Differential Reinforcement: A Complete Guide

Applied Behavior Analysis (ABA) has many tools in the toolbox for improving behavior but one of the most important is reinforcement, including differential reinforcement. Differential reinforcement relies on 2 primary components, providing reinforcement for desired behavior and withholding reinforcement for target behavior. Sounds simple, right? It’s actually a bit more complicated, but the fundamentals really are simple. Although differential reinforcement is commonly used to reduce maladaptive behavior, adept practitioners also use this strategy to strengthen desired behaviors.

Differential reinforcement (DR) is an intervention that reinforces one topography of behavior while putting all other responses on extinction. Five main varieties offer options for behaviors to reinforce:

  1. DRA-Differential reinforcement of alternative behavior
  2. DRI-Differential reinforcement of incompatible behavior
  3. DRO-Differential reinforcement of other behavior
  4. DRL-Differential reinforcement of lower rates of behavior
  5. DRH-Differential reinforcement of higher rates of behavior

It can be difficult to keep these procedures straight. Download the free resources below for a reference.

Differential Reinforcement Guide

Differential Reinforcement Infographic

Differential reinforcement infographic

Contents

Jump to any section:

Types of Differential Reinforcement DRA: Differential Reinforcement of Alternative Behavior DRI: Differential Reinforcement of Incompatible Behavior What is the Difference Between DRA and DRI? DRO: Differential Reinforcement of Other Behavior DRL: Differential Reinforcement of Lower Rates of Behavior, DRH: Differential Reinforcement of Higher Rates of Behavior How to Implement DR Procedures

Related Posts

Types of Differential Reinforcement

There are several types of differential reinforcement (DR) procedures that address a variety of behavior. Although DR interventions are commonly used to reduce challenging behavior, they can also be used to increase desired behaviors. Karsten and Carr (2009) compared 2 DR procedures on their impact on skill acquisition. The authors found that differentially reinforcing unprompted responses may lead to more rapid skill acquisition.

The variations in DR interventions described below are distinguished by which behaviors receive reinforcement or when reinforcement is delivered. Understanding when and how to provide reinforcement determines the success of the procedure. Use the links here to jump to the DR procedure you’re most interested in:

Watch the video below for an overview of differential reinforcement.

Type of Differential ReinforcementWhat to ReinforceApplications
DRAA functionally-equivalent alternative behaviorWidely applicable for reducing maladaptive behavior
DRIA functionally-equivalent incompatible behaviorWidely applicable for reducing maladaptive behavior, used to differentially reinforce independent responding
DROThe absence of the target behaviorUseful in reducing potentially dangerous behavior
DRHBehavior occurring at a rate above a pre-determined minimum rateUseful in increasing the rate of desired behaviors that are in a learner’s repertoire but occur too infrequently
DRLBehavior occurring at a rate below a pre-determined maximum behaviorAdjusting the rate of appropriate behavior that occurs too frequently

Differential Reinforcement of Alternative Behavior (DRA)

Differential Reinforcement of Alternative Behavior (DRA) procedures reduce a problematic behavior by reinforcing an appropriate alternative behavior that serves the same function. Begin by identifying the likely function of the behavior, then choose a behavior that earns the child the same reinforcer as the challenging behavior to reinforce.

If you work with a child who consistently chews on his shirt and you identify that the behavior is reinforced by direct access to a sensory experience, provide an appropriate alternative item for him to chew on. In addition to providing this item, you may need to provide some additional, unrelated form of reinforcement, especially in the beginning, to encourage him to choose the item you provide over his shirt. This might be praise, an edible, or even an unrelated tangible reinforcer.

Here’s another example: You work with a child who has begun to swear. You identify that this behavior is maintained by socially mediated access to attention from his peers. You can implement a DRA procedure that reinforces joke telling which will serve the same function as swearing.

In the examples above, the behavior reinforced is an appropriate alternative behavior that serves the same function of the challenging behavior.

During a DRA procedure, we often choose some form of communication as the functionally equivalent replacement behavior, although this is not always the case. Functional communication training (FCT) is a common form of DRA procedure. For more information about FCT, read our article: FCT: 7 Things You Need to Know about Functional Communication Training.

Back to Top

Differential Reinforcement of Incompatible Behavior (DRI)

Differential Reinforcement of Incompatible Behavior (DRI) procedures are similar to DRA procedures; however, the behavior you choose is incompatible with the problem behavior. For a behavior to be incompatible, the child cannot perform the chosen action and the problem behavior at the same time. Again, the behavior you select to reinforce should serve the same function as the problem behavior.

Consider this example: You work with a child who hits his supervising adult when he is asked to do something he doesn’t want to do. You determine that this behavior is maintained by socially mediated escape from a relatively difficult task. You decide to reinforce writing the word “help” on a dry erase board when he doesn’t want to do his work. Writing the word “help” is incompatible with hitting. You can provide him with an escape from the difficult task by giving him the answer or part of the answer and you can also offer an additional reinforcer (i.e. praise, an edible, tickles, access to a preferred activity, a token, etc.) for selecting the incompatible behavior.

Back to Top

What is the Difference Between DRA and DRI?

Both DRA and DRI reinforce a functionally equivalent replacement behavior while limiting or eliminating reinforcement for the maladaptive behavior. The subtle yet highly important difference between these 2 interventions lies in the relationship of the alternative behavior to the target behavior. If the alternative behavior you choose to reinforce is incompatible with the target behavior, then the intervention is a DRI procedure. If the alternative behavior is not incompatible with the target behavior, then the intervention is a DRO procedure.

Target BehaviorFunctionExample Replacement Behavior (DRA)Example Incompatible Behavior (DRI)
ScreamingAccess to a tangibleManding for the desired itemWhispering the mand for the desired item
HittingEscape from sensoryWearing headphones Covering ears with hands to block the sound
ElopingAccess to an activityManding for the desired activityTapping your shoulder to get your attention prior to manding
RefusalEscape from a difficult taskManding for a breakSaying “I don’t want to do this” or “I need help” while continuing to complete the task

DRA in the Research

Legray, Dufrene, Mercer, Olmi, and Sterling (2013) evaluated the effectiveness of differential reinforcement of alternative behavior (DRA) with 4 typically developing young children in a classroom environment with the intent reduce problem behavior while also increasing more appropriate behavior. The authors determined that using DRA along with pre-teaching an alternative behavior had a significant impact on the behavior of the children. The study expands the research on DRA into general education, demonstrating the utility of behavioral intervention with individuals without disabilities.

