ABA Toolbox‎ > ‎

Natural Environment Teaching

Natural Environment Teaching (NET) is utilizing principles of Applied Behavior Analysis to teach in the natural environment, “the real world”.

Because children with autism are not typically motivated to learn new things, Natural Environment Teaching can be particularly necessary and particularly difficult.  Since a child is taught away from a structured setting, following the child’s lead, using activities and materials in the child’s home, outside, day care setting, etc. that the child is showing an interest, an ABA professional unfamiliar with NET may opt away from NET in order to teach specifically in a structured setting.  As a matter of fact, many programs don’t teach new skills in the natural environment at all!  Unfortunately, this can lead to a child only being able to learn skills in one environment.  It also leads to skills being gained in therapy but not utilized outside of therapy.  What a waste! Even though the idea is to follow the child’s lead, a curriculum is usually planned ahead of time and followed through a session.  The curriculum would focus on a child’s specific needs and embed them within his/her interests. Characteristics of NET also include capturing motivation (EO/MO) pairing, errorless learning, and using probe data as opposed to trial by trial data.

Capturing the motivation of the child to teach new skills is especially important at the beginning of a program when a therapist/teacher is still developing a rapport with a child.  The idea is to capture a moment of high interest/motivation and use it to teach a new skill, whether it is naming/identifying colors, talking about categories, naming familiar people or characters, or requesting preferred items, etc.  

For example, a need that a child may have is to learn to vocally request items.  In NET, the situation may be arranged so that in order for the child to gain access to his/her stuffed “Barney”, which is slightly out of reach, he/she would ask for it.  When the child sees the toy and reaches for the item, the ABA provider may initially prompt the request by modeling the word “Barney”.  If the child says, “Barney”, the item would be given to him/her.  Quickly, the prompt would be faded, until soon enough, the child would be requesting Barney on his/her own.

This is just a simple example.  There is definitely opportunity to teach more and more complex behavior in the natural environment when motivation is captured appropriately.  The key is to target or program in such a way that skills are being taught in the way a child will most likely learn.  If a skill is too advanced for a child, it will most likely end in frustration.  If the skill is too easy, the child would not be learning new things.  E.g. If a child is not yet imitating words but is beginning to imitate sounds, a whole word would most likely not be required for a request.  However, as the child demonstrates better and better ability to imitate during requesting, better and better approximations would be required.

Teaching in the natural environment can be especially difficult when instructional control is not demonstrated.  In structured settings, instructional control is somewhat easier for an ABA provider to gain for several reasons:

1.) Usually the provider limits access to everything the child is interested in so they can be used as reinforcers.  Therefore, the environment itself is essentially, boring and not distracting.  

2.) Access to highly reinforcing items is in the provider’s control, so the therapy is teacher/provider led, rather than child-led.


3.) If reinforcement is not provided well, instructional control can still exist in such an environment because the child can still work to “escape” having to work… i.e. “go on a break”.  Therefore, an illusion of enjoyment can exist, in which the therapy provider feels that the child is enjoying working because he/she is compliant.  When in reality, the child may simply be compliant to get away from working.

In Natural Environment Teaching, instructional control is gained through pairing.  A therapist/ABA provider identifies things in the child’s environment that already serve as reinforcers to the child.  Then the provider lets the child gain access to those reinforcers only through the provider.  Over time, this procedure “pairs” the provider with reinforcing things and the provider him/herself becomes a conditioned reinforcer.  Now, if a therapy provider is a reinforcer, a child will be more likely to want to be around/gain access to that therapist!  Pairing is essential for NET because of the changes in environment that can occur at any given moment during therapy.  We’ve all seen children in a room bouncing from activity to activity.  Instructional control is necessary to be able to insert teaching this movement, keeping the activities “child-led”.

Probe data is typically taken during NET, though it is not used exclusively. Typically, when moving from environment to environment, probe data is simply easier to take than taking data on every single trial.  Since NET also relies on teaching quickly and bouncing from teaching program to program, probe data may also be more efficient.

Similar to Natural Environment Teaching are: 'Incidental Teaching" and "Milieu Language Teaching".  

*Incidental Teaching varies from NET in that the curriculum is not typically planned ahead of time.  The focus is usually to expand or enhance language that the child is already exhibiting.   

*Milieu Language Teaching typically refers to "in setting language instruction".  This includes the combination of incidental teaching, mand-modeling, and something called the time delay technique.  

*The time delay technique is used in the following way:  When an adult notices that a child wants something, rather than prompting right away, the adult waits to see if the child is going to attempt a request.  If the child does request but not "correctly", another delay occurs, where the adult waits to see if the child will offer another request.  If he doesn't or is not successful, the mand-model procedure is used.  

*The mand-model procedure is as follows:  When an adult notices a child wants something, first the adult asks the child what he/she wants.  If the child does not respond or responds incorrectly, the adult models the correct response.  When the child imitates the response, he/she is praised and given the desired item.