How do we know if the behavior we measured change across time? We display it visually. We have two broad choices for time-series visual graphics: arithmetically scaled and ratio scaled line charts. Each chart displays the behavior and can tell us if the change stays the same, goes up, and goes down. Beyond that we will find major differences.
Before I get into major differences let us do an analogy. Pretend you have a child who needs medical treatment. You go to a doctor who offers a diagnosis of the problem with two levels of information. Version one diagnostics has very basic information and uses qualitative (i.e., words) descriptors. For example, the doctor says your child’s pulse is very fast. Your child also has a high temperature. Furthermore, your child has abnormally high respirations. That’s all the information you get. Version two diagnostics has descriptive and wonderfully quantitative (i.e., numbers) information. The doctor says your child’s pulse is very fast at 110 beats per minute. Your child also has a high temperature, 101 degrees. Furthermore, your child has abnormally high respirations, 27 breathes per minute. Which version of information do you want the doctor to use to figure out what is going on with your child? Is there any parent on planet Earth that would pick version one diagnostics? Any parent that wants the doctor to use less information rather than more? Is there?
Let us now bring this analogy closer to home. Suppose you have a child who needs educational, behavioral, or psychological treatment. You go to a teacher, behavior analyst, or psychologist who offers a diagnosis of the problem with two levels of information. Version one diagnostics has very basic information and uses qualitative (i.e., words) descriptors. For example, the teacher/behavior analyst/psychologist says your child’s reading behavior is very poor. Your child also makes many mispronunciations. Furthermore, your child recalls very little information after reading a passage. That’s all the information you get. Version two diagnostics has orders of magnitude more descriptive and quantitative (i.e., numbers) information. The teacher/behavior analyst/psychologist says your child reads text at 57 correct words per minute. Your child also mispronounces words at a rate of 8 word per minute. Furthermore, your child can only retell 2 thoughts units per minute after reading a passage for two minutes. Which version of information do you want the teacher/behavior analyst/psychologist to use to figure out what is going on with your child? Guess what version our education system, discipline of behavior analysis and psychology use? If you guessed the more basic, version one form of diagnostic information you are sadly correct.
Why do we find ourselves in this state of affair? I will attempt to answer that complicated question in time. But in the present, I tell anyone who will listen to me we must, must use the version two level of diagnostics. That is, let’s use a visual display that provides numbers and understandable comparisons of the data (i.e., Standard Celeration Chart) versus the most basic, and prone to provide misleading rate-of-change information, visual graphic (i.e., nonstandard arithmetically scaled chart).
The version almost every single teacher, behavior analyst, and psychologist uses: nonstandard arithmetically scaled line chart. For example, I just completed an extensive survey of behavior journals and found out of 2039 graphs 99.997 were of the arithmetically scaled variety. But you don’t have to take my word for it, open up any journal and see for yourself (of course don’t pick the Journal of Precision Teaching and Celeration 😉
Below please see a table from The Precision Teaching Book. After reading the table, which visual graphic provides the most information? Which visual graphic will make the most difference for teachers, behavior analysts, and psychologists?
The measurement and visual display landscape of education, behavior analysis, and psychology needs to change because every single person whose data is scrutinized deserves the best, most ethical intervention possible. The visual display filters and tells the chart reader what is happening, and it does so through the lens of the selected visual display (e.g., nonstandard arithmetically scaled chart, Standard Celeration Chart). As a rational, concerned consumer or deliverer of educational, behavioral, or psychological services, let’s work to make the SCC the preferred visual display system for time-series data. Our connection to the data, and the people whose behavior is measured, will dramatically improve. In the end, education, behavior analysis, and psychology are founded on the principle of helping people grow. Shouldn’t we use the most informative visual display possible to analyze, interpret, and communicate data?