Although extinction is often included as an element of a DRA procedure, the potentially serious results of this component (i.e. extinction burst) may make this element impractical or even unethical. Athens and Vollmer (2010) evaluated the effects of manipulating different variables of the reinforcer (i.e. duration, quality and delay) on the occurrence of problem behavior. The authors differentially reinforced by implementing longer duration, higher quality or shorter delay to reinforcement for the desired behavior and a shorter duration, lower quality and longer delay to reinforcement for problem behavior. They found that when all 3 of these elements were manipulated, there was a more positive impact on behavior. Even when reinforcement is available for challenging behavior, manipulating elements of reinforcement and differentially reinforcing an appropriate alternative behavior improves outcomes.

DRI in the Research

An evaluation of existing research conducted by Carr et al. (1990) found that DRI was most effective when the incompatible behavior was functionally related to the target behavior. Some studies found a much smaller treatment effect when the intervention simply reinforced a behavior that was topographically incompatible with the target behavior. While considering the compatibility of one behavior to the target behavior, keep in mind that you ultimately must address the function of the target behavior in some way.

A study by Daly and Ranalli (2003) looked at using DRA or DRI as part of a self-monitoring program for children with disabilities. The program uses Countoons as a visual representation of desired and undesired behaviors, as well as the reward for meeting the specified goal. The children then learn to count each time they engage in either behavior. If the child meets the goal for the session, the child earns the reward. This interesting application of the technology may also build independence for many individuals with developmental disabilities, although this is also likely to be effective with children without disabilities as well.

BehaviorIncompatible Behavior
ScreamingDrinking water, whistling, singing a lullaby, whispering the letters of the alphabet, quiet voice
HittingHands in pockets, walking to a quiet space without people, sitting quietly at a table
ElopingSitting in a chair, spinning in a circle, giving a high five
RefusalTask completion, saying, “yes, I will do that”

Think about your learner’s challenging behaviors. How would teaching any of the above incompatible behaviors serve him? Reinforcing “quiet voice” might eliminate screaming, but it doesn’t help the learner long-term. In the example in the previous chart screaming was maintained by access to a tangible. Consider when you want something tangible. If someone gave you a reinforcer each time you were quiet when you wanted something, how would this impact you? This actually happens often in our society, especially in the workplace. Often, those who are quiet and “toe the line” receive rewards. This is disempowering and not what we want for our learners. When using DRI, make sure the behavior you choose to reinforce helps the learner achieve the same reinforcer as the challenging behavior.

Semantic Difference Between DRA and DRI

While understanding the difference between DRA and DRI allows you to make critical treatment decisions, some of this difference is merely semantic. This difference falls along the lines that “all rectangles are parallelograms but not all parallelograms are rectangles” (for all of you geometry stars). Here’s the basic, semantic difference between DRA and DRI:

All DRI procedures are DRA procedures, but not al DRA procedures are DRI procedures.

What this means is that any incompatible behavior you choose is also an alternative behavior, but not all alternative behaviors are incompatible with the target behavior. This information helps you write behavior plans that clearly describe intervention procedures. Understanding the distinction assists you in communicating effectively with other professionals.

The Real Difference Between DRA and DRI

The primary difference between DRA and DRI is the relationship between the appropriate behavior that is reinforced and the target behavior. Many times your client benefits more from either an intervention that focuses more on the function of the target behavior or the form of that behavior. As stated earlier:

If the alternative behavior you choose to reinforce is incompatible with the target behavior, then the intervention is a DRI procedure. If the alternative behavior is not incompatible with the target behavior, then the intervention is a DRA procedure.

Which Intervention is Best?

As with all interventions in ABA, which intervention is best depends on your specific situation. Read the research on both of these interventions before making a final determination. Consider whether or not there are alternative behaviors that serve the same function and are also incompatible with the target behavior. These might be the most effective behaviors to teach your learner.

If you choose to implement a DRI procedure and the alternative behavior is not functionally related to the target behavior (i.e. it doesn’t serve the same function), then seek to include another intervention in your treatment package that addresses the function of the behavior.

A meta review of existing research by Wong et. al (2015) included DRI and DRA as evidenced-based interventions. Go ahead and give these interventions a try.

Back to Top

Differential Reinforcement of Other Behavior (DRO)

Differential Reinforcement of Other Behavior (DRO) procedures reinforce the absence of the target behavior. This procedure is easy to implement because you aren’t looking to reinforce a specific behavior, rather you provide reinforcement at the end of an interval that was free of the target behavior. Because you are reinforcing the absence of the challenging behavior, it’s not necessary that you accurately identify the function of the behavior. Use DRO for those behaviors that are multiply maintained or where you have been unsuccessful in identifying the function.

Although this procedure is simple, there are significant disadvantages as well. Because you target one behavior or class of behaviors, you often inadvertently reinforce other challenging behaviors. Despite this concern, DRO procedures can be highly effective in reducing significantly challenging behaviors. Check out this example:

You work with a child who often becomes aggressive. It’s unclear what the function is because he appears to engage in the behavior at random times across a variety of situations and a functional analysis is too risky. The behavior typically occurs several times (approximately 3x) per hour. You decide to implement DRO on a 15 minute interval (just under the baseline rate of behavior). You set the timer for 15 minutes. Each time the timer goes off without the occurrence of aggression, your client receives reinforcement. If your client engages in aggression, the interval is immediately reset at 15 minutes.

In this example, the only time that the interval will be reset is if aggression occurs. This means that your client may receive reinforcement after engaging in any other behavior, including other undesired behaviors such as screaming, flopping or having a tantrum. You must decide if the benefits of DRO for your client outweigh this possibility. In any event, DRO does not teach an appropriate replacement behavior so if using this strategy, you MUST combine it with another intervention to teach an appropriate replacement behavior.

When to Use DRO

As with any intervention, you must choose interventions that are supported by research and fit your specific situation. Differential Reinforcement of Other Behavior will be appropriate for some clients or behaviors and not others. Consider factors such as:

  • The skill of the interventionist
  • Reinforcers available to your client
  • The severity of the target behavior
  • Information available in the research
  • Your client’s reinforcement history
  • Other competing contingencies

DRO is a behavior reduction procedure. By providing reinforcement only when the behavior does not occur, you ultimately delay reinforcement when the behavior does occur. This lower rate of reinforcement following the target behavior leads to the decrease.

