Applied behavior analysis

Applied Behavior Analysis

Applied Behavior Analysis (ABA) is defined as the science of applying principals of behavior to improve social behavior. The analysis of behavior is also used to find the reasons for changes in behavior. ABA is used as a treatment for a variety of psychological disorders and developmental disabilities. These treatments are called behavioral interventions. It is also used to explain and improve various types of social behavior. Lastly it can be used to teach new skills to an individual or group.

History

The Journal for Applied Behavior Analysis (JABA) is the primary source for academic and scientific articles about ABA. It was founded in 1968. For this reason, the formal founding of ABA as a scientific discipline is believed to have started in 1968. In the first publication of JABA Baer, Wolf, and Risley wrote an article defining the dimensions, or characteristics, of Applied Behavior Analysis. This article is still used today to define exactly what ABA is. These dimensions will be explained in greater detail later in the article. So, while ABA is believed to have formally started in 1968 it can be traced back much further in history.

Behaviorism

The experimental analysis of behavior provided the basis for ABA. The experimental analysis for behavior is simply the science of behavior. Likewise, behaviorism is the study of the science of behavior. Behaviorists believe that behavior is caused by external sources in the environment therefore it can be studied objectively.

Psychologist John Watson founded behaviorism in 1913 and he is regarded as one of the first psychologists to study the science of applied behavior. He studied conditioned responding in humans and believed that internal feelings should not be studied at all. His findings led him to believe that behavior is shaped by the environment over time due to the outcome of an individual's actions or external stimuli. In other words he believed all behavior is learned and can therefore be changed. The concept of conditioning responses a prominent aspect of ABA treatments.

One of the most famous behaviorists was the psychologist B.F. Skinner. Skinner first published his ideas about behaviorism in 1938 and he continued to publish scholarly writings that focused on behaviorism throughout his career. In 1945, Skinner came up with the concept of radical behaviorism. Radical behaviorism is a more moderate form of behaviorism. It recognizes that internal feelings have an effect on behavior but says that experiments should still only focus on observable behavior. Skinner also proposed that behavior is partially a result of the environment but also partially determined by genetics. He studied behavior in many different areas using experimental analysis. So he used scientific experiments to come up with behavior principles. He was a very well known behavior analyst who was also involved with the founding of JABA. For these reasons he is often believed to be one of the founders of Applied Behavior Analysis. However, he never actually applied his behavioral principles to social situations like ABA seeks to do so he is not actually the founder. That being said, he still made huge contributions to the field of ABA by providing the basic behavior principles it is founded on.

Other factors in the development of ABA

Wilhem Wundt established the first experimental psychology lab in Leipzig, Germany in 1879. The founding of an experimental lab was an important moment in the history of psychology because it allowed psychologists to study behavior scientifically. Since, ABA is largely based on the science of behavior the foundation of experimental psychology was crucial to it's development.

Physiologist Ivan Pavlov's findings also proved important for the development of ABA. In 1897, he started to conduct the experiments about the physiological aspect of behavior that he is now famous for. His experiments led him to come up with the theory of classical, or pavlovian, conditioning. Classical conditioning states that if a neutral stimulus is consistently paired with a stimulus that produces a physiological reaction, overtime the neutral stimulus will produce the physiological reaction on its own. A stimulus is a physical event. A neutral stimulus is any stimulus that does not naturally produce a physiological reaction. A physiological reaction is a bodily response that is instinctive and automatic such as salivation or elevated heart rate. Pavlov's development of classical conditioning was important to the foundation of ABA because it was one of the first times that behavior was studied experimentally. It also proved that there was a relationship between environmental stimuli and behavioral reactions.

ABA emerged when the experimental analysis of behavior was applied to social situations. There are various people that began the process of applying these behavioral principles but one of the most famous was O. Ivar Lovass. In the 1960s, Lovass became one of the first people to use applied behavior analysis as a treatment for autism. ABA eventually became one of the most widely used treatments for autism spectrum disorders but this was not always the case. When Lovass began using ABA based behavioral interventions to treat children with autism he revolutionized the treatment of developmental disorders. He also believed that treatments will be more effective if they are started early in childhood. This belief in the importance of early intervention is still widely seen in ABA based interventions today.

