chapter 15: differential reinforcement

what is differential reinforcement? where have we seen it been used before?

• used before in shaping; reinforcing one behaviour, while not others.

  • extinguishing undesirable behaviours (withholding reinforcement), while reinforcing desirable behaviours.

    • undesirable behaviours ↓ while desirable behaviours ↑

DRA

• Differential reinforcement of alternative behaviour

  • problem behaviour is no longer reinforced (extinguished) + alternative behaviour is reinforced instead.

• the undesirable and alternative behaviours do not have to be related.

DNRA

• a subtype of DRA where the reinforcement of the problem behaviour is negative → an aversive stimulus is removed

two steps in using DRA

1. determine the most appropriate alternative behaviour

2. determine the most effective reinforcement

what makes a good alternative behaviour?

• requires the same or less effort than the problem behaviour

• reinforced by natural environment

• is already in the learner’s repertoire

what makes a good reinforcement

• potent/powerful

• can be delivered immediately and consistently

• same reinforcer as the one maintaining the problem behaviour

thinning

continuous reinforcement → intermittent reinforcement

DRI

• extinguishing an undesirable behaviour while reinforcing a behaviour that cannot simultaneously occur in the presence of the problem behaviour

  • eg. arriving early to class vs arriving late to class, squirming in the dentist’s chair vs. sitting still in the dentist’s chair.

FCT

• functional communication; reinforcement of functionally equivalent behaviour

  • 2 behaviours that have the same consequence

• the “alternative behaviour” is an appropriate communication response

  • e.g. a child screams to avoid difficult homework - a child says “may I have a break” to escape difficult homework; these two behaviours have the same consequence (escape), but one is more socially acceptable.

“other behaviour” is not synonymous with ________ behaviour

alternative

DRO

• problem behaviour is extinguished; the absence of the problem behaviour is reinforced

  • after a period of time, a reinforcer is delivered, and the undesirable behaviour does not occur.

  • e.g. a teacher provides reinforcement if the student goes 5 minutes without leaving their seat.

whole interval DRO

• reinforcer is only delivered if problem behaviour is absent for the entire interval

  • if problem behaviour occurs, interval resets.

momentary DRO

• reinforcer is only delivered if problem behaviour is absent at the exact end of the interval

how to use DRO (6)

1. functional assessment → correct reinforcer

2. choose an effective reinforcer

3. choose the time interval length

4. Extinguish problem behaviour

5. Reset time interval if needed (e.g. whole interval)

6. increase time interval

the initial DRO time interval should be based on two things:

• baseline data

• should be slightly less than the IRT

behavioural vacuum

• the problem behaviour is eliminated, but there is no alternative behaviour that is implemented to replace it.

  • occurs in those with limited behavioural repertoire; DRA and DRI may be better alternatives

DRL

• reinforcement of lower frequency behaviours; the aim is to lower the frequency of behaviour, not eliminate it.

  • e.g. sleeping, quickly eating, raising hand in class

full session DRL

• reinforcement provided only if behaviour occurs under x amount of times.

  • e.g. a child is rewarded with dessert if they do not leave the dinner table more than 3 times.

spaced responding DRL

• reinforcement provided only if x amount of time has elapsed between behaviours.

  • e.g. DRI: 10 seconds

    • behaviour → 10 seconds → behaviour → reinforcement

interval time DRL

• a session/observation period is divided up into intervals; behaviour must occur a maximum of once in each interval for reinforcement to be provided

DRL vs FI

• DRL: based on response (e.g. only provide reinforcement if 60 seconds have elapsed between two responses)

• FI: based on reinforcers (e.g. only provide reinforcement if 60 seconds have elapsed between two reinforcers)

how to use DRL

1. determine whether DRL is appropriate (reduction > extinction)

2. determine the rate of behaviour

3. decide whether to use full session or spaced responding DRL

4. communicate criterion to the client

5. give feedback to client → monitor progress