Experimental Design Flashcards Preview

Cooper, ABA > Experimental Design > Flashcards

Flashcards in Experimental Design Deck (78):

Experimental Control

When a predictable change in behavior can be reliably produced by the systematic manipulation of some aspect of the environment.


What are four important elements of behavior regarding exeperiments?

Behavior is individual, continuous, determined and extrinsic.


Can groups have behavior?

NO. However, each individual can have behavior.


What do you do when you see variability in data?

Attempt to experimentally manipulate what is suspected to be the cause.


What is an experimental question?

Brief, specific statement of what researcher wants to learn


What are six components of Experiments in ABA?

1- At least one subject
2- At least one behavior
3-At least one setting
4- At least one treatment (IV)
5- A measurement system and ongoing data
6- Experimental Design


What does "Single Subject" mean?

That the subject acts as his own control. The person is exposed to the condition several times.


What is a collateral Effect?

When multiple DV are measured to assess the IV's effects on behavior that is outside of the given response class.


What are extraneous variables?

Unplanned environmental variations


What do you do when there are extraneous variables?

Wait them out or incorporate them into the design.


What are the two types of Experimental Design?

Nonparametric Analysis and Parametric Analysis


What is Nonparametric Analysis?

The Independent Variable was either present or absent during study.


What is a Parametric Analysis?

The value of the independent variable is manipulated. This looks at the differential effects/range of values.


What is one important Rule of Experimental Designs?

Change one IV at a time


What is a treatment package?

When multiple IV's are grouped into one intervention


What is a Component Analysis?

Looks at the effectiveness of each IV in a Treatment Package


What is Steady State Responding?

Pattern of responding that shows little variation. Provides the basis of Baseline Logic.


What is Baseline Logic?

Prediction, Verification, Control


Steady State Strategy is:

Repeated exposure of a client to a condition to get rid of extraneous variables.


What is the function of baseline?

It serves as Control.


Does baseline mean the absence of an intervention?

No. However, it is the absence of a single IV.


What are some benefits of baseline data?

It can create a guide of initial criteria, to see if the DV really requires change, basis for objective illustration of change.


What are four patterns of Baseline data?

descending, ascending, variable and stable.


What does a descending or ascending baseline show you?

That the bx is already changing.


If Baseline data is descending or ascending, should introduce the IV?

No. Unless the graph shows that behavior is getting worse (going down when you want it up, vice-versa).


What does a variable baseline mean?

There is no clear trend. The variability is probably due to environmental factors.


What do you do if a baseline is variable?

Wait it out ad try to control for the extraneous variables.


What does stable baseline data tell you?

There is no ascending or descending trend. This is a perfect way to look at the IV effects on the DV. You can intro the IV!


What is prediction?

The anticipated outcome of unknown measurement.


How many data points do you need in order to practice prediction?

There is no magic number. But, the more, the better.


What is affirmation of the consequent?

Your prediction means that, without the IV, the data would look the same. The process of Identifying if the IV had a definite effect on the DV (the data will show a change).


What is verification?

Verifying the predictions by withdrawing or reversing the effects. The data should mimick baseline


What is replication?

Making it happen again!


What are the five types of experimental designs?

Multiple Baseline
Changing Criterion


What is the most widely used design?

Multiple Baseline


What is a multiple baseline Design?

staggered implementation of IV, step-wise across behaviors, settings and people.


If it's unethical or impossible to reverse conditions, which design should you use?

Multiple Baseline Design


How do you show a functional relation with Multiple Baselines?

Apply the IV to Baseline one (once stable). If baselines 2 and 3 do not change, then you are closer to your prediction. Then Introduce IV to Baseline 2. It should have the same effect as Baseline 1. And so on with BL3


How many tiers are usually in a Multiple Baseline Design?



What does a Multiple Baseline Design across behaviors look like?

Multiple behaviors are targeted for one person.


What does a Multiple Baseline Design across settings look like?

One behavior is targeted in multiple settings.


What does a Multiple Baseline Design across subjects look like?

One target behavior across multiple people.


Which type of Multiple Baseline Design is the most widely used?

Across subjects


What are two variations of Multiple Baseline Design?

Multiple Probe Design, Delayed Multiple Baseline Design. They are both weaker than the other four.


What is a Multiple Probe Design?

Instead of simultaneous baselines, there is a simple probe to see if behavior has changed during baseline.