DRO has proven to be valid under a variety of conditions. Research has demonstrated the effectiveness of Differential Reinforcement of Other Behavior to reduce challenging behaviors such as:

  • Self-injurious behavior
  • Aggression
  • Disruptive behavior
  • Hyperactivity
  • Pica
  • Thumb sucking
  • Stereotypy

As with anything, this procedure offers some risks and benefits that you must evaluate prior to implementation. Use DRO only once you carefully consider the advantages and disadvantages and you determine that the advantages outweigh the disadvantages.

Advantages and Disadvantages of Differential Reinforcement of Other Behavior

DRO procedures are widely implemented in the field of ABA. Experienced professionals account for the risks associated with the intervention through the development of a comprehensive treatment package. Here’s what to consider before choosing DRO as your differential reinforcement procedure.

Advantages of DRODisadvantages of DRO
Easier to implement than other DR procedures You risk reinforcing behaviors not targeted by the intervention
Directly addresses the problem behavior DRO does not teach an appropriate behavior

Many times, you rely on parents, teachers, paraprofessionals or less experienced staff to implement interventions on a daily basis. In these instances, you must choose interventions that are easy to implement in the natural environment. If the intervention can’t be implementing reliably, then you must choose a different intervention. Research demonstrates that people who are unfamiliar with behavioral interventions are able to implement DRO with fidelity.

Many behavior analytic procedures for decreasing maladaptive behavior seek to reinforce more appropriate functionally equivalent replacement behaviors without directly addressing the target behavior. While many of these interventions are effective and appropriate in some situations, if you want to directly address the problem behavior, you may find DRO a desirable intervention.

However, realize that during DRO, you provide reinforcement only when the target behavior does not occur. this applies only to the target behavior. In traditional DRO procedures, you still provide reinforcement even when other undesired behaviors occur. This can pose a serious threat if your client engages in a wide variety of topographies of inappropriate behavior.

Here’s an example:

You target aggression with DRO, providing reinforcement only if aggression did not occur during the predetermined time period. You are not currently targeting tantrums, so you must provide reinforcement even if your client engages in a tantrum during that time period.

In the above example, you inadvertently reinforce tantrum behavior, should tantrums happen to occur during the interval. In many instances, it may be more beneficial to decrease a potentially dangerous behavior, even if you risk reinforcing a disruptive behavior. You then either target the disruptive behavior with a different intervention or you target behaviors sequentially, adding new target behaviors as the more severe behavior decreases.

The other main disadvantage of DRO is critical. DRO alone does not teach an appropriate behavior. At its foundation, it simply teaches the child to stop engaging in a behavior. As professionals in the field, we all know that behavior occurs continuously and when you reduce one behavior, another one must take its place. If you don’t teach the learner what behavior you want to see, then the learner may choose a different undesirable behavior.

Whenever you choose to implement DRO, you must consider including additional interventions as part of a treatment package. You must teach a functionally equivalent replacement behavior to mitigate this concern. If the function of the behavior is unclear, continue with your functional assessment or practical functional assessment (PFA).

Comparison Between Noncontingent Reinforcement (NCR) and DRO

Vollmer, Iwata, Zarcone, Smith and Mazaleski (1993) conducted a study comparing the effects of noncontingent reinforcement and differential reinforcement of other behavior. The intent behind the study was to determine if noncontingent reinforcement (NCR) could serve as an alternative to DRO and if that intervention assuaged the disadvantages of DRO.

The authors found both interventions were effective, possibly due to the relationship between the selected reinforcer and the identified function of the target behavior. NCR did in fact avoid some of the limitations of DRO, and was therefore, preferred by the authors. When choosing interventions you must choose the ones that support the learner and provide the best experiences for him.

Professionals might find it difficult to convince parents faces with severe challenging behavior to provide reinforcement noncontingently. Because DRO more closely aligns with traditional parenting techniques resulting in a specified consequence for challenging behavior, it can be easier to get parent buy-in. Easy doesn’t mean better. Take time to consider what is truly in the best interest of the learner, not everyone around him.

The Role of Extinction in DRO Procedures

DRO procedures are most commonly effective when combined with extinction. Extinction is a behavior reduction procedure in which you withhold reinforcement for a previously reinforced behavior. While it may be impractical or unsafe to withhold reinforcement for some behaviors, the effects of DRO can be compounded when the target behavior receives no reinforcement.

Keep in mind the risk of an extinction burst (the behavior escalating once extinction is implemented). Make sure that you are prepared to work through the extinction burst prior to implementing extinction. Extinction can also be unethical, harmful, or perceived as abusive in some situations. Although some studies show differential reinforcement procedures may be more effective when combined with extinction, there is an important growing recognition of the potential harmful effects of extinction.

The Matching Law is an effective alternative to the use of extinction in DR procedures. Learn more in our post Matching Law: Practical Applications in ABA and in the video below:

Token Economy and DRO Procedures

If you want to delay reinforcement more gradually or over longer periods of time, consider including token economy in your DRO procedure. This allows the child to receive attention and acknowledgement for a shorter interval while seeing progress toward the backup reinforcer.

There are many different ways to incorporate the use of token economy with your DRO procedure. Consider creating fun boards featuring your client’s favorite characters. If you are short on time like I am, you can order the “I Can Do It” Caterpillar Token Board pictured below on Amazon for only $6.95. There’s also a 10 pack available for only $29.95 which is a huge time saver if you’re working with more than 1 client! This is truly one of my favorite time savers, especially for children who love The Very Hungary Caterpillar!

Token economy offers a way to provide frequent reinforcement, especially for children who engage in attention maintained behavior. Finding efficient ways to utilize this intervention in your practice puts a valuable tool in your ABA toolbox. You can learn more about using token economy by reading our post: Token Economy: Examples and Applications in ABA

Examples of DRO from the Research

Professionals in the field of ABA are held to high standards when selecting appropriate interventions. We must choose evidence-based interventions, supported by research. Differential reinforcement procedures are well documented in the research literature and commonly used in a variety of different situations. Despite the important disadvantages to Differential reinforcement of other behavior, the research continues to support its use under certain conditions.

Differential Reinforcement of Other Behavior Combined with Visual Schedules to Reduce Problem Behavior During Transitions

A study by Waters, Lerman and Hovanetz (2009) looked at the impacts of combining DRO with visual schedules on problem behavior that was triggered by transitions for 2 children with autism. The authors found that visual schedules alone were insufficient at reducing target behavior,. When they added visual schedules to a treatment package including DRO and extinction, they observed a decrease in problem behavior. Interestingly, this decrease was observed whether or not visual schedules were used. Although visual schedules have been supported by research, this study demonstrates the power of DRO in reducing challenging behavior.

