Home > Flashcards > Print Preview
The flashcards below were created by user
on FreezingBlue Flashcards. What would you like to do?
What are the 3 reasons for assessing IOA?
- Consistency of measurement
- Minimizes observer bias
- reflects the quality of operational definition
Be able to calculate IOA using the frequency ratio method, Point-by-PointAgreement Ratio method (scoring agreements as occurrences and non-occurrences), andfrequency-within-interval method.
FREQ RATIO - smaller total/larger total X 100
Point By Point - Agreements/agreements +Disagreements X 100
Freq within interval (Disagreements-Total intervals)+(sum of all agreements when small agreement/large agreement)/total intervals x 100
how can you reduce the likelihood that a high percentage of agreements between two observers is the result of chace
For High rate bx define agreement as bx not occuring
For low rate bx define agreement as bx occured
report IOA on agreements and disagreements seperatly
Purpose of baseline and intervention
- Baseline -
- describe current level of Bx
- Predict future level if no intervention
- Test is Bx departs from baseline prediction
- Describe current level of Bx
- Predict future level if intervention remain in place
Describe reversal of behavior procedures
DRO for any bx other than target bx
problems with a reversal design
can not reverse learning
other variable may have control over target Bx
change agent may not be willing to reverse Bx
control of Bx could have transfered to different form of Sr
How to indentify the effect of an IV
all variables present during baseline are held constant during Tx
which experimental designs are most commonly intergrated
- reversals and multiple baselines
- baselines are assessed across multiple settings, participants, bxs
Changing criterion with reversal
Multiple tx with reversal
define partial and sequential withdrawal designs and their purpose
partial - slowly removing the entire tx from one baseline while no changes are made in the other to assess for generalization and transfer of stimulus control
Sequential - removing one component of a tx package to assess for maintenance requirements
how to maximize probe evaluations
no programmed consequences
When and why would a multiple probe technique be recommended?
When extended baselines are not possible because exposure would cause Bx to change before tx is implemented.
When implementing sequential treatment because it may not be possible to complete the later steps before the first step is mastered
what info can be gathered when conducting probe sessions throughout the course of an intervention applied to successive approximations
initial level of bx
what happens to bx during treatment
what happens to bx before training or treatment
what happens to remaining steps when prior step meets mastery
true baseline of bx
if using a multiple probe design in place of multiple baseline to avoid having to take continuous data that may result in bx change before interverntion, do BLs have to be concurrent and why?
Yes, baselines should remain concurrent so that time as an IV can be effectively eliminated as a confound
How is experimental control demonstrated in a multiple treatment design
by showing a difference in the paths of data in each treatment phase
Describe the purpose ofusing a multiple-schedule design.
to show that Bx is under control of stimulus conditions
What is multiple treatmentinterference?
- The effect of one
- treatment influencing the effect of another.
Two design modification that can enhance experimental control when using a multielement design
most effective treatment associated with all conditions in the final phase
conduct a mini reversal by applying the least effective treatment to the other conditions
Given that the inclusion of a BL phase is not a technical requirement of an alternating tx design, why is it that most alternating tx studies use a BL phase
incase there is no difference in the data paths of treatment
To enhance experimental control
to demonstrate the absolute effect of the intervention
If performance surpasses the criterion in a changing criterion design can a functional relationship be demonstrated?
Yes, through mant criterion changes and mini-reversals of criterion
what difficulties are associate with changing criterion designs that have unidirectional changes over the course of the intervention.
how can experimental control be enhanced
changes can be so small they dont approximate a stepwise change
changes may exceed criterion leaving room for alternative explanations for the change in Bx
Experimental control can be enhanced through mini reversals
clinical implications of bi-directional changes in a changing criterion design
Termination of the treatment may result in a return to baseline performance.
additional steps to transfer stimulus control is neccessary
What is the problem with having too many shifts in criteria (changing criterion)
- Changes in criterion are too small
- may need reversal to enhance experimental control
may not have enough time for each phase to stablize
three guidelines in steps or changes in criterion
change should be gradual
changes should be large enough so that changes in Bx are detectable
Cirterion changes do not have to be equal
what is meant by the statement there is a general relationship between the variability in the clients performance and the amount of change in the criterion that may need to be made
The variability in performance should guide criterion selection
- high variability - large magnitude
- Low variability - small magnitude
Distinct characteristics of skinner approach to single case research
- one or a few participants
- freq as a measure of data
- visual inspection
what is wrong with using group design to investigate clinical or therapeutic interventions
group designs mask individual differences
therapeutic interventions are often conducted 1 on 1
group designs not conducted in natural environmet
two considerations when intervening on problem bxs
intervention is socially acceptable
intervention and outcome is age appropriate
two social validation methods
social comparison - gathering normative data from peers
social validation - gathering information from experts (teachers, doctors, parents)
define internal, external and construct validity
Internal - manipulation of IV is responsible for changes in DV
External - outcome of intervention is generalizable
Construct - understanding the underlying mechanism responsible for changes in the DV
how to minimize reactivity
train observers to ignore extraneous variables
having observers take treatment integrity and IOA data
minimize cues associated with target Bx
when would sample session records differ from whole session records
when the target bx is low frequency
when sample session is short and whole interval is long
when the duration of target bx is short
describe the characteristic primary measures used to evaluate performance must have?