One of the more recent developments in the field of Applied Behavior Analysis is the development of the Behavior Analysis Certification Board (BACB). The BACB is a non-profit corporation that was founded in 1998. This board certifies behavior analysts and ensures that they are providing the best service to those in need. The BACB was founded to provide consistency among behavior analysts. They make sure that the board certified behavior analysts (BCBAs) are using ethical, legal, and professional practices in their work.

Characteristics

There are 7 characteristics of ABA that were defined in detail in the 1968 article by Baer, Wolf, and Risley.

Applied

The term applied means that the behavior being studied must be socially important. Experimental behavior analysis generally studies behavior that is not socially relevant, like the behavior of animals in a laboratory. However, in order for something to be considered applied behavior analysis it must study behavior which provides some social benefit or causes social harm. Social activity is different for everyone so behavioral research that is applied is usually specific to an individual.

Behavioral

ABA is the scientific study behavior so naturally an actual behavior must be present. In other words, what an individual actually does is the important factor to study rather than what they say or feel. The behavior being studied must be reliably measured. If it is not reliably measured then the study of it cannot be scientific and data based.

Analytic

For behavior to be analytic is must be able to be controlled by a researcher. If a researcher can control a behavior they can further study it and prove the cause of it. ABA is focused on changing behavior but to change a behavior the behavior analyst must first figure out why that behavior is happening. There are two ways that researchers can prove they have control over a behavior. They both involve the use of an experimental variable. An experimental variable is something a reseacher intentionally applies during an experiment. Researchers want to prove that the variable they used produced the desired effect. In order to do this they have to compare data from before they applied the variable, known as baseline data, to data they collected after they applied the experimental variable. The first technique is called the reversal technique. WIth the reversal technique an experimental variable is applied and then the behavior being studied is measured to see if the variable had any effect on the behavior. If a change in behavior is noted it could be due to the experimental variable. In order to prove it is the experimental variable that effected the behavior and not another factor the variable must be removed, or reversed, and then the behavior is measured again. If the behavior changes back to its original rate it can be proven that the variable effected it. The second other option for proving control is called the multiple baseline technique. This is used when the behavior is impossible to reverse or when it should not be reversed due to safety reasons. With this technique, multiple baselines of different behaviors are taken first to produce a rate of the behaviors before intervention. The researcher then applies an experimental variable to the different behaviors one at a time and studies the effect it has on the rates of behavior. If a single experimental variable changes the rates of multiple kinds of behavior it will help to prove its effectiveness. Both of these techniques help researchers prove reliability. Reliability is the likelihood that a certain action produced a behavioral change. So if an action has high reliability it is believed that that single action was the reason for a change in behavior. The ability to analyze and control behavior is necessary to prove reliability.

Technological

Baer, Wolf, and Risley use the term technological to describe the amount of explanation necessary for behavioral interventions. Aspects of a behavioral intervention must be explained well so there is no confusion. Anyone with the proper training who reads a description of a behavioral intervention should be able to reproduce that behavioral intervention and get the same results. In order for a description to be considered technological it must define all aspects of the behavior being studied as well as all possible scenarios that could happen during the behavioral intervention.

Conceptually Systematic

For an aspect of ABA to be conceptually systematic it must be related to already known behavior principles. Decisions must be made based on data and aspects of behavior that are already defined. This makes ABA more effective and scientific.

Effective

Simply put this characteristic says that behavioral interventions must produce good results to be considered applied. The purpose of applied interventions is to improve an aspect of socially relevant behavior so if it does not improve social ability then it is not effective. Interventions that are not effective should not be used once it is proven that they are not effective. It can be difficult to tell how much behavior change makes a intervention effective because different people can have different opinions on the acceptable or ideal rate of behavior. So, the minimal rate of behavior change for an intervention to be considered effective is not a scientific factor. Instead it is up to the people that have to deal with the behavior regularly to determine if they intervention is effective or not.

Generality

A change in behavior has generality if; it continues for a long time after the intervention, is found in multiple environments, or causes a change in related behaviors. As a behavior analyst the goal is to produce behavioral changes that have high generality. High generality means that the behavior change was meaningful and will likely provide more consistent benefits in social behavior.