Video Example of DRO

In this video, Hitomi Wada uses a fun work example to demonstrate the use of DRO to reduce workplace gossip.

Back to Top

Differential Reinforcement of Lower Rates of Behavior (DRL)

Differential Reinforcement of Lower Rates of Behavior (DRL) procedures are similar to DRH procedures except you reinforce gradually decreasing rates of behavior. This intervention is ideal for a behavior you want to reduce but not eliminate. For example:

Your client says “hi” to everyone she sees, even if she has said “hi” to them in the last 5 minutes. You calculate her baseline responding at 10x/5 minutes. In order to ensure success, you set the initial criteria to achieve reinforcement at 8x/5 minutes. Over time, as she reaches this criteria, you establish new criteria that help her approach a more sustainable rate.

Austin and Bevin (2011) used DRL to reduce the requests for teacher attention of 3 students in primary grades. Not only was the intervention effective, but the teacher found the intervention highly acceptable. Gaining the buy-in (social validity) of the interventionist directly impacts the success of interventions.

Use DRL to Reduce Behaviors, Not Eliminate Them

DRL provides a system for reducing a behavior without eliminating it. This form of differential reinforcement can be used in an assortment of different situations to gradually shape behavior over time. Add this valuable tool to your ABA toolbox.

When to Use DRL

Differential reinforcement procedures are some of the most widely used interventions within ABA. DRL, although less common can be an effective means for reducing behaviors that are situationally appropriate, yet occur at too high of a rate. Situations where this might be beneficial include:

  • Asking for help
  • Raising a hand to answer questions
  • Eating
  • Completing homework

The above behaviors are desirable, yet if they occur at too high of a rate, they can become problematic. Take a look at how problems occur:

  • A child who asks for help with every task fails to build independence.
  • A student raises their hand to answer every question, does not allow other students to answer the question and may become frustrated if not called on
  • Eating too quickly leads to digestive problems and obesity
  • Completing homework too quickly leads to errors and sloppy work

Targeting these behaviors through other differential reinforcement procedures may result in unintentionally extinguishing them rather than just reducing them. When choosing which interventions to include as part of your package, carefully consider the following:

  • Severity of the behavior
  • Current levels of behavior
  • Potential for harm of the behavior
  • Skill of your interventionist
  • Other responsibilities of your interventionist
  • Ultimate goal for the rate of the behavior
  • Reinforcers available to the child

Prior to making your treatment decisions, take a look at how the advantages and disadvantages of DRL impact your specific situation.

Advantages and Disadvantages of DRL

While Differential Reinforcement of Low Rates of Behavior is not appropriate for every situation, there are many advantages that may make it a desirable intervention, but also consider the disadvantages before making your decision.

Advantages of DRLDisadvantages of DRL
Lends itself to changing criteria to continue to reduce the behavior to a predetermined level Not intended to completely eliminate behavior, would need to ultimately add a component of extinction to eliminate the behavior
Reduces the chances of an extinction burst Requires counting each instance of the behavior
Some procedures require minimal intervention making it practical for group environments such as classrooms Not practical for behaviors occurring at very high rates

Video Example of DRL

In the video below, Dr. Mark Berg discusses the use of DRL to slow rates of responding in regard to the rate of eating candy and doing homework. He also demonstrates a wait program with his furry friend. This wait program demonstrates the use of Spaced Responding DRL to increase the interresponse time between occurrence of the behavior (eating) resulting in a decrease in the overall rate of the behavior.

Back to Top

Differential Reinforcement of Higher Rates of Behavior (DRH)

Sometimes learners engage in a desirable behavior at a rate far below what we want to see. Using Differential Reinforcement of Higher Rates of Behavior (DRH) systematically reinforces incremental increases in the rate of that behavior. Take a look at this example:

Your client seldom raises his hand in class to answer a question, especially during math block, even though he often knows the correct answer. You decide to use DRH to reinforce a higher rate of hand raising. When you calculate his baseline rate, you get 2/45 minute math block so you set your first criteria at 3. You provide reinforcement if he raises his hand to respond to questions at least 3 times during that block of time. Over time, as he becomes successful, you gradually increase the criteria he needs to meet to achieve reinforcement.

When to Use DRH

Use DRH when the learner knows how to perform a task or skill but doesn’t do it fast enough to encounter natural reinforcement. This procedure helps build fluency and confidence for many learners. Situations where DRH may be beneficial include:

  • Raising a hand in class to answer questions
  • Walking at a pace that aligns with peers
  • Completion of problems during homework
  • Work related tasks such as sorting mail

Consider how performing each of these tasks too slowly or infrequently becomes problematic for the learner:

  • Receives low grades for participation in class discussion despite knowing the material
  • Falls behind his peers when walking between classes and misses important social opportunities
  • Spends hours on homework each night instead of engaging in recreational activities with peers
  • Struggles to maintain employment

Make sure when setting reinforcement criteria that the goal is achievable for the learner. If you set the initial reinforcement criteria too high, you might unintentionally put the desired behavior on extinction.

Take a look at an example of using DRH to reinforce a higher rate of typing:

Back to Top

How to Use DR Procedures

The steps for implementing differential reinforcement procedures are essentially the same as for any DR procedure you choose. Here we will review a couple of examples using DRO and DRL.

Example of DRO Procedure

DRO requires many specific steps for effective implementation. DRO reinforces the absence of the target behavior, making it a time-based intervention.

  1. Define the target behavior
  2. Identify the function of the target behavior
  3. Choose reinforcers
  4. Collect baseline data
  5. Determine the type of DRO procedure you will use
  6. Set the criteria to advance or reduce the interval
  7. Define your procedures
  8. Implement the intervention and collect data

1. Define the target behavior

The definition of the target behavior must be clear enough so that others who collect data do so accurately. This is a critical step, so don’t skimp on the definition. DRO requires analysis of data to determine if the child meets criteria to advance or reduce the interval. If your definition is unclear, your data won’t be accurate and your intervention will not be as effective as it could be.

Writing a good operational definition takes practice and can be tricky at first. If you need some help with this step, you can learn how to write an operational definition in our free course Writing Operational Definitions. Take a look at the example below to get you started:

Aggression: Any instance of Liam making forceful physical contact with another person using his body or an object with enough force to cause an audible sound and/or visible mark on their skin. 

