Preparing a Graph

Preparing a Graph

• Before graphing data, ensure data accuracy.
• Inaccurate data makes graphs meaningless and unreliable.
• It won't show if a student is learning or making progress.

Reliability vs. Validity

• Reliability: Two observers agree on whether a behavior occurred.
  • If both agree the behavior happened, data is reliable.
  • If there's disagreement, data is unreliable, indicating issues with observational definitions.
• Validity: Data represents actual behavior change.
  • Important to show actual change occurred.
  • Example: Child initially leaves room 20 times a night, then 0 times.

Reliability and Validity Differences

• Data can be reliable without being valid (no behavior change).
• Unreliable data requires action – usually tightening the operational definition.

Training on Measurement Systems

• Essential for RBTs, especially when working with children.
• Taking data during training videos is different from real scenarios.
• Practice data collection with a supervisor role-playing as a child.
• Gradually escalate behavior expectations during practice.
• Real-world data collection is challenging and requires practice, especially in group settings.

Observer Drift

• Changing definitions leads to scoring differently.
• Example: Changing the definition of "joining in" during playtime.
  • Initially it was going up and playing with the kids.
  • It changes to standing next to the kids.
• Another Example: Marking "bah" as incorrect initially, then counting it as correct.
  • In expressing language, the kid needs to say "ball", but he says "bah".
  • initially it was counting as wrong.
  • But observer drift it starts counting is right.
• Observer drift leads to inaccurate and invalid data, affecting program decisions.
• It affects the learner. You engaged an observer drift.
• Programs might be stopped prematurely or prompting systems faded out incorrectly.

Student Reactivity

• Learners might change behavior when observed (e.g., cameras or new staff).
• They might be excited or revving down in behaviours.
• Behavior might not represent typical behavior, affecting intervention decisions.
• To avoid: Introduce cameras/observers gradually before scoring.

Inter-Observer Agreement (IOA)

• Supervisors should observe RBTs to ensure data reliability.
• It important because if you are both using continuous measurement systems if you guys are unreliable, then all the data is not valid.
• Unreliable data indicates training opportunities on operational definitions, whether continuous or discontinuous.
• If data's reliable, it shows proper data collection.
• IOA can also validate estimation data.
• Encourage RBTs to seek IOA with supervisors or fellow technicians.

Treatment Fidelity

• Assess whether procedures are implemented correctly.
• Hard to collect on clinical judgment due to varied thought processes.
• In discrete trial training, it can be assessed if the trial is implemented one time, you varied your SDs. if you closed out your trial.

Selecting Data and Measurement System

• BCBA or supervisor determines the measurement system.
• RBTs can provide input based on their experience.
• Depends on multiple variables:
  • Funding sources
  • IEP goals
  • Group setting
  • Aberrant behavior levels
• Balance accuracy with quality teaching.
• Prioritize quality teaching if a balance can't be achieved.
• Use a combination of estimation data in some sessions and continuous data collection procedures in other sessions.

Preparing for Data Collection

• Know the operational definition.
  • Clarify any ambiguities or loopholes in the operational definition with the supervisor.
• Understand data collection procedures.
  • Know time intervals for momentary time sampling and partial interval recording.
• Implement procedures accurately.
  • Never put yourself in the position to collect data that you don't know.
• Seek training if needed.
• Understand data collection procedures for every goal because data collection changes based on the goal.
• Have materials ready (pens, paper, erasers).
• Include name and date on data sheets.
• Include child identifying information (number, pseudonym, or initials) following HIPAA regulations.
• Include session information.
• Adequate training is crucial for fluency and accurate data.

Watching the modules is not enough you need the practice in order to make you fluent in it once you can do it to a high degree of treatment finally then that data is going to be accurate then that data is going to be usable which can help you train. you can use a data to graph.