chosen, offer value B (dU) if B chosen |
aIn any given trial, QA and pA indicate the quantity and probability associated with juice A, respectively. QB and pB indicate the quantity and probability associated with juice B, respectively. The two parameters ρ and α are derived from the behavioral analysis and represent, respectively, the relative value of the two juices and the risk attitude of the animal in that session. The variable selection analysis described in the main text included variables 1–21.
bVariables 22–30 were considered the analyses performed to establish whether value-encoding responses reflected the subjective risk attitude (see Material and Methods). In those analyses, variables offer value A, offer value B and chosen value are referred to as offer value A (dP), offer value B (dP) and chosen value (dP), respectively.