Ch 7 - Schedules of Reinforcement

Schedule of Reinforcement

response requirement that must be met to gain a reinforcer

• varying either the time b/w reinforcements or amount of responses needed

Continuous Reinforcement Schedules

every time a response occurs, it’s reinforced

• useful experimentally, X always realistic

Intermittent Reinforcement Schedules

only deliver a reinforcer occasionally, set # of responses, period, etc

• ratio schedules

• interval schedules

Ratio Schedules

change the amount of responses needed

• type of intermittent reinforcement schedules

• how many times rat has to press the lever (ex → 10x to get food)

Interval Schedules

change the minimum time b/w reinforcers

• type of intermittent reinforcement schedules

• press as many times, must have a 30 sec gap in b/w, end = food

Fixed Ratio Schedules

require a set number of responses per reinforcer, do things a certain amount of times

• ratio needs to be gradually inc w/ training

• rich dense = FR5, reward often

• lean = FR 100, reward rarely

• subject to ratio strain

• start w/ continuous reinforcer (FR1/CRF)

Ratio Strain

occurs w/ fixed ratio schedules

• failure to respond caused by ratio rising too quickly/too high

• X start at FR100, fail to respond, too much for them to respond to

• start @ dense → gradually work to lean

Variable Ratio Schedules

reinforce randomly on an avg # of responses

• modifying amount needed to make behavior

• addictive, gambling, small rewards along way → think big one is coming

• VR5 equally likely to reward after 1 response or 10

• reward is always the same

What do variable ratio schedules result in?

very aggressive response rate w/ few pauses

• predatory or maladaptive behaviors rely on VR schedules

What do fixed ratio schedules result in?

rapid responses, brief pause after reinforcement (animal satisfied w/ reward)

Fixed Interval Schedules

deliver a reinforcer on the 1st response after a set time

• long pause after reinforcement

• animal knows it can respond around this time

• ex → Na channels close for certain period of time

What do fixed interval schedules result in?

response rate increases as interval time comes up

• scalloped appearance for response rate

Variable Interval Schedules

produce steady & predictable responses at slow to moderate rate

• X pause, slower than VR

• random amount of time passes then can do again

• most moderate & boring

• response only works when variable takes place

• ex → grading, text & waiting for reply, getting affection in abusive relationship

Pauses in FI & FR Schedules

common but may disappear at very rich reinforcement rates

Appetitive Value of Reinforcer

matters for rate of response, as it determines how motivating the reinforcer is for the subject. A higher appetitive value generally results in increased response rates.

• FR50 w/ food will be moderate for a satiated rat but very strong for hungry rat

Ratio/Interval Modification

in most cases, subjects need to be progressively steppe up to higher ratios or intervals or behavior collapses

Duration Schedules

specify the length of time a behavior must occur for reinforcement to be delivered, often used in conjunction with other schedules

• difficult, X discriminate quality of behavior

• X just practice → need to learn/have good quality of thing to be learned

Fixed Duration Schedules

reinforce based on time spent doing behavior

• having org do behavior for a set peiod of time

• good for behaviors like running, working, etc

Variable Duration Schedules

reinforce based on a random time spent w/ set avg

• based around a random time, doing it for that time

Response Rate Schedules

DRH, DRL, DRP, noncontingent

• organisms want to find patterns in things & sometimes adopt superstitious behavior

Differential Reinforcement of High Rates (DRH)

reinforces for meeting or exceeding a mi rate of reinforcement

• type of response-rate schedule

• ex → reward kids for being quiet

Differential Reinforcement of Low Rates (DRL)

reinforces for staying below a max rate of reinforcement

• type of response-rate schedule

• ex → Reinforcing a child for asking "are we there yet?" only if they wait 10 minutes between questions, or reducing hand-raising/speaking out to a manageable level

Differential Reinforcement of Paced Responding (DRP)

reinforces for meeting a specific rate of reinforcement

• type of response-rate schedule

• get best response for # done

• reward inc/dec rate

Noncontingent Schedules

reinforcer is delivered w/o regard for response

• type of response-rate schedule

• fixed or variable time

Complex Schedules

conjuctive, adjustive, chained schedules

Conjunctive Schedule

rule of 2 basic schedules apply simultaneously to obtain reinforcement.

• type of complex schedule

• It requires meeting the criteria of both schedules before reinforcement is delivered.

