Chapter 4 - Reliability and Validity of Everyday Observations

Question 1

A measure of the accuracy of your observations.

Answer 1

Reliability

Question 2

A measure of whether you are observing what you are really interested in.

The correlation between ratings by outside judges and observations by trained observers.

Answer 2

Social Validity

Question 3

Getting the same value about an observation.

Answer 3

Repeated Observations

Question 4

Learning a new task for several days as it improves and becomes more

Answer 4

Stable level of behavior

Question 5

The agreement between two independent observers.

Answer 5

Reliability

Question 6

The agreement between two observers is evidence of…

Answer 6

Accuracy

Question 7

Used by behavior analysts when they can compare each observation of two observers. Most RIGOROUS form of reliability.

Cannot be used with OUTCOME or EVENT recording.
Can be used with INTERVAL or TIME SAMPLE recording.

Compares EACH observation of the two observers.

The percentage of the total number of trials that are agreements between two independent observers.

ie. every time pitcher throws a ball to the batter, we can observe whether the batter hits it well.

Answer 7

Trial Reliability

100%xA/(A+D) Multiply 100% times the agreements.
You divide by the sum of the agreements and disagreements. Agreements plus disagreements equal the total number of trials. Add the agreements and disagreements before dividing.

Question 8

Used by behavior analysts when they can only compare the total observations.

Answer 8

Frequency Reliability

Question 9

Every chance to observe a behavior.

Answer 9

Trial