Chapter 2 (Lecture): Observing and Recording Behavior

What are the ethical responsibilties that the APA follows?

1. Beneficence and Nonmaleficence: Benefit others and do no harm

• Harm is measured by the outcome

2. Fidelity and Responsibiity: Establish relationships of trust, uphold professional standards, responsibilities, and obligations

3. Integrity: Promote accuracy, honesty, and truthfulness

4. Justice: Fairness and justice for all, equal access, benefit, and quality

5. Respect of People’s Rights and Dignity

How is behavioral intervention evidence based?

• It is operationally defined

• Measurement occurs at baseline (before intervention) and repeatedly over time to assess progress and to determine intevention efficacy (before and after assessments)

• Intervention procedures are clearly specified (need to be reproducible)

What does it mean by behavior is “operationally defined”?

The behavior is turned from an abstract concept into something concrete, observable, and measurable for a scientific study

What is behavior?

Behavior is what people do and say

What are the 5 primary characteristics of behavior?

Behavior is:

1. Action

2. Multiple measurable dimensions

3. Observable by others

4. Produces an effect on the environment (what we can see in the external environment)

5. Is lawful (related to the environment)

Does behavior focus on overt behavior?

Yes, behaviorism, where a person’s behavior is influenced by the environment focuses on visible behavior

• Not private events

What are the measurable dimensions of behavior?

There are multiple observable and measurable dimensions

• Frequency (number of times the behavior occurs)

• Duration (how long the behavior lasts)

• Intensity (how much force is required)

• Latency (how long does the behavior take to start)

Depending on the behavior, you look at the dimension you are the most interested in!

Target behavior

Behavior identified for change

Behavioral excess

Behavior targeted for reduction

Behavioral deficit

Behavior target for increase

Controlling variables

Factors in the immediate environment that influence behavior

How do you operationalize behavior?

To turn a behavior into something observable, and measurable for a scientific study:

• You need to describe the behavior using active verbs

• Use objective language

• No unobservable internal states (emotions, intentions)

• No labels

How can behavior be assessed?

Behavior can be assessed using:

• Indirect assessment: questionaires and rating scales, interviews

• Direct assessment: live or recorded observation

What are the strengths/ limitations of direct assessment?

Direct assessment is good for:

• Giving context to the behavior

• Good for frequent unnoticed behaviors (wouldn’t be good to self record)

• Clear evidence and high accuracy

Limitations:

• Resource and time costly

What are the strengths/ limitations of indirect assessment?

Indirect assessement is good:

• Time efficient

• Convenient

Limitations:

• Subjective

• Suspectable to bias

What are key considerations for choosing an observer?

Observer:

• Either a professional or a person from the natural environment

• Proximity to observe (unless using video)

• Need sufficient time to observe and training

What are the recording considerations?

Need to decide:

• When to record (Select a period when behavior is most likely to occur)

• Where to record (Natural environment vs laboratory environment)

• How to record (Structured vs unstructured recording)

• Structured = options are given for activities/ directions are given, while unstructured, nothing is given

Whar are the types of recording procedures?

The types of recording procedures are:

• Continuous Recording (can be frequency, duration, latency, intensity, or percentage of opportunities)

• Interval Recording (Partial/ whole interval)

• Time Sampling

• Product Recording

Continuous Recording: Frequency

Continuously record each instance of behavior during the observation period

• Need to define each instance of behavior and break down the definition of a behavior for reliability

Calculate:

• Total frequency count

• Rate (especially important if observation periods differ in length)

Continuous Recording: Duration

Continuously record the duration of each instance of the behavior during the observation period

• Record the start and stop of the behavior

Calculate:

• Average duration of behavior

• Proportion of observation period with behavior

Continuous Recording: Latency

Continiously record latency for each instance of behavior during the observation period

• Time from the stimulus to when the behavior occurs (the amount of time that lapses before behavior occurs)

• How long does it take us to see the behavior?