Examples Include:

  • Biting- teeth making contact with any part of the skin, constricting and leaving a mark
  • Hitting- using a hand or arm with a closed or open fist to hit (making forceful physical contact) with another person
  • Kicking – using the foot/feet or leg/legs to kick or hit another person
  • Spitting – any instance of saliva leaving the mouth (excluding while the student is talking or yelling) with the saliva landing within one foot of a person
  • Throwing objects – any instance of him throwing items that are not designed to be thrown with enough force that the object lands at least 2 feet from another person’s body

Non-Examples Include:

  • Giving a high five
  • Hugging
  • Common social physical interactions
  • Saliva leaving his mouth as a result of talking or yelling

2. Identify the function of the target behavior

Correctly identifying the function of the behavior allows you to identify the variables that support the behavior. Although with a traditional DRO procedure, you don’t teach a functionally equivalent replacement behavior, as stated when discussing the disadvantages, you must include a plan to teach the child how to get what he wants in a more appropriate way.

In addition to teaching other skills, identifying the function of the target behavior can help you with the next step, choosing reinforcers. Learn more about the different methods of identifying the function of a behavior by reading our article: Functions of Behavior in ABA: Complete Guide

3. Choose reinforcers

While all of the steps listed above are important, without an effective reinforcer, your intervention is likely to fail. Choosing a reinforcer that serves the same function as the target behavior may amplify the effects of your intervention. Take a look at this example:

You decide to implement a DRO procedure to reduce tantrums maintained by access to the iPad. You set up your DRO and decide to use access to the iPad to reinforce the absence of the tantrums.

In the example above, your intervention has the added benefit of reducing the Motivating Operation (MO) for the iPad. Your client knows that he will have access to the iPad during the reinforcement period, thus may be less likely to engage in problem behavior to get it.

4. Collect baseline data

Baseline data helps you determine the appropriate length of the DRO interval. These data should be collected over at least 3 data points and during relevant activities (activities that are often associated with the target behavior).

There are many different data collection methods for you to choose from. The most appropriate method will depend on a variety of factors, including: the behavior you target, who will be collecting the data, and how many other responsibilities the person collecting the data has. For more information about choosing an appropriate data collection system, read our article: How do I Choose the Right Data Collection Method for my ABA Program?

For whole interval DRO procedures, it often makes the most sense to use a whole interval data collection system. Similarly, for momentary DRO, it makes sense to use momentary time sampling. You can use a simple interval data sheet like the one below, just make sure that you fully define the behavior for the person or people who will be collecting data.

Interval Data Sheet
Interval Data Sheet
Frequency Data Sheet
Frequency Data Sheet
Duration Data Sheet
Duration Data Sheet

5. Determine the type of DRO procedure you will use

Repp, Barton and Brulle (1983) described 2 main types of DRO procedures: interval and momentary, although there are some variations you may also choose to add. The best type of procedure depends on your specific situation and impacts your outcomes. Let’s compare the 2 types.

Whole-Interval

Whole-interval DRO procedures provide reinforcement when the behavior does not occur for the entire interval. For these procedures, you set the timer for the specified amount of time and provide reinforcement only if the behavior did not occur during that period of time.

If the behavior occurs, you have 2 choices: reset the interval once the behavior stops or wait until the next specified interval. For example:

Your interval is 2 minutes and your client engages in the target behavior at 1 min 30 seconds, will you reset the interval to 2 minutes or will you wait for the end of the next interval to deliver reinforcement?

The decision must be made ahead of implementation and remain consistent. This should be specified in step 7, define your procedures. In either event, the behavior must not occur at all for the specified amount of time.

Momentary

Momentary DRO is an interval method with some important differences from whole-interval DRO. For momentary DRO set the timer for the specified amount of time and if the behavior is not occurring at the moment the timer goes off, then deliver reinforcement.

The major disadvantage to this method is clear: you run a very high risk of providing reinforcement even if the behavior occurred extensively throughout the interval. By shortening the interval, you begin to mitigate this concern. The advantage is the ease of use of this system. It can be easily implemented in busy environments when 1:1 support is unavailable.

Types of Intervals

When establishing your interval system, choose from two types of intervals:

  • Fixed
  • Variable
Fixed

With a fixed whole-interval or momentary DRO procedure, each interval is the same duration. This method is easier for staff to implement as it is one less thing to remember or to keep track of. The child learns to predict when the reinforcer will be available.

Variable

For a variable whole-interval or momentary DRO procedure, each interval is an average duration of time. For example, if the interval is a VI 15 minutes, then reinforcement would occur at intervals that average 15 minutes. Reinforcement could occur at 10 minutes, 20 minutes and 15 minutes. In this method, the child can’t predict when the reinforcement will occur. Although this is an advantage for some children, this might also be a disadvantage for others who find they are not encountering reinforcement sufficient to refrain from the behavior. The major disadvantage to this method is the complexity it adds to the intervention. If you can find a systematic way for staff to determine when to provide reinforcement on a variable schedule, you are likely to achieve some positive results.

Although some research has shown that a fixed whole-interval DRO is more effective than a variable whole-interval DRO, there may be occasions where you may choose to vary the interval.

Whole Interval DROMomentary DRO
Reinforcement dependent on behavior across the entire intervalReinforcement dependent on behavior at a specified moment in time
Requires continuously watching for behavior to occurRequires looking for behavior at the specified time
Fixed or variable intervalFixed or variable interval
Reset the interval if the behavior occursWithhold reinforcement if the behavior occurs
Comparing Methods for Programming DRO

Repp, Barton and Brulle (1983) compared the effectiveness of whole-interval and momentary DRO. In the randomized study where some subjects received the momentary DRO procedure first followed by the whole-interval and others received the whole-interval DRO procedure first.

Their study determined that whole-interval DRO was more efficient at reducing behavior than momentary DRO. The authors suggest that the utility of momentary DRO may be in using it to maintain low rates of the target behavior. They suggest beginning with whole-interval DRO and then once the behavior is reduced to acceptable levels, implement momentary DRO to ensure that these levels are maintained over time.

6. Set your criteria to advance or reduce the interval

Use the information you acquired during your baseline data collection (step 4) to determine how long the interval should be. The goal is to increase the length of the interval over time, but the initial interval should be set well under what you collected during baseline (about half). Estimate the amount of time between incidents of the behavior (# of intervals x length of intervals = inter-response time). If you want a more precise measure, you can collect inter-response time data (IRT) in step 4.