Definitions and concepts

The following terms are important to the application of behavior principles. These terms and ideas are used in ABA. This information comes from Behavior Modification Basic Principles.

Behavior

Behavior is defined as anything an organism does. So, anything that a living thing does is considered behavior. When studying behavior, it is assumed that all behavior is lawful. Traditionally, internal feelings and thinking are considered causes of behavior. However, radical behaviorism defines feelings and thinking as a behavior. A response is any time a behavior occurs. The rate of a behavior means how many times the behavior occurs in a given time period. When measuring the rate of a behavior you measure one specific behavior at a time.

In ABA, a behavior that is bad is called problem behavior. Problem behavior is something that harms the person doing it or harms other people around them. It could be physically hurtful but it does not have to be. A lot of ABA treatments try to decrease problem behaviors.

Respondent

Respondent behavior is is reflexive and involuntary. It is not under an individuals control. Environmental stimuli elicits respondent behavior. It is associated with classical conditioning because the responses are physiological and cannot be controlled.

Operant

Operant behavior is voluntary and able to be controlled by an individual. With operant behavior, the likelihood of a response is dependent on the outcome of the behavior. For instance, if good things happen after the behavior it is likely to be repeated but if bad things happen as a result of a behavior it is not likely to be repeated. Operant behavior is the basis for ABA because unlike respondent behavior it can be controlled which means it can be changed.

Three-term contingency

Three-term contingency is a concept for understanding operant behavior that was first used by B.F. Skinner. It continues to be used in ABA. It says that every behavior has an antecedent and a consequence. An antecedent is what happens right before the behavior. It can influence the likelihood of a behavior occurring. The consequence is simply what happen after the behavior occurred. In other words, it is the response to the behavior. So, an antecedent occurs before a behavior and a consequence occurs after a behavior. Keeping track of antecedents, behaviors, and consequences is called taking ABC data. ABC data can be used to determine behavioral trends. For instance if the same thing always happens before or after a behavior it can help determine the cause of that behavior. ABC data is not enough to prove the cause of a behavior in ABA because the behavior analyst needs to use experimental variables to prove the cause. However, ABC data is helpful for determining potential causes which can then be studied.

Reinforcement

When a stimulus increases the likelihood of a response occurring again it is said to be a reinforcement. A reinforcement is often thought of as a reward. But, there is a important difference between a reward and a reinforcement. A reward is any stimulus that is given after a response occurs. A reward can be a reinforcement but only if it increases the rate of response in the future. If it does not increase responding then it is not a reinforcement. All reinforcers dont work for all people. So it is important to try different types of reinforcers to find out what works best for a person.

Primary reinforcers are consequences that fill a biological need. Primary reinforcers are food, water, warmth and sex. They are reinforcing so they can alter behavior. The ability for them to alter behavior is genetic and natural. This means it exists in every person and does not need to be learned. Since primary reinforcers do not need to be learned they are also called unconditioned reinforcers. Reinforcers do not fill biological needs are called secondary reinforcers. Secondary reinforcers are not naturally reinforcing. A person needs to learn that they are reinforcing before they will increase the rate of behavior. They learn that they are reinforcing because they are paired with other known reinforcers. For this reason they are also called conditioned reinforcers.

A discriminative stimulus (SD) is something that tells a person that reinforcement is available. Ideally, the presence of a SD will lead to a behavioral response because the individual will make the connection between the reinforcement and the SD.

Reinforcement should be given right after the behavior that is being reinforced happens. If they are not given right away the person may not remember what they are getting reinforced for. So, the sooner a reinforcer is given the more effective it will be. Lastly, reinforcement can either be positive or negative.

Positive reinforcement

Positive reinforcement (SR+) is when the addition of a stimulus leads to an increase in responding. So something is added to the environment after a behavior which increases the likelihood an individual will produce that behavior again. For example, social praise could be a positive reinforcement. A paycheck also serves as a positive reinforcement for most adults because the addition of money increases the chance that people will go to work. It is important to think of the term positive in terms of adding something to the environment rather than being good. In this instance positive is not the same as good. There are four different types of positive reinforcement:

1. Tangible Reinforcers

Tangible reinforcers are things that can be touched or held. So, toys or books would be considered tangible reinforcements. Other tangible reinforcers could be clothes or electronics. Consumables are types of tangible reinforcers also. Consumables are anything that can be eaten or drank. With people with very serious developmental disabilities consumables may be one of the only types of reinforcers that will work. Tangible reinforcers can be expensive. They can also be seen as socially unacceptable to use. For these reasons they should not be the first choice for positive reinforcement.

