The Function of Our Behaviors
When we look at patterns of behavior, they can for the most part fall in to four categories: Access, Escape, Attention, and Sensory. Why do we do what we do? And why do we do it again? The study of the functions of human behavior, and the factors that make up the modern science of behavior analysis are could fill a library. This is just a brief glimpse in to the topic of functional relations, and the four functions.
Before we get into that, let’s get into this.
Have you ever heard of the choice-supportive bias? The gist of it is, we make a choice, and then later ascribe either more positive or negative rationalizations to it as though it were the actual reasons for the choice itself. We, in essence, attribute a more complex reasoning process to our choice for the sake of a narrative. This is a process that takes place after the behavior occurred. Great for imaginative and alluring story-telling, but not great when we want to actually understand what maintains patterns of what we do. 
In order to study these maintaining factors without a bias like this, we use what’s called a Functional Behavior Assessment. A Functional Behavior Assessment (FBA) is used to formulate and support theories on the ongoing “functional relations” between behaviors and their environment. A functional relation is a term used to describe that when one event or stimulus (antecedent) occurs, a behavior is more (or less) likely to follow it than it would be random. This is based off of past history of reinforcement (reward). [1,2,3] Behavior is adaptive, it evolves as we gain experiences. One doesn’t just “grow up” and get handed a book of things that we then do; it is a process of behaviors in our past that have worked or not worked for us that gives us the repertoire we have today. Certain medical conditions or physical disabilities exhibit behaviors that are not directly controlled by these functions. Ruling these out first is the best idea before looking specifically at a functional analysis. 
Let’s go with an example:
Bob walks to the store. He sees an orange on the shelf, picks it up, pays for it, and gets to eat that orange. The next day, he also walks in to the same store, looks for an orange, and goes through this same process again.
A choice-supportive bias explanation might be that Bob really loves oranges, and that he knew the juiciest most amazing oranges come from that store, and he really loves the people in that store because they are so friendly when he pays, and he saw his friend John there the other day and John is from California, which must mean he remembered something about oranges, etc, etc, etc.
Does that help explain the base question? Maybe. Is it something we can reliably test? Not very easily, and even if we did, many of those factors might be superfluous. In the most common practices of behavioral science, private events (thoughts, feelings, etc) exist, but they provide poor data and are actually subject to interpretation (self-report biases, attribution errors, poor recall, untruthful responses, etc), and are moderators rather than causal. [1,2] We rely mainly on observable events. Setting. The environment and past experience appears to play the greatest role on what we do and why we do it. It can explain these patterns more precisely than recollections which may be subject to bias. [1,2.3] Looking at this scientifically, we want a pattern. Patterns and replication are very helpful when producing theories. [1,2,3]
All we need for the sake of this pattern are 3 things:
The Antecedent (what happens before the behavior): Bob enters the store, and he looks at the orange.
The Behavior (what did Bob do?): Bob purchases that orange.
The Consequence (what did Bob get?): Bob eats that orange.
From these three things, we can see the first step of the pattern. If he returns to that store, and he sees a similar orange, and buys that orange, we can reasonably say that his behavior was reinforced (rewarded). Reinforcement is an effect on future behavior. It follows an initial behavior and increases the future likelihood of it occurring in similar situations (antecedent settings). The consequence? That was the reinforcement step. Bob ate the orange, and then pursed that same experience the next day or week. If he did not go back to that store, we could say that the behavior was not actually reinforced to any observable effect.
Seeing these functional relations, we can pick apart the factors that hold us to our specific likes or dislikes, or why we do certain things in certain ways. They can help us change these patterns to a more socially appropriate, or even more functional alternative. An example might look like this, if Bob robbed the store instead, he would still access and be reinforced by the orange. Buying the orange was the “better” alternative and satisfied the same function.
Access, Attention, Escape, and Sensory:
Let’s take a look at these four classes of maintaining functions.
Access: This pertains mainly to tangible items. Items we can get. This could be food, vehicles, money, etc. These are things that you can see with your eyes and interact with.
Attention: This pertains to an interaction, praise, or even a signal, from others. This could be a look, a conversation, etc. Any response or increase of awareness from others.
Escape: This function relies on removal of an aversive item, situation, or stimulus related to a prior event which underwent punishment. Moving away from fires, for example. Avoiding interaction with specific people. This function is to get away from something, or avoid it entirely.
Sensory: This function, unlike the others, does not rely on outside factors. These are internal rewards from the body itself. Physical contact, for example.
Are they exclusive? Can’t a behavior be maintained by more than one?
Absolutely, and that is more common than behaviors specifically maintained by individual factors.  Let’s take Bob for example, and his orange. You could hypothesize from what we have observed already that his orange buying behavior could be maintained by both Access, and Sensory. Going to that store, he gains access to oranges. While he eats it, it may provide reinforcement through sensory means as well (taste). The hypothesis is the first step of many. Testing each factor, or observing changes in the environment which has an impact on that behavior’s use in the future would yield more and more accurate representations of that functional relation between behavior and environment. [1,2]
Other things to consider.
There is a factor that also decreases the future probability of behavior. It follows the same rules as reinforcement, except that the behavior does not become a stronger habit or pattern, it lessens. In the same way that Bob may go to the store to get that orange, and has been for weeks, one day he might pick up a spoiled orange, and eat that orange. That would have an effect on this pattern of behavior. Instead of strengthening his pattern of buying and consuming oranges, it would either lead to an adaptation of his behavior (checking the orange for spots, for example), or reduce his visits to that store entirely. These are factors that we could track if we were keeping a log of how the setting, or environment, influences his behavior.
Can you think of any examples you could fit in to this paradigm? How about some things that you have trouble fitting? Reach out! Let’s explore it.
 Iwata, B. A. (1994). Functional analysis methodology: Some closing comments. Journal of Applied Behavior Analysis, 27(2), 413-418. doi:10.1901/jaba.1994.27-413
 Hanley, G. (2012). Functional Assessment of Problem Behavior: Dispelling Myths. NCBI.
 Oliver, A. C., Pratt, L. A., & Normand, M. P. (2015). A survey of functional behavior assessment methods used by behavior analysts in practice. Journal of Applied Behavior Analysis, 48(4), 817-829. doi:10.1002/jaba.256
 Mather, M., Shafir, E., & Johnson, M. K. (2000). Misrememberance of options past: Source monitoring and choice. Psychological Science, 11, 132-138
Photo Credits: http://www.unsplash.com