What is a Delayed Multiple Design?

When the subsequent baselines and IV's are delayed.


What are 5 guidelines regarding Multiple Baseline Design?

1. Select independent and functionally similar baselines
2. Select concurrent and related multiple baselines
3. Do not apply the IV too early
4. The more baselines differ in length, the stronger the comparison.
5. Intervene on the most stable baselines first


What are some advantages of multiple baselines?

The successful intervention does not need to be removed, easy to implement and can evaluate generalization


What are some disadvantages of multiple baseline designs?

Functional relationship is not directly shown, Function of behavior is not given, It take proper resources to implement correctly.


What is Changing Criterion Design?

After baseline, the intervention consists of multiple phases where the criteria for R+ or P is raised.


How is a functional relation illustrated with Changing Criterion?

The extend the level of responding changes to conform to each new criterion. The greater the vertical distance between criterion lines, the more experimental control there is.


What are 3 guidelines for Changing Criterion Design?

1. Length of Phases: The length of the phase must show steady responding
2. Magnitude of criterion changes: The size of change between criterions should change, to illustrate a true functional realtion
3. Number of Criterion changes: the more # of changes, the better proof of experimental control.


What are some advantages to Changing Criterion Design?

Does not require reversal, Experimental analysis is achieved while gradually improving behvior


What are some disadvantages to Changing Criterion Design?

The target bx must already be in the client's repertoire, Not a comparison design and not appropriate for shaping.


What is the difference between shaping and Changing Criterion Design?

Shaping is a behavior change strategy, Changing Criterion Design is an experimental design. Shaping deals with novel bx, CCD deals with bx already in a repertoire.


What is a Reversal Design?

The level of responding is REVERSED to that of a previous condition. This can include an alternation between baseline and intervention.


How is a functional relation illustrated with Reversal Design?

Each reversal illustrates more experimental control.


For Reversal to occur, the level of behavior must do what?

Approximate that of baseline.


What are the three required phases of Reversal?

B- Intervention
A- Return to Baseline


What are 5 types of Reversal Designs?

1. Repeated Reversals
2. BAB
3. Multiple Treatment Design
4. NCR Reversal Technique
5. DRO/DRA/DRI Reversal


Would you conduct a reversal design on a client that had unsafe behavior?

NO. Immediately start providing treatment.


What are repeated Reversals.

continuing to repeat the reversal and baseline phase.


What is a B-A-B Design?
When is it best to use it?

Phase 1: IV (B)
Phase 2: IV removed (A)
Phase 3: IV reintroduced (B)

This is the best design when the client is displaying severe and dangerous bx. Appropriate when IV is already in place.


What are sequence effects?

Effects on a client's behavior because of exposure from a previous condition


What is a Multiple Treatment Reversal?

A reversal design that includes multiple treatments that are being compared to baseline or to each other.

Major disadvantage: Sequence Effects


What is Non-Contingent Reinforcement (NCR) Reversal Technique?

NCR is provided as a control phase, versus using it as baseline.


What is DRO/DRI/DRA Reversal Techniques?

Using DR+ as a control versus a baseline


What are advantages of Reversal Designs?

Clear demonstration of functional relations.


what are disadvantages of Reversal Designs?

Irreversibility!! Ethics (dangerous behaviors)


What is irreversibility?

The level of behavior from a previous condition cannot be replicated (i.e. riding a bike).


What is an alternating Treatment Design?

Two or more conditions are presented in rapidly alternating succession


Alternating Treatment Designs are based on what?

Stimulus Discrimination. Each intervention has a clear Sd.


How is a functional relation illustrated with Alternating Treatments?

One data path is clearly higher than the other. No data points overlap.


What are three variations to alternating treatment Designs?

Single Phase w/o Baseline
With Baseline
With Baseline and Final Best Treatment Phase


What three problems are avoided by using alternating treatment designs?

Irreversibility, Sequence events, unstable Data


One disadvantage of Alternating Treatment Designs is "Multiple Treatment Interference". What is that?

When there are multiple treatments going on at once.


What is a withdrawl design?

Same as Reversal.


What are three practical and ethical guidelines to demonstrate treatment effectiveness in single-case experimental designs?

1- Baseline Trends: wait for stable baseline, rule out extraneous variables or choose a method that does not require stable baseline.
2. Excessive Variability: Search for causes in the variability
3. Duration of Phases: enough data to create a stable level, obejctive.