Imagine your child just dropped cereal all over the kitchen floor. They look at you and you say, “don’t step on the cereal; it will make a mess!” They lift their foot while looking at you and stomp on it. “Ugh” you say as they laugh. Now, you are left with the question: why did my child do that?

What causes behavior? To start to find a cause, we want to first define behavior itself. What is behavior? Behavior is any interaction between someone and the environment. Hailing a cab? Behavior. Taking a bite? Behavior. Hitting a sibling? Behavior. Saying, “I love you”? Behavior. If you can do it, it is behavior.

So what causes behavior? As you may suspect, this is not a simple answer. To illustrate the complexity of identifying causes of behavior, let’s use a food analogy. Consider the omelet – what CAUSES an omelet? Is it just eggs? No. Is it the grumble in your stomach that makes you grab a pan and turn on the stove? Not alone. Maybe it is going into the kitchen and getting a whisk? Still not the only thing that causes an omelet. To cause an omelet, you need lots of ingredients and events to come together in the right way. Behavior is a little like that; no single event or ingredient leads to behavior. There are many variables that have to come together to cause a behavior. Functional behavior assessment (FBA) and functional analysis (FA) are ways to figure out what causes behavior.

Technically, the reasons for behavior are called functions. This refers to how the behavior works (functions) as part of the environment. The omelet, for example, is a function of eggs, milk, a whisk, a pan, and some heat; the omelet functions to reduce hunger, increase energy, impress your mother-in-law, maybe. You can see from this analogy that the omelet (our stand in for behavior) is both a function of some events, and functions to change some other events. Behavior is the same way. Essentially, all behavior is a function of a few basic events:

  • Learning history (you can’t learn something you haven’t been exposed to)
  • Biological function (think neurons, hormones, muscles)
  • Environmental variables (basically everything in the world outside your skin).

The environmental variables part is the focus of FBA and FA. Although the other parts are equally important in determining behavior, they just aren’t the focus of FBA and FA.

Why do we care about function? We care about function because behavior happens for a reason and once we find that reason, we can change something about the behavior. Knowing the function tells us what to change before the behavior occurs to make it less likely, and what to teach the child to do instead to achieve the same outcome that previously was a function of problem behavior.

  • We can prevent problem behavior to begin with: Knowing what leads to problem behavior can help us prevent it – if Bobby screams when he has been playing alone for too long, we can work on giving him more social interaction to prevent screaming in the first place.
  • Function-based treatment works best: Evidence shows that teaching children to meet their functions with appropriate skills is more effective at getting rid of problem behavior than punishment or medication alone.
  • It tells us exactly what to teach instead: Knowing this information helps appropriately design interventions to decrease problem behavior and increase appropriate behaviors. For example, if Sally hits every time you say “I-Pad is all done”, and then you give her back the I-Pad, hitting probably functions to get toys back. A good intervention will teach Sally to get the I-Pad by requesting it appropriately.

Functional Behavior Assessment (FBA) includes all the different ways to learn about behavior by testing behavior directly (an analysis) and ways that do not (all other types of assessment):

  • Review of existing data
  • Interviews (asking the parent, teacher, or even the student themselves about behavior)
  • Observation of the behavior without testing anything directly(further ABC data)
  • Functional analysis

For our omelet, we might start by reviewing recipes for omelets – but you will find they are not all the same. We can also ask some people who make lots of omelets (what happens before? What happens during? What happens after?), or we could watch someone make an omelet. All of these examples are like FBA, in that they do not necessarily involve testing omelet making by yourself. You might find out a lot about making omelets, but you can’t be absolutely sure you know how to make one unless you try it and see what happens.

