Learning: Exam 4

Choice

• the distribution of behavior among alternative sources of reinforcement

Preference

• when several schedules of reinforcement are available concurrently, one alternative may be chosen more frequently than others

How is choice studied in the lab?

•  setting up 2 schedules of reinforcement that are concurrent

Concurrent

• available at the same time

• Schedules function independently of each other

Problem with concurrent schedule

• Indiscriminate switching

  • Switching may be reinforced accidentally 

How to prevent indiscriminate switching

• Change Over Delay (COD)

Change Over Delay (COD) as a procedure

• control procedure to stop rapid switching between concurrent alternatives

Change Over Delay… what is required

• After switching to a new alternative, a brief time is required before a response can be reinforced

Which type of schedule is preferred for concurrent schedules

variable interval

Experimental Procedures to Study Choice

1. Arrange 2 or more concurrently available schedules of reinforcement

2. Program interval schedules on each alternative

3. Use variable rather than fixed-interval schedules

   1. Organism unable to discriminate time to reinforcement

   2. Must respond on both alternatives - switching doesn’t always result in reinforcement

   3. Operant behavior sensitive to the rate of reinforcement on each alternative

4. Require a COD in order to stop frequent alternation between or among schedules

Matching Law

• the proportion of behavior allocated across alternatives will equal (match) the proportion of reinforcement available across those alternatives

Hernstein’s Equation in words

• Relative rate of behavior on key A equals relative rate of reinforcement from key A… same for key B

Hernstein’s Equation (actual equation)

IV: relative rate of reinforcement

DV: relative rate of response

Hernsteins equation example

• Only have to calculate one side of the equation because they are EQUAL 

• Alternative 1 = 20 reinforcers/hour (calculated below)

• Alternative 2 = 10 reinforcers/hour

• The proportion of alternative 1 is for reinforcement so 67% of behavior should be allocated to alternative 1

  • And 33% of behavior will be allocated to alternative 2

In order to calculate matching law you need the

rate of reinforcment

example of converting to rate of reinforcement

• Lever 1 = VI 15s food

• Lever 2 = VI 30s food

15s food means the animal is getting 4 pellets a minute and 30s food means the animal is getting 2 pellets a minute, so…

• Lever 1 = 4 pellets/minute

• Lever 2 = 2 pellets/minute

Perfect matching occurs when…

• On concurrent VI VI with COD of a few seconds

• When target behavior is same on both schedules

• Same effort is required on both schedules

• Individual has no spontaneous preference for either schedule

Departures from Matching

• bias

• undermatching

• overmatching

bias

• unknown asymmetry

  • preference for one alternative

undermatching

• changes in response ratio are less than changes in reinforcement ratio

  • response proportions less sensitive to changes in reinforcement proportions

overmatching

• changes in response ratio are greater than changes in reinforcement ratio

  • response proportions are more extreme than reinforcement proportions

extensions of matching law

• time: matching time on an alternative

• alternative: matching on more than 2 alternatives

• social: matching in a social situation

Significance of matching

• we are constantly presented with choices

  • matching allows us to predict that outcome of those choices

choices for smaller sooner reinforcer over the larger more delayed reinforcer

choices for larger more delayed reinforcer over the smaller sooner reinforcer