2. Social Reinforcers

Social reinforcement is one of the best types of positive reinforcement. It is good because it is not expensive and easy to use. Social reinforcers are any action or object that is valued by society. Examples of objects that can be used as social reinforcement would be awards or nice letters of praise. Actions that are socially reinforcing could be compliments, smiles, high fives or attention. Social reinforcement can be very effective.

3. Activity Reinforcers

Activity reinforcers are any events that a person would like to earn. The activities could be getting to play games, go to the movies or even just earning a break from work. Activity reinforcers can work really well depending on the activity. A common issue with activity reinforcers is that they may not be able to be given soon enough. This would result in a delay of reinforcement. The reinforcer will be less effective if it is delivered a long time after the behavior.

4. Token Reinforcers

A token reinforcer is a neutral stimulus that can be traded for a object or activity at a later time. This object or activity would be something of value to the person who is earning the token reinforcers. The items or activities they can get are known as back-up reinforcers. The token reinforcers do not have value on their own. They are only valuable because the person knows that they can trade them for something they want. Token reinforcers are similar to money. They can be really effective because the person will get to choose the reward they want. They are also effective because they are given right after the target behavior happens.

Negative reinforcement

Negative reinforcement (SR-) happens when the removal of a stimulus increases the likelihood a individual will produce a behavior in the future. In this case the term negative does not mean bad it just means that something from the environment is removed. The stimulus that is removed from the environment is something aversive, meaning the person does not like it. There are two types of negative reinforcement. The first type is called escape. Escape occurs when the response causes the stimulus to be removed. So a response will stop the stimulus from continuing. An example of escape negative reinforcement would be turning off an alarm clock. The buzzing of an alarm clock is not pleasant and when a person turns the alarm clock off the unpleasant noise stops. So the removal of the unpleasant noise serves as reinforcement because it increases the likelihood that they will turn off the alarm clock the next time it goes off. The second type of negative reinforcement is avoidance. Avoidance occurs when the response prevents the stimulus from occurring at all. Avoidance occurs due to a history of contact with the aversive stimulus. With repeated contact a person learns how to avoid the stimulus. An example of this would be a child who cries when they have to go to school but do not want to go. If they dont have to go to school as a result of the crying then they will continue to cry because it helps them avoid the unpleasant stimulus which in this case, is going to school.

Reinforcement schedules

Schedules of reinforcement are the way in which reinforcements are given. The schedule will define the ways reinforcement will be earned. Different schedules are useful for different behaviors. The schedule used depends on the person and the behavior that is being reinforced

Continuous reinforcement schedules

Continuous reinforcement schedules (CRF) reinforce a person every time they do the target behavior. So every time the person performs a specific behavior they will get reinforcement. This generally leads to high rates of the target response. Meaning the person will perform the target behavior a lot. It is good for teaching a person new skills. When teaching new skills a high response rate will make the person better at the new skill. There are a few problems with CRF schedules though. The rates of behavior are not permanent. If the person stops being reinforced they will stop doing the target behavior. Another issue is satiation. If the person becomes satiated it means that they do not want the reinforcer anymore. So because they got the reinforcer so many times they are no longer motivated to work for it. A reinforcer must be changed when using CRF so the person does not get satiated. CRF schedules do not happen a lot in the real world. It is unlikely that a person will get reinforcement every single time they do a certain behavior in real life. Also CRF schedules should be changed into intermittent schedules over time. Changing them to intermittent schedules makes the behavior more reliable and long-lasting.

Intermittent reinforcement schedules

CRF schedules reinforce a person every time they perform a target behavior. Intermittent schedules of reinforcement only reinforce the target behavior sometimes. They are many types of intermittent schedules of reinforcement. The four most common types will be explained in more detail.