- Administered repeatedly - can be administered continually over time
- Consistency of measurement - minmal error
- Capacity to reflect change - must show changes
- Dimensional Scale - not binary (yes, no, complete, non complete)
- Relevance of measure - should assess problem directly
- Importance of measure - should be socially important
What are the 5 points related to validity ofmeasurement?
single measure does not convey the complete construct
observations are limited to specific situations(Will behavior be the same in less structured environments)
Does the measure assess something important
the human filter can alter data
Is the change observed important
What are the critical differences between discrete categorization and frequency measures?
Freq - single response performance measures are tallied- no real time limit
Discrete - several Bxs included- limited # of opportunities to perform
Describe the two main versions of time-basedmeasurement.
Interval recording - bx observed for a single block divided into several small blocks of time per day
Response Duration- the amount of time the response is performed, good for continuous rather than discrete bxs
Separate groups are used in research when one is interested in evaluating different levels of variables that are introduced initially to the subject(s) who have not previously been exposed to any experimental phase. Can you think of another way this couldbe accomplished, other than separating participants into two differentgroups?
You can use the same groups just randomize the order of treatments
Why is steady state responding not necessarilyindicative of the independent variable being applied accurately?
The fallacy of affirming the consequence
If the IV is functionally related to the DV then the DV will have a steady state
DV responding is steady thus the IV must have been applied accurately
Explain why anticipated changes in behavior when the treatment variable is applied do not necessarily indicate a robust functional relationship.
because the experimenter must prove that those changes in the DV are a product of IV manipulation and nothing else
Explain why discontinuous sampling methods may be contraindicated for small treatment effects.
because it increased the likelihood of false negative interpretations of a functional relations
How can a functional relationship be demonstrated using a non-concurrent multiple baseline
use more baseline phases
number of datapoint during baseline
When using a multiple baseline design across behaviors or settings, what mightbe a risk of keeping one behavior or setting in baseline for an extendedperiod of time? How do you know when to intervene with the secondbaseline?
When the data is stable. No trends
What does interdependence of baselines refer to?
changes in one baseline effect bx in other baselines
When baseline phases are extended for a prolonged period, performance maysometimes improve even before the intervention has been applied. What aresome reasons that may account for this?
interdependence of bxs
learning may have occured - changes do to observation or direct practice
social environment changes based on changes to Bx in the first intervention
How can problems with prolonged baselines be avoided?
staggered baseline assessment
What are the requirements for making comparisons across baselines in multiplebaseline designs? Explain the rationale for each requirement. (Will discuss further in class)
each baseline must be independent
demonstrate sensitivity of second baseline to the intervention
When intervention is ongoing in one baseline the control condition must be ongoing in the others
overlap between control and intervention must be long enough to allow extraneous variables associated with the IV to emerge during control
extraneous variables must be present in treatment and baseline conditions
1. Whatare quasi-experimental designs and why are they important?
- a. Designs
- in which the condition of a true experiment is approximated and or post hoc
- b. They
- allow for the evaluations and inferences to be drawn from programs that cant
- afford costs or the time associated with experimental rigors
1. What are the 5 ways in which we can remove ambiguities from quasi-experimentaldesigns?
- a. Collect
- systematic data
- b. Assess
- behavior on multiple occasions
- i. Looking
- at behavior across intervals (ie. months)
- c. Consider
- past and future projections of performance
- d. Consider
- the type of effect associated with treatment
- i. Looking
- for immediate and dramatic changes associated with treatment
- e. Use
- multiple and heterogeneous participants
- Replication of the
- change in behavior across multiple different individuals
1. Whatdoes the evaluation of treatment procedures in the quasi-experimental designinvolve?
a. Listing all known threats to the validity of the interpretation of the effects of the
treatment, decide whether they are plausible as explanations
1. What isan interrupted time-series design?
- a. A
- series of observations at time removed from the immediate initiation of the
- treatment both before and after