Accuracy Vs Reliability

Accuracy Vs Reliability

1. Accuracy

   • The extent to which a measure reflects or closely approximates the true value
   • A reliable measurement system must be used
   • The only way to know if data are accurate is to compare them to true values
   • 5 head hits and 5 count recorded

2. Reliability

   • repeated measurement of the same events yields the same result
   • Observing two seperate sessions and comparing their measures
   • Same system every time
     • must yield consistent data
   • AKA: precision

3. Reactivity

   • the presence of an observer influences behavior

4. Observer drift

   • over time observers modify operational definitions

5. complexity

   • complex data collection systems are prone to unreliable data collection

6. Expectancy

   • Bias due to preconceived notions about the behavior and/or person being observed

7. Believability

   • When steps are taken to ensure their accuracy
   • obtained by reliable measures
   • "Being convinced"
   • More believable when ACCURATe

8. Validity

   • A measure measures what it purports to measure
   • A person can put someone else's name on a test they took thus questioning the TEST's validity
   • Lesson plans are matched to the SOLs
   • If target is tacting apple, tacts of apple are recorded