Once you have established the appropriate interval, determine your reinforcement (R+) interval and progression to increase the length of the DRO interval. The chart below provides an example, given that the behavior occurred, on average about every 6 minutes during baseline and the goal is to increase the interval to 10 minutes. Above the chart is the criteria to advance of regress a step.

Criteria to advance step:  Two consecutive days with fewer than 2 incidents of target behavior.

Criteria to regress step:  Four consecutive days with more than 4 incidents of target behavior.

StepDRO IntervalR+ Interval
13 minutes1 minute
24 minutes2 minutes
35 minutes3 minutes
46 minutes4 minutes
57 minutes5 minutes
68 minutes6 minutes
79 minutes7 minutes
810 minutes8 minutes

7. Define your procedures

To ensure that your intervention is implemented with fidelity, you must thoroughly define your procedure. Here’s an example of a DRO targeting aggression for Liam, a 3 year old diagnosed with autism:

Differential Reinforcement of Other Behavior (DRO):
  1. Offer Liam a visual menu of preferred items he can work for (i.e. bubbles, iPad, etc.).
  2. Staff should set a timer for the prescribed step (see below).
  3. At the end of every interval that Liam is free of target behaviors (i.e. aggression as defined above), staff should present verbal praise (i.e. “Great job having a safe body! You’re doing awesome!”) along with the item selected in step 1. Staff should set the time for the prescribed R+ interval (see below). If Liam selected an edible, the R+ interval ends once the edible is fully consumed.
  4. If Liam engages in a target behavior, immediately stop the timer.  Provide as little attention as possible while monitoring for safety.  Do not respond verbally.  It’s important that staff present a calm demeanor and neutral facial expression/body language when Liam is escalated.  Once at baseline, reset and restart the timer.  
  • When the timer signals the end of the R+ interval, Liam should return to his regular activities.
  • Repeat steps 1-7.
StepDRO IntervalR+ Interval StepDRO IntervalR+ Interval
130 seconds1 minute52 mins, 30 sec2 minutes
21 minute1 minute63 minutes3 minutes
31 mins, 30 sec1 minutes73 mins, 30 sec3 minutes
42 minutes2 minutes84 minutes4 minutes

Criteria to increase step:  Two consecutive days with fewer than 4 incidents of target behavior.

Criteria to decrease step:  Four consecutive days with more than 6 incidents of target behavior.

8. Implement the intervention and collect the data

This is where all of your hard work and preparation pay off. The key to successful implementation is sufficient staff (i.e. RBT, teacher, parent, etc.) training. Many will resist providing reinforcement when various maladaptive behaviors occur that aren’t targeted. Withholding reinforcement for a wide variety of behaviors confounds the intervention and runs the risk of the child not encountering sufficient reinforcement to discriminate between conditions. Be clear with staff that you will eventually target the other behaviors (or you will target them in a different way), but only the behavior targeted by the DRO should result in withholding the reinforcement.

Continue to collect data to ensure you move through the steps outlined in your definition of the procedure. This is critical to the success of your DRO. Remaining at the same interval length too long can result in lack of progress and ineffective intervention.

Back to Top

Example of DRL Procedure

As an example, here we will walk through the steps to implement DRL procedures. Becoming fluid with these steps will help you utilize all of the tools in your ABA toolbox. Here are the basic steps:

  1. Define the target behavior
  2. Identify the function of the target behavior
  3. Choose reinforcers
  4. Collect baseline data
  5. Determine the type of DRL procedure you will use
  6. Set the criteria to advance or reduce the reinforcement criteria
  7. Define your procedures
  8. Implement the intervention and collect data

1. Define the target behavior

The definition of the target behavior must be clear so that others who collect data can identify whether or not the behavior occurs at any given moment. The definition is critical to the success of your intervention plan. Don’t take shortcuts here. DRL requires an accurate count of the behavior to determine whether or not the child met reinforcement criteria. If your definition is unclear, your data won’t be accurate and your intervention will not be as effective as it could be.

Writing a good operational definition takes practice and can be tricky at first. If you need some help with this step, you can learn how to write an operational definition in our article: Operational Definitions: Clearly Define the Behavior. Take a look at the example below to get you started:

Asking for help: Approaching an adult or raising hand to get adult’s attention during independent work time. 

Examples include:

  • Raising hand while sitting at desk during writer’s workshop
  • Walking to the teacher’s desk during silent reading
  • Waving arm above head during math time to ask for help

Non-examples include:

  • Raising hand to answer a question during morning meeting
  • Walking up to an adult at recess

2. Identify the function of the target behavior

Correctly identifying the function of the behavior allows you to identify why the behavior continues (i.e. access or escape). Although during a DRL procedure, you don’t teach a functionally equivalent replacement behavior, having an understanding of why the behavior occurs helps you develop your intervention package. When implementing DRL, you will want to include a plan to teach a functionally equivalent replacement behavior. For more information about the possible functions of behavior, read our article: Functions of Behavior in ABA: Complete Guide.

In addition to building your intervention package, identifying the function of the target behavior can help you with the next step, choosing reinforcers. Learn more about choosing reinforcers in our article: Choosing Reinforcers: Reinforcer Assessments or Preference Assessments.

3. Choose reinforcers

Identifying an effective reinforcer is critical to the success of your DRL procedure. If you choose a reinforcer that serves the same function as the target behavior your intervention may prove more effective. Take a look at this example:

You decide to implement a DRL procedure to reduce the rate of asking for help maintained by access to teacher attention. You set up your DRL procedure to include access to the teacher’s attention as a component of the reinforcer available for meeting reinforcement criteria.

In the example above, your intervention has the added benefit of reducing the Motivating Operation (MO) for teacher attention by providing contingent attention for meeting criteria. Your client knows that she will have access to the teacher’s attention during the reinforcement period, thus may be less likely to engage in problem behavior to get it. Read our post Choosing Reinforcers: Reinforcer Assessments or Preference Assessments if you need help choosing a reinforcer.

4. Collect baseline data

Baseline data helps you determine the appropriate reinforcement criteria for your DRL procedure. These data should be collected over at least 3 data points and during relevant activities (activities that are often associated with the target behavior).

