Provide Reinforcement Without Delay
For a consequence to be most effective as a reinforcer, it should occur immediately after the behaviour occurs.- Miltenberger (2008, p. 83).
With Little or No Delay!
When running an educational programme and a student engages in a target behaviour it is important to deliver reinforcement with little delay, and when possible no delay at all (Malott & Trojan-Suarez, 2004).
When we say “target behaviour” we just mean the behaviour (or response) you want the child to engage in. For example, if you asked your student to “point to the dog” then the target behaviour would be the student pointing to the dog. In a home environment, if you asked your child to put their toys away in the cupboard, then the target behaviour would be the child putting their toys into the cupboard.
Immediacy of Reinforcement
This principle of reinforcing without delay is sometimes referred to as the “immediacy” of reinforcement. Basically, the ideal situation during a teaching session is for reinforcement to be delivered immediately after the target behaviour (Miltenberger, 2008).
The best case scenario would be a delay of 0 seconds between behaviour and the delivery of reinforcement; so effectively no delay at all (Cooper, Heron, & Heward, 2007).
Why is this Important?
To understand why this is important, consider the two examples below. In these examples, a student is being taught to discriminate between colours by pointing to different cards with colours printed on them; the student’s correct responding is being positively reinforced by the delivery of verbal praise (e.g. “Great job!”) and tokens for a token economy.
Example One: the teacher says “point to blue,” the student points to the blue card, the teacher says “Great Job! You earned a token!” and then gives the student a token straight away.
Example Two: the teacher says “point to blue”, the student points to the blue card, the teacher then turns to their right to pick up a card that fell on the ground but while doing this the student accidentally knocks a pencil off the table without the teacher noticing it. After picking up the photo the teacher turns back around and says “Great job! You got your token!” and gives him a token.
In the first example, the student will know exactly why they got their token but in the second example they might think they’re getting it for knocking a pencil on the ground.
Of course, this example is overly simple just to illustrate the point, but there are many things that can happen between the time a child responds correctly and the delivery of reinforcement.
The longer it takes the teacher to deliver reinforcement the greater the chance an unintended behaviour will occur and this unintended behaviour could end up being the one reinforced.
Importance for New Skill Acquisition
When first teaching a new skill it is especially important to follow this principle of providing reinforcement immediately after every target behaviour has occurred.
It is crucial that students understand exactly what response to give when their teacher instructs them to do something or asks them a question. By delivering reinforcers immediately after target behaviours the teacher is making sure her/his students understand exactly how to respond.
If there is any confusion about what constitutes a correct response then there could be an unnecessary delay of skill acquisition and a prolonging of the teaching sessions.
Essentially, teachers need to make sure there is a “pairing” of behaviour and reinforcement so everything is clear for their students.
Without Delay…but Sometimes Not After Every Response
A distinction should be drawn between delivering reinforcement without delay and delivering reinforcement after every target behaviour.
As already mentioned, providing reinforcement without delay is particularly important when teaching a new skill, however as a student makes progress in learning the skill, reinforcement will often not continue to be provided after every single correct response (Cooper, et al, 2007).
This is particularly important if tangible reinforcers such as tokens or edibles are being used. Note that verbal praise would almost always be provided after each correct response on a continual basis regardless of whether the student is becoming skilled at a task.
Basically what we’re saying is that, if you are going to deliver reinforcement after a target behaviour make sure the time between the behaviour and reinforcement is as short as possible but that reinforcement does not necessarily have to be delivered after every target behaviour.
Reinforcement could be delivered after every second, third or fourth (or more) target behaviour but remember to do this only when you feel your student fully understands the task.
For example, when first teaching a student to discriminate between red and blue you would deliver reinforcement after every correct response; then when teaching them to discriminate between yellow and green you would do the same. But when you move to teaching them to discriminate between red, blue, yellow and green you might decide that for the first three discrete trials (teaching attempts) you will deliver a token after every correct response but from then on you would deliver a token after every second correct response because you would already know that the student understands the difference between red and blue, and yellow and green.
Of course, in the example of above, if you found that your student was getting a lot of the discriminations incorrect you might need to go back to delivering reinforcement after every correct response to build the skill up again.
Three Reasons Why This is Important
Firstly, if you have a student who is becoming skilled at a task and is still earning one token or edible for every correct response, they may spend less time learning and more time with back-up reinforcers (i.e. too much time spent chewing a sweet or using their tokens to “pay” for a 5 minute break outside in the playgound). Of course, if using a token economy, you could just increase the number of tokens required to earn a back-up reinforcer.
Secondly, the delivery of tokens or edibles after every single correct response can become distracting and reduce the concentration levels of your student. Sometimes getting into a flow when teaching can keep your student's motivation levels up but continuously having to take a token and put it on their token board can prevent them getting into that flow.
Thirdly, it is generally understood that by increasing the number of responses needed to earn reinforcers you will typically find that ‘the rate of responding is greater’ (Miltenberger, 2008, p. 88). In other words, by making reinforcers harder to earn, you will often find that students will work harder to get them.
Relationship to Schedules of Reinforcement
This change to the amount of reinforcement that is delivered relates to what we call schedules of reinforcement. Basically, when first learning a skill, a student might typically be on what is called a “continuous schedule of reinforcement” which basically means that throughout the teaching session, reinforcement is “continuously” provided for every correct response.
When it is clear the student is beginning to learn the skill, the schedule of reinforcement might change to an “intermittent schedule of reinforcement” which means reinforcement would be delivered after some but not all correct responses (reinforcement provided intermittently).
The main message to take away from this teaching tip is to make sure that you deliver reinforcement to your students straight after they engage in a target behaviour. Doing this will ensure that there is a "pairing" of the target behaviour and the delivery of reinforcement, and in doing this your students should clearly understand what responses to give in order to earn their backup reinforcers.
- Cooper, J., Heron, T., & Heward, W. (2007). Applied Behaviour Analysis. New Jersey: Pearson Education.
- Malott, R. & Trojan-Suarez, E. (2004) Principles of Behaviour. New Jersey: Pearson Prentice Hall.
- Miltenberger, R. (2008). Behaviour Modification. Belmont, CA. Wadsworth Publishing.