Calculate:

• Average latency of behavior

Latency is very specific and only applies to some behaviors

Continuous Recording: Intensity

Record each instance of behavior continuously during the observation for intensity

• Look at intensity (magnitude + scaling)

• Good for: aggression, yelling

Percentage of opportunities or % correct for recording

Out of the total number of times someone has the opportunity to respond, how often do they respond?

• Percentage of times the person responds/ the number of times someone has the opportunity to respond

What are the strengths of continuous recording?

Strengths:

• You’re measuring every instance of a behavior

• More accurate (less measurement error) than discontinuous methods

• Best for discrete behaviors (clear onset and offset)

What are the limitations of continuous recording?

Limitations:

• May be less accurate when behavior has ambiguous start/stop or occurs at very high rates

• Requires significant attention, effort, and time

• May be difficult to record mulitple responses

What is discontinuous recording?

It is a behavioral observation method) where behavior is sampled in intervals rather than tracked continuously

• Observation period is divided into intervals of 20-30 s

• Only captures whether a behavior occured or not within that interval

What are the types of interval recording?

There is:

• Partial Interval Recording: Record if the behavior happened during each consecutive interval

• Whole Interval Recording: Record whether the behavior occured throughout the entire interval

Partial Interval Recording

Records whether the behavior occured at all during each interval

• Divide the observation period into intervals (reduces the load)

Calculate:

• Proportion (%) of intervals during which the behavior occured

What are the strengths of partial interval recording?

The strengths are:

• It’s efficient and behavior is recorded once regardless of frequency

• Helpful for low frequency behaviors

• Useful if the behavior reduction goal is 0

What are the limitations of partial interval recording?

The limitations are:

• It overestimates behavior (ex. 1 sec and 10 sec are equivalent occurences)

• Not a good choice when you want to increase a behavior that should happen very often, because it can make the behavior look more successful than it really is.

Whole Interval Recording

Record whether behavior occured throughout the entire interval

• For behaviors with high frequency

• Looking to see if the behavior occured for the entire duration

What are the strengths of whole interval recording?

It’s strengths are:

• It’s useful if the behavior is expected to have a long duration

• Infrequently used

What are the limitations of whole interval recording?

It’s limitations are:

• It underestimates behavior (1 s and 10 s response are equivalent non-occurences)

• Not recommended when targetting behavior reduction, makes it look like less happened

Frequency-Within-Interval Recording

Record the frequency of the behavior within each interval

• Provides frequency data and specific intervals within which behavior occured

• Not used as much because of time and labor costs

Time Sampling

Record behavior only during part of each interval or at a specific time in the interval

• Divide the observation period into intervals

Calculate:

• Proportion (%) of intervals during which behavior occured

• Can increase accuracy

Momentary Time Sample Recording

Behavior recorded only if it occurs at exact end of the interval

Strengths of Momentary Time Sample Recording

It’s strengths are:

• It may provide a more accurate estimate of behavior than partial or whole interval recording

• Recommended for use in targeting behavior increase

Limitations of Momentary Time Sample Recording

It’s limitations are:

• May have difficulty capturing low frequency behavior

• May be less sensitive to small intervention effects

• Not reccomended when target for behavioral reduction is 0

Data Recording Decision Tree

Product Recording

Indirect measure focused on the product of behavior

• Ex. Number of homework sheets completed, number of injuries produced by aggression

However, there is uncertainity about factors not directly observed (ex. did the child complete homework independently

What is reactivity?

It ishow peoplechange their behavior when they know they are being observed

• May reduce representativeness

• Behavior change is typically temporary

How do you reduce reactivity?

Reactivity is reduced by:

• Waiting until people are accustomed to observer’s presence

• Observe without the individual’s knowledge (with consent)

• Use video recording

• Reactivity can be beneficial and induced by self monitoring

Interrater (interobserver) Reliability

Level of agreement between 2 or more observers/ raters

• Ratings can’’t be valid if they are unreliable

• Percent agreement may be applicable with small samples