• ex → attend class regularly (FI) & complete coursework (FR)

• can be slow (X test week) or crazy (finals)

Adjusting Schedules

altered based on performance on previous responses or reinforcements, allowing for dynamic changes in the requirements to obtain reinforcement

• type of complex schedule

• failure to respond might result in a more rich schedule, while responding well might result in a more lean schedule

• FR changes w/ response rate affecting the criteria for reinforcement.

Chained Schedules

require responding to multiple tasks in a specific order

• type of complex schedule

• 1 thing completed before other

• ex → log in, put in password, multifactor, open email

Training Chained Schedules

involves teaching multiple tasks that must be completed in sequence to earn reinforcement. This approach often utilizes small steps to ensure that each component of the chain is learned effectively.

• earlier schedules (FR10 & FI1m) have weaker response & more pausing

• easiest to train backwards (VR15 then FI1m then FR10)

• limits to how far back gone

Discriminating Stimulus

A cue that signals the availability of reinforcement for a specific response, indicating when a behavior will be reinforced. It helps in differentiating which behaviors are likely to produce favorable outcomes.

• tells you when response is valuable/worthwhile → lets you know when to do response

Drive Reduction Theory

Hull, events were reinforcing because they led to the reduction of some kind of behavioral drive

• main/incentive motivation → pressing a lever delivers food, reducing hunger drive

• X every behavior reduces a need (ex - animals running on treadwheel just for fun)

Incentive Motivation

derived from the reinforcer indep of a physiological drive

• eating bowl of oatmeal would satisfy hunger, but is less attractive than eating a cheesesteak

• hobby or skill is often self rewarding

Premack Principle

everything is behaviors, low/high probability (LPB/HPB)

• issue w/ reinforcers → determining if/when they’re actually appetitive

• view reinforcers as behaviors X stimuli that can reinforce other behaviors

• more probable behaviors can be used as reinforcers for less probable behaviors, effectively increasing the likelihood of the less frequent behavior occurring.

High Probability Behaviors (HPB)

behaviors the organism is likely to perform in a given condition

• eating is a HPB when organism is hungry, LPB when it’s X

Low Probability Behaviors (LPB)

behaviors an organism is unlikely to perform in a given condition

• eating is a LPB when an organism is full

• lever-pressing X typical behavior for rats (if trained = HPB)

Premack & Operant Association

argued that operant association uses HPBs to encourage engaging in LPBs

• grandma’s rule → first you work, then you play

Premack Principle Application

give spaghetti 1st = organism might X run much

• if we make getting spaghetti contingent on running = promote running

• running is COMPARATIVELY LPB to eating spaghetti (both are HPB, just 1 seems lesser to other)

Response Deprivation Hypothesis

have existing knowledge of whether behaviors are H/LPBs & must use them accordingly

• could try to restrict access to thing & see if org will work to get it

• elasticity, sometimes people go w/o & sometimes people work hard to get something even if banned (ex - alcohol prohibition = mafias, becomes HPB w/ restriction)

• depends on baseline of behavior

Baseline of Behavior

preferred level of a behavior, amount an org will engage in w/ no restrictions

• related to response deprivation hypothesis

• assumed to be desirable amount = preferred

• ex → wheel put in hamster cage, how much does it want to interact

What does the response deprivation hypothesis consider?

if something is/X fulfilled, X classifying as high/low prob

• if rat wants 30g of food pellets/day & gets them, they’re satisfied

• if a rat wants to run for 1 hour a day but only gets 30 minutes, they are deprived

• could make runing contingent on eating: eating 40g food (R) → unlock wheel for 10 extra minutes (S^R), eating beyond preferred just to get other thing

• organis is trying to meet preferred amount of behaviors = moving away from preferred amount/satisfaction

Behavioral Bliss Point

The point at which an individual achieves a balance of preferred behaviors, maximizing overall satisfaction while minimizing restriction. It represents the optimal allocation of time and effort among various activities.

• if add a contingency = now impossible to meet all preferred levels

• org will adopt a best fit behavior pattern, minimize error (ex = regression)

What does the behavioral bliss point assume?

regression line, orgs will try to maximize overall reinforcement across all potential reinforcers

• inherently some things will be over/underrepresented

• assumes perfectly rational orgs, trying to minimize error/be as ideal as possible (even we aren’t fully rational & knowingly make bad decisions)

• some behaviors & inherent drives might be more important or attractive, so balance can be hard to acheive