Fixed-ratio schedules

Fixed-ratio (FR) schedules give reinforcement after a certain number of responses. So once a person does a target behavior a certain number of times they will get reinforcement. The number of responses needed to get reinforcement can change depending on the person or behavior. For example on a FR 5 schedule reinforcement would be given for every 5 responses. On a FR 15 schedule reinforcement would be given for every 15 responses. FR schedules will produce high and steady rates of response.

Variable-ratio schedules

With variable-ratio (VR) schedules the amount of responses needed for reinforcement changes. The number of responses will be different every time. For example with a VR 10 schedule reinforcement would be given for an average of every 10 responses. This means that reinforcement could be given after the 2nd, 14th, 7th, 9th or 18th response. If it given on the 4th response one time it may be given on the 12th response the next time. It is random and not predictable. This leads to high and steady rates of response also.

Fixed-interval schedules

With fixed-interval (FI) schedules reinforcement is delivered for the first response that happens after a fixed amount of time passes. The reinforcer is delivered after the time interval finishes and the person responds once. For example, a FI 10 schedule would mean that the first response that happens after 10 minutes is reinforced. Reinforcement will not be given until after the 10 minutes passes. With FI schedules there will be low rates of responding in the beginning of the time interval. At the end of the interval there will be an increase in the rate of response. FI schedules are not very effective.

Variable-interval schedules

Variable-interval (VI) schedules give reinforcement after an variable amount of time passes. The amount of time that needs to pass before reinforcement will be given changes a lot. It is based on an average time period though. So, a VI 10 schedule will give reinforcement for the first response after an average of 10 minutes has passed. The response rate for VI schedules is higher than FI schedules. But the response rate is not as high as it would be with FR or VR schedules.

Differential Reinforcement

Differential reinforcement is a way that people can learn to make correct responses over time. The correct behavior is reinforced and the incorrect behavior is not reinforced. Over time people will learn to make the correct response because they want to receive the reinforcement. Differential reinforcement can also be explained as the process of teaching someone the difference between a SD and a non-discriminative stimulus (S-delta). A SD signals that reinforcement is available while a S-delta does not tell the person anything about the availability of reinforcement. A S-delta is not likely to produce responses because there is no reinforcement involved. There are four types of differential reinforcement that can be used to decrease a problem behavior. They are:

DRL

DRL stands for differential reinforcement of low-rate of response. DRL is used when the behavior normally happens at very high rates. Before starting a DRL program data needs to be collected to find out how often that behavior happens. A target frequency also needs to be set. A target frequency means how often you want the behavior to happen at the end of the program. It can be thought of as the end goal. A DRL program uses reinforcement to decrease rates of behavior a little at a time. Guidelines are set on how many behaviors are allowed to happen in a time period. If the person has less than the number of behaviors allowed then they get reinforcement. If they have more behaviors than the number that it allowed they do not get the reinforcement. The number of behaviors that is allowed in the time period will decrease with time. So the person must have less behaviors if they want to keep getting reinforcement. The number of behaviors allowed in a time period keeps decreasing until they reach the target frequency. DRL is helpful when trying to decrease a behavior but not get rid of it completely.

DRO

DRO stands for differential reinforcement of other behavior. With DRO, a person gets a reinforcer if they do not do the problem behavior for a set amount of time. The time period will be set before the program starts. If they have a problem behavior at any time during the period they will not get reinforcement. This program would be faded overtime as the behavior is decreased. The time period will increase little by little. This means the person will have to go longer without doing the problem behavior.

DRI

DRI stands for differential reinforcement of incompatible behavior. A DRI program reinforces behavior that is incompatible with the problem behavior. Incompatible means that it will not let them do the problem behavior. So, an incompatible behavior would stop them from being able to do the problem behavior. The incompatible behavior will increase because it is being reinforced. If the incompatible behavior increases then the problem behavior has to decrease.

DRA

DRA stands for differential reinforcement of alternative behavior. For a DRA program the person is taught an alternative behavior that they can do instead of the problem behavior. The alternative behavior would serve the same purpose as the problem behavior. So they would get the thing they wanted by doing the alternative behavior. The alternative behavior is reinforced and the problem behavior is not. So over time the person will increase the number of times they do the alternative behavior. Because the alternative behavior gets them the same result as the problem behavior they will not have to do the problem behavior anymore. So the alternative behavior replaces the problem behavior. The alternative behavior must be easy for the person to do. It should be easier to do than the problem behavior. If it is not easier then the person will keep doing the problem behavior. Before a DRA program is used the reason for the problem behavior must be understood. If the behavior analyst does not know why the person is doing the problem behavior they cannot find a alternative behavior for them. This is because the alternative behavior must be something that lets the person get what they wanted. So the behavior analyst must first find out exactly what it is they want.