Functional Analysis (FA) is a scientific way of trying it and seeing what happens. It involves testing how certain changes impact the occurrence of problem behavior. Like the omelet, you can find out lots about what people think, write, or do when making an omelet, but that information is not as accurate or exact as information found through direct testing.  An FA is the only way to know for sure what causes behavior (Cooper, Heron & Heward, 2014). Causes of behavior fall into four basic categories:

  • Attention: from teacher, parent, peer, etc.
  • Escape/avoidance: of a situation, person or activity
  • Tangible: Obtaining a preferred object or activity
  • Sensory: the behavior fulfills sensory need

In an FA, we test one thing at a time. For example if we are testing attention, there will be nothing in the room, except a person. We will only talk to the child when they engage in the behavior we are testing. That way we find out if behavior might function to gain attention. We repeat these tests multiple times until we have stable results that make us confident behavior is most likely to occur when certain things happen – that gives us the function! Here are some other ways we test:

  • Escape/avoidance: something the child does not like is removed when the behavior occurs.
  • Tangible: Something the child likes will be made available when the behavior occurs.
  • Ignore/toy play: the child is observed to see if the behavior occurs without being provided with a task, attention, or preferred item when the behavior occurs. We want to see how much behavior occurs when we don’t change anything or we just let the child play with toys. Seeing a lot of behavior during this time leads us to believe the behavior is making its own reinforcement (automatic function).

How do we use the information? By knowing what situations the behavior occurs in more, we can decrease problem behavior by making behavior plans that focus on the function. We can also use this information to teach appropriate behaviors that gives the child what they want. Appropriate behaviors include raising your hand for the teacher’s attention, asking for a break from academic work or asking to play with a toy. It is possible that the analysis will tell us there are more than one reason the behavior is happening (behavior can be really complex!). It may also tell us more about the situations where the behavior does not occur. All of this information is useful to the behavior analyst when making behavior plans.

What are the limitations? Ideally these assessments will provide us with a clear function(s) of the behavior, which will help to develop a successful intervention. Sometimes in the FA, making a testing situation leaves out something small, but important, that happens in the natural environment (maybe the child only wants a green train). This is a great example of when other sources of information such as descriptive assessments, direct observation, and interviews are helpful in providing more information to make the testing situations seem real or help with treatment decisions. Another limitation, is an FA may produce a temporary increase in problem behavior. The problem behavior may continue or happen more often if the assessment is ended before an appropriate behavior is taught.  To avoid this we end the analysis as soon as we have enough data to teach us something about the behavior.

Conclusion FBA and FA are assessments – they are ways of gathering information about behavior. Specifically, they help us identify what makes problem behavior likely in the first place, and what outcomes keep it going strong. Assessment includes everything we do to learn about behavior – analysis includes only direct tests of the causes of behavior.  Once we know what the function of behavior is, we can teach a better way to get that function met.  Both FBA and FA answer the questions: “why does my child do that?” and lead us towards figuring out “how do we change it?” These procedures are how behavior analysts and families quickly and effectively figure out how to teach kids what to do instead of problem behavior.

Remember – diagnoses don’t cause behavior. Sometimes people can be tempted to cite medical or psychological diagnoses as causes of behavior. Example: “Why does my child have social and communication deficits – because of Autism! How do I know my child has Autism? Because of the symptomatic social and communicative deficits!” This is circular logic, and actually doesn’t help identify anything we can treat that causes behavior. Diagnoses are a standard way of describing a set of symptoms. Diagnoses, in themselves, don’t cause anything at all.


Cooper, J. O., Heron, T. E., & Heward, W. L. (1987). Applied behavior analysis (2nd ed.). Columbus, OH: Merrill Pub. Co.Iwata, B. A., Dorsey, M. F., Slifer, K. J., Bauman, K. E., & Richman, G. S. (1994)

Toward a functional analysis of self-injury. Journal of Applied Behavior Analysis, 27(2), 197–209.

Mayer, G. R., Sulzer-Azaroff, B., & Wallace, M. (2014). Behavior analysis for lasting change (3rd ed.). Cornwall-on-Hudson, NY: Sloan Publishing.

Other Informational Links

By Allison Piper and Jennifer Chase