Design Basics
DICE
1. Behaviour is Determined - it's determine by the functional relation it holds to other events
2. Behaviour is Individual - based on a person's interaction with their environment
3.Behaviour is Continuous - it changes over time and requires constant evaluation
4.Behaviour is Extrinsic - behaviour change/variability is the result of the environmental changes not some internal unseen process
Single subject designs means that there is only one subject per experiment
False - it's called single subject because the subject acts as their own control condition.
4-8 subjects per experimental design is actually more common
A pattern of responding that exhibits very little variation in its measured dimensional quantities over a period of time is called
There are 5 different types of multiple baseline designs. List all 5 types and variations
1. Multiple baseline across behaviours
2. mutliple baseline across settings
3. multiple baseline across subjects
4. multiple probe
5. delayed multiple baseline
This design should be used when it is unethical or impractical to reverse conditions or when the behaviour is irreversible
Multiple baseline design
This is a brief but specific statement of what the research wants to learn from conducting an experiment
What is the research question
A collateral effect is when the IV effects behaviours other than the targeted behaviour
TRUE
* investigate if other people's behaviour has changed which influenced the subject (ex. experimenter's behaviour)
* reason we measure multiple DVs sometimes
1. Prediction
2. Verification
3. Replication
You would use the weaker variations of multiple baselines (multiple probe and delayed multiple baseline) when
1. extended baseline measurement is unnecessary
2. its impractical to collect extended baseline measures
3. its too costly to do so
4. extended baselines were not available (i.e. a subject joins the experiment after it has begun for other subjects)
This experimental design is recommended if your client is displaying severe or dangerous behaviour and you can not collect baseline data before providing an intervention
B-A-B reversal
List the 6 components of an experiment in ABA
1. at least one subject
2. at least one behaviour
3. at least one setting
4. at least one treatment
5. a measurement system and ongoing analysis of data
6. an experimental design
Baseline data implies the absence of the intervention
False - it can be the absence of a specific IV but it does not mean that no intervention at all is in place - it simply services as a control condition for the IV your manipulating
List 3 benefits of baseline data
1. helps to set the initial criteria for reinforcement
2. confirms if the behaviour targeted for change really needs an intervention
3. it provides an objective basis for detecting changes from the IV
4. obtains descriptions of ABC correlations for the planning of an effective treatment
List two advantages of a multiple baseline design
1. successful interventions do not have to be removed
2. evaluates generalization
3. easy to implement
One of the main ethical concerns with a reversal design from an ethical stand point is
That you're potentially removing an effective IV for the sake of experimental control - this has no social validity in school and home settings
The behaviour selected for change is the ________ variable while the intervention/treatment is the _________ variable
* dependent
* independent
Nonparametric analysis deals with the value of the IV and it's manipulation. It seeks to discover the differential effects of a range of values
False - parametric analysis does this
What are the 4 patterns of baseline data
1. Descending Baseline
2. Ascending Baseline
3. Variable Baseline
4. Stable baseline
List two disadvantages of multiple baseline designs
1. does not demonstrate a functional relationship
2. does not provide information on the function of the target behaviour
3. IV may be delayed for certain behaviours, subjects or settings
If there is excessive variability in your data how could this be addressed
1. block consecutive dat points and plot blocked averages rather than day-to-day performance
2. search for causes of the variability or the situation (eg. variation among environmental stimuli)
What should an analyst do when there is variability in the data?
* attempt to experimentally manipulate the suspected factors causing the variability (causal factors)
* seek treatment variables that are robust enough to overcome variability
A component analysis involves looking at the effects of each part of a treatment package one at a time to determine the effective components of a treatment package
True - you want t keep the effective components going and eliminate the ineffective ones to make the intervention more acceptable to the mediators and efficient for the learner
You should collect this minimum number of baseline data points _____ before implementing the IV
When intervening on a behaviour/subject or setting in a multiple baseline how do you know which behaviour/subject or setting to start with
Select the behaviour/subject or setting with the most stable baseline first.
List two ways to address increasing or decreasing trends in your baseline data
1. continue observations for a longer period of time
2. try to reverse the trend (eg. using a DRO schedule of reinforcement)
3. select designs that do not require a stable baseline
4. use statistical techniques to take initial trends into account