Extinction

Sometimes a problem behavior will be reinforced by accident. If a problem behavior is reinforced it will increase rates of the behavior. An example of this would be a parent letting their child have candy if they scream and cry. The child will learn that if they scream and cry enough they will get the candy they want. So the candy served as reinforcement for the screaming and crying behavior. Since the child got reinforced they will scream and cry again the next time they want candy. In order to fix accidental reinforcement the behavior can be put through extinction. Extinction is a way to decrease rates of problem behavior. A problem behavior can only go through extinction if it has been reinforced in the past. To put a behavior through extinction the person must stop getting the reinforcement they expect. If they stop getting the reinforcement they want then the problem behavior will decrease. The problem behavior will decrease because the person will learn that this behavior no longer gets them what they want. Extinction can take a long time especially if the problem behavior has been reinforced for a long time. It is possible to make the extinction process go more quickly though. This can be done by teaching an alternative behavior to the person. The alternative behavior would have to be something that lets the person get what they want. So, a DRA program can be used at the same time as an extinction program. This will make the extinction program more successful.

Punishment

Reinforcement is always the first thing that is tried when trying to change or improve a behavior. This is because reinforcement produces longer lasting results and is generally better for the individual. However reinforcement does not always effectively reduce problem behavior. If all of the types of reinforcement prove to be ineffective it may be necessary to try punishment. Punishment is effective but not desirable because of its potential harmful effects. A punishment consists of any consequence of a behavior that decreases the likelihood of a behavior happening in the future. If a punishment has to be used it should be used with reinforcement. So, punishment of a problem behavior should be paired with reinforcement of the appropriate behavior. Just like reinforcement, punishment can be both positive and negative.

Positive

Positive punishment is the addition of an stimulus that decreases the chances of future responses. This stimulus is something undesirable and aversive. An example would be yelling because yelling is aversive so it will decrease the probability of the behavior happening again.

Negative

Negative punishment is the removal of a stimulus which decreases the chance of future behaviors happening. The stimulus with negative punishment is something that is desirable. So the removal of something that someone likes will make them less likely to repeat the behavior that caused the removal. An example of this would be not letting a child play with their toys when they do something wrong. Taking away the toys they like will decrease the chance that they will repeat the problem behavior.

Prompting

Prompting increases the likelihood that a behavior will be performed the right way. Using prompts will make the learning process easier and quicker. Prompts could be hints, modeled behavior verbal emphasis or physical assistance. The two broad categories of prompts are stimulus and response.

Stimulus prompts

With stimulus prompts the teacher will change the antecedent in the three-term contingency to make a response more likely to occur. An effective stimulus prompt makes it so the learner will realize the SD and associate it with the correct response. Some examples of stimulus prompts would be using a list or picture to guide responses and help the learner remember what the correct response is. Another example of a stimulus prompt is using a highlighter to make the important parts of a piece of writing more visible.

Response prompts

Response prompts do not involve a change in the antecedent. Instead the teacher provides the learner assistance with the actual behavior part of the three-term contingency. There are three types of response prompts. The first type is verbal prompts. Verbal prompts are spoken instructions. Verbal prompts are useful for teaching new behaviors.The second type of response prompt is a gestural prompt. A gestural prompt is when a teacher uses a gesture or some kind of movement of their body parts to indicate the proper response. The third type of response prompt is a physical prompt. A physical prompt is when the teacher physically touches the learner in order to help them perform the correct behavior. Physical prompts are often used when teaching new skills that involve movement of the body.

Fading

Prompting is helpful for teaching new behaviors to learners but once the behavior is learned the teacher should decrease prompting. The process of gradually decreasing prompting is called fading. Fading will help increase the independence of the learner because they will no longer have to rely on the actions of the teacher to produce responses. Fading is easier if you are using weak prompts such a verbal or gestural rather than physical which is always a stronger prompt. For this reason, the teacher should always use the weakest form of a prompt that will still produce the correct response. The amount a teacher can fade prompts depends on the capabilities of the learner. For instance, learners with really profound disabilities may always require some form of prompting to complete tasks.

