What is the Difference Between Comparative, Component, Parametric Analysis and Nonparametric Analysis?
Having a hard time understanding the difference between Comparative Analysis, Component Analysis, Parametric Analysis, and Nonparametric Analysis? Let The Bitches explain to you, SNABA style.
Comparative Analysis: Looks at two different types of treatment in an attempt to determine which is more effective.
The Big Question: Which one worked better?
Which keyboard types faster, a T-Mobile Sidekick or a Blackberry?
Clinical Example: In working with a client whose behavior is primarily maintained by negative attention, the team implements planned ignoring vs. functional communication training (learning to recruit attention by saying, “excuse me”) and measures which is more effective in reducing challenging behavior.
Component Analysis: Dissects each part of a treatment package to determine exactly which piece is affecting behavior change (i.e., what we’re measuring).
The Big Question: Which part is the best part (or the worst part)?
You’re at a boujee Sweet 16 and recognize you hate everybody because they’re annoying and entitled. You start demanding that everybody leave. You yell at one person at a time to grab their bejeweled shoes and leave the building immediately. After you ask one specific person to leave, you recognize that they were the reason you were so annoyed— not all the other kids.
Clinical Example: The BCBA develops a treatment package including one antecedent intervention (e.g., noncontingent access to magazines), one functional communication training procedure (e.g., requesting time alone), and one response cost procedure (e.g., fining the client for every instance in which they make an inappropriate sexual remark). The team removes one part of the treatment package at a time (they may begin with removing the antecedent strategy, collecting baseline data, and then re-implementing the antecedent strategy and removing the functional communication training procedure) and continues to collect data to determine the most effective piece of the treatment package.
Parametric Analysis: Looks at different dosages, or amounts, of one treatment and how these dosages affect behavior change.
The Big Question: How much is best?
How much Adderall do I need to write this blog post? If I take too little, I may never even start writing. If I take too much, I may become hyper-focused on cleaning my keyboard with a toothpick. We basically use a parametric analysis to be like Goldilocks and find the bed that’s just right.
How many White Claws do I have to chug before I blackout?
Clinical Example: The team is working with a school district to implement a school-wide incentive system. In doing so, the team must determine how often students may exchange their tickets for a prize that remains effective to maintain target behavior. Too infrequently (i.e., once every 2 weeks), the students may quickly lose interest and engage in other behavior because the delay is too long. Too frequently, though, the “lure” of an incentive system may lose its luster, as children have access to it every single day and are more likely to grow satiated by the system.
Nonparametric Analysis: This is when we look at the effect on behavior when the independent variable (intervention) is present or absent.
After seeing a psychologist, you decide that Lexapro is in your near future. Constantly moody and feeling a bit on edge, this magic pill can supposedly take the edge off and make you feel alive again. For a straight month, you take your Lexapro religiously. Following this month and seeing an increase in your positive vibes, you decide to go #coldturkey on the Lexapro and completely stop taking it. The purpose in doing so would be to determine if you actually need it to maintain those positive vibes, baby.
Clinical Example: The clinical team decides to implement Greg Hanley’s “My Way” Protocol for 2 consecutive weeks. Following the client meeting a pre-set mastery criteria (i.e., using functional communication independently for 5 consecutive sessions), the team is instructed to remove the intervention and collect baseline data without the use of the intervention.
By Kayla Perry