Tantrums & Numbers

Erica Kastan, MA, LMFT, BCBA

There are numbers all around us. From gas prices to knowing what channel number NBC is on. Each one of these numbers has a symbolic and important meaning to its own context. Last month I wrote about understanding the functions of behavior and briefly stated that data collection is needed to understand if you are hypothesizing the right function, baseline, and progress of behavior. There are many ways to take data in ABA: Frequency, duration, latency, andInter Response Time (IRT). Each one of these has its own merits, applications, and meaning. Knowing the one(s) to use for which behavior is the key. Otherwise the numbers have no meaning, like stating the gas price when someone is asking what channel NBC is on. Frequency is the amount of times a behavior happens (usually done with tally marks and over set period of time).Duration is the amount of time engaged in the behavior at one time. Latency is the amount of time from when told to do something, and when the person starts doing it. Inter Response Time (IRT) is the amount of time between sets of targeted behavior. Having an operational definition of what the behavior looks like, amount, length of time per each occurrence, space between each occurrence to be considered a new one, more on this next time of what the behavior you are tracking is) is essential to take accurate data. The correct data lets us know what the baseline is and if the interventions are working.  Using the wrong data set will not help and can hide what is really going on. Often data tracking a behavior may require using multiple data collecting methods. I have worked with many clients on tantrum behavior. 

I have worked with many clients on tantrum behavior. When taking data (during the initial set up) I will take multiple sets of data to get a baseline to understand the behavior itself. Once I place an intervention I stay with multiple sets, and here is why: If I only data track tantrum behaviors using frequency, I may see that my intervention has reduced form the baseline of 12 tantrums a day to 2 per day. That sounds great and the intervention is a success! However, the duration of those 12 was at an average 5 minutes each, and now the average of is 45 minutes each, is it so great? Is it better? Using both measurements allows us to see the full picture and where things are successful and where things still need to be adjusted. If I use the wrong data collection, the measurements are incomplete and meaningless. Some behaviors happen so frequently and have such a small time between occurrences (i.e., hand flapping, is each had flap one, how long between each up or done movement is a new one?), that frequency numbers would be impossible to collect and highly inaccurate that one would never know if it is getting better. So using duration (how long in total an event happens, and the space between events (IRT), to start the time collection for the next occurrence duration) is needed to understand the behavior and progress. Then by marking a set in duration as a primary collection, you get frequency as a secondary and IRT, and now have an average per occurrence and space between.  Latency is great for other behaviors and knowing if the person is starting tasks in needed amount of time (usually used along with frequency of prompts). IRT is great to see if one can learn to space out behaviors and delay targeted behaviors. Correct data collection takes lots of practice, and knowing how to collect the correct set for the behavior takes experience and guidance.