Baseline data helps you determine the appropriate reinforcement criteria for your DRL procedure. These data should be collected over at least 3 data points and during relevant activities (activities that are often associated with the target behavior).

Because DRL evaluates either the interresponse time (IRT) or the rate of the behavior, these are the data you will need to record.

To collect IRT, simply use a stopwatch to record the time between the end of one behavior and the beginning of its next occurrence. Data will be presented as an amount of time (i.e. minutes/seconds, etc.). For example: 6 minutes, 23 seconds.

To collect rate, count the number of occurrences of the behavior and divide by the unit of time you will be using. Data will be presented as a number over a predetermined amount of time (i.e. minute, hour, etc.). For example: 4.33 instances per hour.

5. Determine the type of DRL procedure you will use

There are 3 basic types of DRL procedures you can choose from based on your overall goal for behavior change, spaced responding, interval and full-session.

Spaced Responding DRL

Spaced Responding DRL seeks to increase the interresponse time (IRT) as a means for reducing the overall rate of the behavior. To begin, you will need accurate IRT baseline data to determine your initial reinforcement criteria. Set your initial reinforcement criteria below baseline data to ensure success. Take a look at this example:

Your baseline data reflect IRT of: 6 min 23 seconds, 8 min 47 seconds, and 7 min 14 seconds. You choose to set your initial reinforcement criteria at 5 minutes to ensure that your client contacts reinforcement. When your client has an IRT greater than 5 minutes, he receives reinforcement.

Interval DRL

Interval DRL sets a maximum criteria per interval in order to access reinforcement. This allows for reinforcement to be delivered more frequently than during full-session DRL and is more appropriate for individuals who are not yet able to delay reinforcement. Here’s an example:

Your baseline data reflect rate data of: 6 instances/hour, 5 instances/ hour, and 9 instances per hour. This is an average of 6.67 instances/hour. You set your initial reinforcement criteria at 5 instances/hour to ensure your client contacts reinforcement. When your client engages in fewer than 5 instances of the behavior in an hour, she receives reinforcement.

Full-Session DRL

Full-session DRL establishes a maximum criteria per session (i.e. day, class, etc.) in order to access reinforcement. This type of DRL offers an efficient intervention when 1:1 intervention is unavailable. Here’s one more example:

Your baseline data reflect the following data: 10 instances/class, 8 instances/class and 13 instances/class, for an average of 10.33 instances/class. You set the initial reinforcement criteria at 9 instances/class to ensure your client contacts reinforcement. When your client engages in fewer than 9 instances of the behavior during the class, he receives reinforcement.

Spaced RespondingIntervalFull-Session
IRT dataRate dataRate data
Increases time between responsesDecreases number of responses over timeDecreases number of responses over time

6. Set the criteria to advance or reduce the reinforcement criteria

Rarely is your initial reinforcement criteria your terminal goal. Often, it will be necessary to shape the behavior so that it gradually occurs at lower rates until it finally reaches the desired rate. Once you have established the appropriate initial interval, determine your reinforcement (R+) interval and progression to increase the length of the DRL reinforcement criteria. Here’s an example of changing criteria for an IRT DRL procedure when the baseline data showed an average of 6 minutes IRT at baseline. The criteria to advance or regress a step is above the chart.

Criteria to advance step:  Two consecutive days with fewer than 2 incidents of target behavior.

Criteria to regress step:  Four consecutive days with more than 4 incidents of target behavior.

StepDRO IntervalR+ Interval
15 minutes1 minute
26 minutes2 minutes
37 minutes3 minutes
48 minutes4 minutes
59 minutes5 minutes
610 minutes6 minutes
711 minutes7 minutes
812 minutes8 minutes

7. Define your procedures

To ensure that your intervention is implemented with fidelity, you must thoroughly define your procedure. Here’s an example of an Full-Session DRL procedure targeting asking for help for Violet, a 7 year old diagnosed with autism:

Differential Reinforcement of Low Rates (DRL):
  1. Offer Violet a visual menu of preferred items she can work for (i.e. bubbles, iPad, etc.).
  2. At the end of every class during which Violet engages in fewer than the prescribed number of incidents of asking for help (as defined above), staff should present verbal praise (i.e. “Great job working independently! You’re doing awesome!”) along with the item selected in step 1. Staff should set the time for the prescribed R+ interval (see below). If Violet selected an edible, the R+ interval ends once the edible is fully consumed.
  3. If Violet engages in more than the prescribed number of incidents of asking for help,  withhold reinforcement and avoid engaging in conversation or otherwise drawing attention to the behavior. 
  • When the timer signals the end of the R+ interval, Violet should return to her regular activities.
  • Repeat steps 1-7.
StepDRO IntervalR+ Interval StepDRO IntervalR+ Interval
130 seconds1 minute52 mins, 30 sec2 minutes
21 minute1 minute63 minutes3 minutes
31 mins, 30 sec1 minutes73 mins, 30 sec3 minutes
42 minutes2 minutes84 minutes4 minutes

Criteria to increase step:  Two consecutive days with fewer than 4 incidents of target behavior.

Criteria to decrease step:  Four consecutive days with more than 6 incidents of target behavior.

8. Implement the intervention and collect data

Here’s where you get to see the fruits of your labor. Teach the interventionist (i.e. RBT, teacher, parent, etc.) to complete each step of the procedure as you defined it in step 7. The interventionist may express concern that you are providing reinforcement even when the behavior occurs far more often than is appropriate. This becomes a more pressing concern for staff when the behavior poses a potential safety risk or happens at so high a rate that it presents a real problem. Assure staff that this is necessary in order to reach your end goal.

Continue to collect data to ensure you move through the steps outlined in your definition of the procedure. This is critical to the success of your DRL. Remaining at the same interval length too long can result in lack of progress and ineffective intervention.

Back to Top

Test Your Knowledge

Follow this choose-your-own-adventure style video to test your understanding of differential reinforcement procedures. Can you help Amanda, a new BCBA choose the right procedure?

Other Uses for Differential Reinforcement

Differential reinforcement has a wide range of applications. Any time you want to value one behavior over another, you can use a differential reinforcement procedure without extinction.

For example, if you are working with a client who responds to questions correctly, but often does so slowly, you can differentially reinforce more rapid responses over slower responses. If he often takes 10s to respond correctly, you can provide a stronger reinforcer for any response within 5s. If you’re using a token economy system, you can provide 2 tokens for rapid responding and only 1 token for a correct response. This type of procedure reduces frustration on your client’s part over not providing any reinforcement for slow responding.