Shaping

Shaping is a way to teach new behaviors or improve behaviors that have already been taught. It involves reinforcing successive approximations of a target behavior until the end behavior is reached. The end behavior is also called the terminal behavior. It can be thought of as the end goal. Successive approximations can be thought of as improvements that bring a person closer to the terminal behavior. So as a person improves an action they are reinforced for the new improvement only. Once they improve the behavior they will not be reinforced for the old behavior anymore. This will keep happening until eventually they reach the terminal behavior.

Chaining

Complex behaviors are made up of simple behaviors that are combined together. This is what the concept of behavior chains is based on. Behavior chains are the combination of simple behaviors that makes up a more complex behavior. The complex behavior is broken down into smaller behaviors that are done in order. So each of the smaller behaviors is a step in the behavior chain. A behavior chain is usually presented as a list of steps. Each step in a behavior chain is considered a response on its own. Each response serves as the discriminative stimulus (SD) for the next step. The SD then reminds the person of the next step in the sequence. It also reminds them that reinforcement will be available at the end. Each response also becomes a conditioned reinforcer for the previous step. It is a conditioned reinforcer because the learner will learn to associate it with the reinforcement they get for completing the entire chain. Chaining a behavior is a good way to teach new skills because it breaks the behavior down into smaller steps. This makes it easier to remember. There are three different ways to teach a behavior chain:

Forward chaining is the most obvious way to teach a behavior chain. The chain is taught starting with the first step. Once the learner can do the first step they move on to the second. This continues until they have completed all of the steps. The person learns how to do the steps in order. They do not move on the the next step until they can do the step they are on. Forward chains can take a long time because each step needs to be taught on its own.

Backward chaining is the opposite of forward chaining. With backward chaining the last step is taught to the learner first. Once can do the last step they move on to the second to last step. This continues until they make it to the first step. Once they make it to the first step they will have learned all the steps. The steps are not actually done in a backwards order though. The behavior is done normally starting with the first step. For example, the teacher will complete the steps for the learner until they get to the last step. Then the learner will do the last step themselves. Next, the teacher will complete all of the steps up until the second to last step. The the learner will do the last two steps on their own. This continues until they can do the entire chain on their own. There are some advantages to backward chaining. The learner gets to watch the behavior chain first before they have to do it themselves on the last step. The first steps will be repeated a lot so they will be easier for the learner to remember. They also get the reinforcement quickly in the beginning of the teaching. For instance they will only have to complete one or two steps before they get reinforcement at first. This will make them want to learn the behavior more.

Total task presentation is the third way to teach a behavior chain. It can also be called total task procedure. With forward and backward chaining only one step in the behavior chain is taught at once. But with total task the learner completes every step in the chain every time they do the program. They should complete every step in the correct order. Total task requires the learner know how to do all of the steps in the chain before they start. The learner may make more mistakes because total task can be complex. But the learner will learn the behavior more quickly with total task. This is because they do every step of the chain every time so they get more practice.

Analyzing Behavior

In applied behavior analysis recording behavior is very important. ABA relies on data to make decisions about behavior. This makes decisions better because they are scientific and objective. Objective means that they are based on facts rather than emotions. To makes these decisions the behavior must be measured a lot. So behavior analysts are always collecting data about a persons behavior. Before collecting data about a behavior the behavior must be defined. In ABA, this is called an operational definition. An operational definition tells people exactly what the target behavior looks like. The target behavior is the behavior that data is being collected on. The operational definition of the target behavior needs to be very good. People that have not seen this behavior before should be able to tell when it happens based on the operational definition.

Data collection

There are different ways that behavior analysts collect data about a behavior. First, they use direct observation. Direct observation is when behavior is watched and then rated by the person who is watching the behavior. In ABA direct observation is used with naturalistic observation a lot. Naturalistic observation means the behavior is studied in its natural environment. So the person watching the behavior would observe the behavior in the place it normally happens. Indirect measures could also be used but they are not as good. Indirect measures compare a persons behaviors to other peoples behaviors. They get information about a person this way. An example would be an intelligence test. One persons score is compared to the scores that other people got. These are not as good because everyone is different. Scores on indirect measures can change a lot too.

