Token Economies Flashcards
Token Economies
What’s a Token?
-Token:
- an object or symbol that is exchanged for goods or services
- object that can be accumulated
- has no phylogenetic value
- exchanged for phylogenetically significant stimuli
- typically manipulable
- Examples
- coins, poker ships, marbles, lights, points, money, credit, levels
- earning “stars” in the classroom
Can Token Establish and Maintain Behaviour
-Wolfe (1936)
- chimps trained to lift bar system to receive either tokens or food
- increased response requirement
- responding effectively maintained for both types of reinforcers
-Cowles (1937)
- Chimps given discrimination task
- White tokens = food
- Brass tokens = no food
- Both food and token groups learned the task rapidly
- Learning was more rapid for primary reinforcers
Schedules of Token Reinforcement
-Token Production Schedule
- schedule that dictates how tokens are earned
- e.g., 50 responses gives 1 token (FR 50)
- e.g., 50 responses gives 1 token (FR 50)
- Exchange Production Schedule
- schedule that dictates when exchange opportunities are made available
- e.g., 2 tokens needed to allow exchange (FR 2)
-Token Exchange Schedule
- schedule that dictates how primary/terminal reinforcement is delivered
- also called the “Exchange Rate”
- e.g., token provides 1 reinforcer (FR 1)
Other Notes about Tokens
- can be brought under discriminative (stimulus) control
- e.g., casino chips only works in a particular casino
- Reinforcing value can be extinguished
- The accumulation of tokens can function as a discriminative stimulus
- the more tokens earned, the closer you are to exchanging them for the reinforcer. The accumulated tokens function as an S^D stimulus that generates increased responding
- depends on the schedules being used
- Can be used to bridge long delays
-Pairing token and food can establish CS functions
- chimps reminding brass tokens (no food paring) and white (food-paired) exhibited consummatory behaviours (i.e., treated the tokens as food), This has been seen with poker chips, marbles, and ball bearings. And in other species such as rats
- Removal of accumulates tokens can have a negative punishment function of behaviour
- Reinforcing value can be manipulates by altering deprivation level of the primary reinforcer
Generalized Reinforcers
- conditional reinforcers that are contingent on multiple forms of unconditional reinforcement
- money
- makes reinforcers less dependent on effects of depreciation and satiation
- Laboratory Research is lacking
- unknowns:
- what types of experiences are necessary for their formation
- are they more resistant to extinction
- are they preferred to specific reinforcers
- is losing a generalized reformer more punishing that losing a specific reinforcer
- will subjects work harder to obtain a generalized reinforcer over a specific one
Essential Features of a Token Economy
Identify and define target behaviours
Identify desirable behaviours to increase
May also identify undesirable behaviours to decrease
Identify the items to use as tokens
Tokens must be easy to deliver and accumulate
Tokens should not be
Available outside the economy
Easily stole or counterfeited
Identify and procure terminal (“back-up”) reinforcers
Terminal reinforcer must be known reinforcer for the participants
Best to have variety
Might be physical, social, or other reinforcers (e.g., privileges, services)
Decide on a schedule of reinforcement for the deliver of tokens
I.e. heat is the Token Production Schedule
How many tokens delivers for each target behaviour
Decide on the token exchange rate
I.e., What is the token exchange rate
How many tokens are requires to obtain a particular reinforcer
Cost should be low to begin with
As token-behaviours increase, cost of a terminal (back-up) reinforcer can be increased
Determine the time and place for exchanging tokens
I.e., what is the Exchange Production Schedule
What de Teri when tokens can be exchanged for terminal reinforcement
Will help avoid non-discriminated extinction effects
Consider the use of a “token store”
Response Cost (Negative Punishment)
Define undesirable behaviours that compete with desirable behaviours
Decide how many tokens are lost for each behaviour
-Response Cost tips
- do not try to punish the “non-occurrence” of a behaviour by removing tokens
- the non-occurrence of a behaviour is NOT a behaviour and therefore is not a behaviour that can be punished
- Punish undesirable behaviours that compete with desirable behaviours
- to be effective, the tokens need to actually be conditional reinforcers
- has been demonstrated? DO they have a reinforcing function?
Other Considerations
- Perform a baseline “field test” of the economy
- Present tokens immediately after desirable behaviour while praising
- remove tokens immediately after undesirable behaviour
Long-Term Considerations
- Over the long-term
- incorporate techniques to foster generalization
- ensure that target behaviours have natural reinforcer associates with them
e. g., Omari token delivery with. Descriptive praise and social approval - Gradually increase the number of responses required to earn a token
- make use of variable schedules
- Incorporate items and activities from the -real-world (criterion) environment as back-up reinforcers
- increase the cost of highly desirable items that are irrelevant to the criterion environment (e.g., toys)
- replace items irrelevant to the criterion environment (e.g., toys) with relevant items (e.g., award sheets)
- keep the cost of items contextually relevant to the criterion environment low
- Gradually fade away physically evidence of the tokens
- Tokens = poker chips
- poker chips are replaced with slips of paper
- paper is replaced with tally marks on a card kept by learner
- Marks replaced by by index cards kept by instructor
- instructor keeps track of tallies but checking is not allowed and totals are announced at the end of day
- token system is no longer running (however, points totals are still kept evaluating the behaviour)