Important Considerations when Using Extinction

An important component of differential reinforcement is extinction. In Applied Behavior Analysis, withholding reinforcement for a previously reinforced behavior is called extinction. Often these behaviors have been unintentionally reinforced and extinction is implemented to reduce challenging behaviors. The challenging behavior should not receive reinforcement.

Although extinction is a common component of DR procedures, there are some risks associated with its use. Take our course Ethics of the Use of Extinction in Differential Reinforcement Procedures to learn more and earn 2 ethics CEUs!

Any time you use an extinction procedure, you will likely experience an extinction burst. This means the behavior will likely get worse before it gets better.

Extinction burst in children with autism

When dealing with intense or potentially dangerous behaviors, be cautious when choosing an extinction procedure. Make sure that you have a plan for keeping everyone safe if behaviors escalate to dangerous levels. Ensure you have the competence to implement this procedure or organize appropriate supervision.

Extinction may not be a practical intervention in all situations. If, after careful consideration, you determine that you can’t risk an extinction burst, consider using the matching law by offering a less potent reinforcer for the challenging behavior and a very strong reinforcer for the alternative behavior. For example, if attention maintains the challenging behavior, provide a quiet reprimand for inappropriate behavior and extended attention for alternative behaviors.

Borrero and Volmer (2002) found that behavior occurred at rates proportional to the rate of reinforcement for both problem and appropriate behavior during differential reinforcement procedures. This important research offers an alternative to the potential risky extinction procedures common when implementing DR.

Functions of Behavior

An important key to several types of differential reinforcement is identifying the function of the undesired behavior. All behavior occurs because it works for the individual. Either the learner gets something they want (a tangible, attention, or a sensory experience) or escapes something they don’t want (a demand, attention, or an unpleasant sensory experience). Your differential reinforcement procedure depends on correctly identifying the function of the maladaptive behavior.

Be sure to read out post Functions of Behavior in ABA: Complete Guide to learn more about identifying the functions of behavior. This post also provides a free download with helpful examples.

Download the Complete Guide

Find more details and examples of differential reinforcement in our complete guide, Differential Reinforcement: A practical guide to using DR, available on Teacher’s Pay Teachers.

References and Further Reading

Athens, E. S., & Vollmer, T. R. (2010). An investigation of differential reinforcement of alternative behavior without extinctionJournal of Applied Behavior Analysis43(4), 569-589.

Austin, J. L., & Bevan, D. (2011). Using differential reinforcement of low rates to reduce children’s requests for teacher attention. Journal of Applied Behavior Analysis44(3), 451-461.

Borrero, J. C., & Vollmer, T. R. (2002). An application of the matching law to severe problem behavior. Journal of Applied Behavior Analysis35(1), 13-27.

Carr, E. G., et al. (1990). Positive approaches to the treatment of severe behavior problems in persons with developmental disabilities: A review and analysis of reinforcement and stimulus-based procedures. monograph no. 4., 1-44. Retrieved from https://search.proquest.com/docview/62974256?accountid=166077

Chowdhury, M., & Benson, B. A. (2011). Use of differential reinforcement to reduce behavior problems in adults with intellectual disabilities: A methodological reviewResearch in Developmental Disabilities32(2), 383-394.

Daly, P. M., & Ranalli, P. (2003). Using countoons to teach self-monitoring skills. Teaching Exceptional Children, 35(5), 30-35. Retrieved from https://search.proquest.com/docview/201148804?accountid=166077

Dille, L. (2009). A comparison of two curricular models of instruction to increase teacher repertoires for instructing students with autism (Order No. 3367979). Available from Education Collection. (304865342). Retrieved from https://search.proquest.com/docview/304865342?accountid=166077

Flynn, S. D., & Lo, Y. (2016). Teacher implementation of trial-based functional analysis and differential reinforcement of alternative behavior for students with challenging behavior. Journal of Behavioral Education, 25(1), 1-31. doi:http://dx.doi.org/10.1007/s10864-015-9231-2

Karsten, A. M., & Carr, J. E. (2009). The effects of differential reinforcement of unprompted responding on the skill acquisition of children with autismJournal of Applied Behavior Analysis42(2), 327-334.

Legray, M. W., Dufrene, B. A., Mercer, S., Olmi, D. J., & Sterling, H. (2013). Differential reinforcement of alternative behavior in center-based classrooms: Evaluation of pre-teaching the alternative behavior. Journal of Behavioral Education, 22(2), 85-102. doi:http://dx.doi.org/10.1007/s10864-013-9170-8

Myerson, J., & Hale, S. (1984). Practical implications of the matching lawJournal of Applied Behavior Analysis17(3), 367-380.

Petscher, E. S., & Bailey, J. S. (2008). Comparing main and collateral effects of extinction and differential reinforcement of alternative behaviorBehavior Modification32(4), 468-488.

Repp, A. C., Barton, L. E., & Brulle, A. R. (1983). A comparison of two procedures for programming the differential reinforcement of other behaviorsJournal of Applied Behavior Analysis16(4), 435-445.

Trosclair‐Lasserre, N. M., Lerman, D. C., Call, N. A., Addison, L. R., & Kodak, T. (2008). Reinforcement magnitude: An evaluation of preference and reinforcer efficacy. Journal of Applied Behavior Analysis41(2), 203-220.

Vollmer, T. R., Iwata, B. A., Zarcone, J. R., Smith, R. G., & Mazaleski, J. L. (1993). The role of attention in the treatment of attention‐maintained self‐injurious behavior: Noncontingent reinforcement and differential reinforcement of other behaviorJournal of Applied Behavior Analysis26(1), 9-21.

Ward‐Horner, J. C., Pittenger, A., Pace, G., & Fienup, D. M. (2014). Effects of reinforcer magnitude and distribution on preference for work schedulesJournal of Applied Behavior Analysis47(3), 623-627.

Waters, M. B., Lerman, D. C., & Hovanetz, A. N. (2009). Separate and combined effects of visual schedules and extinction plus differential reinforcement on problem behavior occasioned by transitionsJournal of applied behavior analysis42(2), 309-313.

Wong, C., Odom, S. L., Hume, K. A., Cox, A. W., Fettig, A., Kucharczyk, S., … & Schultz, T. R. (2015). Evidence-based practices for children, youth, and young adults with autism spectrum disorder: A comprehensive review. Journal of autism and developmental disorders45(7), 1951-1966.

Processing...