There are three common methods of data collection. They are frequency, interval, and time sampling. These methods of data collection are how the people observing the behavior actually record the behavior when it happens.

Frequency is how often a behavior happens. So frequency recording just counts the number of times a behavior happens. Every time a behavior happens it is recorded by the observer. Frequency data is easy to collect so it is used a lot. Behavior analysts use frequency data to determine rate of response. Rate of response is the number of responses for a specific time period. To find rate of response, the frequency is divided by the total time of the observation. For example if the target behavior happened 100 times in 10 hours the rate of response would be 10 times per hour. Rate of response is the most commonly used measures of behavior in ABA. It the the best measure of how often behavior happens.

Interval recording breaks the observation time into smaller pieces. The total time is divided into smaller intervals. If the behavior happens at any time during one of the smaller intervals it is recorded. But only one behavior per interval is recorded. So the total number of behaviors is not recorded. Analysts can get an idea of how often the behavior occurs with interval recording. But interval measurements will not be as accurate as the rate of response measurement. It can also be hard to measure behavior using intervals. It can be hard because the observer has to keep track of the intervals. To do this the observer must be paying attention all of the time.

Time sampling methods are a little different from frequency or interval methods. It can also be called momentary time sampling. It is called this because the observer only watches behavior for a moment. The observer will only check to see if the behavior is happening every once in a while. So the behavior is only recorded sometimes. If the behavior is happening when they check it gets recorded. But if the behavior happens any other time it does not get recorded. The observer does not have to be paying attention the whole time for time sampling. Because the observer is not paying attention the whole time they will not record all of the behaviors that happen. For this reason it is not as good as the rate of response measure.

Data analysts can also collect ABC data for problem behaviors. ABC data records the antecedent and consequence for the behavior. If ABC data is collected it can be analyzed. It can be analyzed for patterns which can be used to tell possible causes of behavior. Indirect methods such as interviews and questionnaires can also be used to analyze behavior. All of these methods should be used before doing a functional analysis. They should not be used instead of a functional analysis because they are not scientific like a FA is.

Functional analysis

A functional analysis (FA) is done to find the reason for problem behavior. The reason for problem behavior must be known before an behavioral intervention can be done. So before treatment for a problem behavior can happen a functional analysis must happen. Behavior analysts came up with the idea of a functional analysis in 1982. Since then it has become an important part of ABA. Functional analysis uses data collection to prove the reason for problem behaviors. Problem behavior is usually reinforced accidentally. This reinforcement increases the rates of problem behavior. FA's use this concept to prove the cause of problem behavior. A functional analysis is run like an experiment. During an FA, a behavior analyst will reinforce problem behavior on purpose. This is done to try to make the behavior happen like it would in real life. Behavior analysts do different things to try to make the problem behavior happen. Once it happens they reinforce it. The behavior analyst is trying to find the things that cause the highest rates of problem behavior. They can then say that the things that cause the behavior to happen are the reasons for the behavior. Once the reasons for problem behavior are known treatment can begin.

Behavioral interventions

Behavioral interventions can be done without doing an FA. A functional analysis is not always done. This is because they are difficult and take a lot of time. But these behavioral interventions are more likely to fail. Treatments based on the results of a FA are more effective. FA's improve behavioral interventions for two reasons. First, if the reason for the behavior is known it will be easier to stop reinforcing it by accident. Once reinforcement stops the behavior will decrease. Second, a new appropriate behavior can be taught to the person. This new behavior gets the person the same consequence as the old behavior. Since the behavior has the same effect it will replace the problem behavior. To know what the consequence the person wants is a FA would have to be done.

Use for treatment of autism

Applied behavior analysis is the most effective treatment for autism spectrum disorders. ABA is the only treatment for autism that has been proven to be effective through research. For this reason, ABA is used most for behavioral interventions for people with autism. The earlier an intervention is started the more effective it will be. That is why behavior analysts try to treatments with children when they are young. ABA has also proven to be effective when used to treat older children and adults. But it is not as effective when used with older people because their problem behaviors have been reinforced for a longer time.