TECT: Kacelnik Package Individual and group decision making under risk. Are groups more or less...

Post on 20-Dec-2015

214 views 0 download

Tags:

Transcript of TECT: Kacelnik Package Individual and group decision making under risk. Are groups more or less...

TECT: Kacelnik Package

Individual and group decision making under risk.

Are groups more or less efficient in handling risky decisions than individuals?

The Experimental Paradigm:

Risk it!

Choice Play Safe

Discrimination images

Confidence?RIRisk it?

Well done…You won 1…

OK

FeedbackSample

10%

20%

30%

90%

100%

Decisions and contingencies

Risk it! Play Safe

Correct identificationP

HIT Missed opportunity

Incorrect identification1-P

False Alarm

Correct rejection

Payoff P * (HIT) + (1-P) * (FA) P * (Miss) + (1-P) * (CR)

This matrix allows us to combine Signal Detection Theory and Metacognition, and to identify absolute, individual and group optimal strategies

CRMFAH

FACRP

Maximum expected payoff:Take a chance IFF:

Hit

False AlarmCorrect

Rejection

Miss

Our experimental paradigm results in estimates of:

1. Accuracy of metacognitive estimates.

2. Influence of metacognition on choice.

3. Non- conscious determinants of choice.

4. Whether groups differ (respect to individuals):

1. Accuracy of choice

2. Accuracy of self-appraisal

3. Risk appetite

4. Achieved pay off

Chance Safe

CorrectIncorrect

Miss

CRFA

Hit

Logistic regression of Outcome vs. Confidence, gives several parameters of self perception

0

0.5

1

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

Confidence

P C

orr

ect

Cor

rect

?

Logistic regression of Decision Versus Confidence gives parameters of actual riskiness in behaviour

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

Confidence

P G

am

ble

Too

k a

cha

nce

?

C*: Optimal Confidence Threshold

A perfectly consistent individual has one estimate of Pcorrect; below it, she plays safe, above it, she chances.

C*:the confidence above which a perfectly consistent individual should switch from playing safe to gambling in order to maximize payoff

C* results from a subject’s distribution of accuracy and confidence, but assumes total consistency, hence it is another theoretical yardstick against which to judge decision making aptitude.

270

290

310

330

350

370

390

410

430

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

Confidence Threshold

Score

C*

God’s Dad

Normal discriminationPerfect judgementPerfect consistency

C*

Normal discriminationNormal judgementNormal consistencyExcellent confidence match