4.1 Value Equation

Question 1

Temporal motivation theory

Answer 1

Integrates four theories: picoeconomics, expectancy theory, cumulative prospect theory, need theory

A (mathematical) model of motivation and decision making.
Temporal means time is important
Balances between precision and parsimony (simplicity)

V (value) = (R * pS)/(D * Imp + 1)
R = reward (things we like)
pS = probability of success (how likely we are to get reward
D = delay (time until we get reward)
Imp = impulsiveness (importance of delay)

We make choices based on what has the most value to us, but the value is affected by each parameter

Question 2

Reward (R)

Answer 2

Can be positive (benefit) or negative (cost)
The value is relative to other values
Also subjective from person to person

Right now, V = R
ex grad = 1000+, studying = 10-

Question 3

Probability of Success (pS)

Answer 3

It should matter how certain or uncertain we believe the reward to be!

From 0.0 to 1.0, as it’s a probability
Also subjective from person to person

Right now, V = R*pS
A reward we know for sure the we won’t get has a value of 0
A reward we know for sure we will get has full R value
When R is negative, low pS is GOOD

THINK
What is the range of V for : R > 0?
What is the range of V for : R < 0?

Question 4

Delay (D)

Answer 4

It should matter how close or distant in time we believe the reward to be

Time: 0 or more
Subjective
More delay -> more temporal discounting (+ve reward decreases and -ve reward increases as the delay increases)
Value converges to 0 as delay approaches infinity

Right now, V = (R*pS)/D + 1
We add the +1 to remove divide by 0 error

THINK
What is the range of V for : R > 0?
What is the range of V for : R < 0?

Question 5

Impulsiveness (Imp)

Answer 5

The weight we give to delays (our relation to time and reward)
How much temporal discounting is actually happening in your mind
More or less fixed for a person or animal
More impulsiveness = more temporal discounting

Now we get the original equation

THINK
What is the range of V for : R > 0?
What is the range of V for : R < 0?

Question 6

+1

Answer 6

To not divide by 0

We get the full reward when D reaches 0

Question 7

Value curves

Answer 7

Review curves
\+ve rewards
R = +100
pS = 1
Delay from large to small as that's how time moves forward

-ve rewards
R = -100
pS = 1

Question 8

Making a Choice

Answer 8

-3 tiered system when making choices and actions
-A choice has multiple actions
Each action can have multiple rewards (for project example: turning it in, graduating, having to do work)
-Each reward has a set of parameters and a value

The total value of an action is adding all the rewards together

The action chosen for the choice is the one with the highest total value