Use LEFT and RIGHT arrow keys to navigate between flashcards;
Use UP and DOWN arrow keys to flip the card;
H to show hint;
A reads text to speech;
41 Cards in this Set
- Front
- Back
|
*Primary goals: - Demonstrate functional relation between IV & DV - Functional relation said to exist when a change in antecedent/consequent stimulus class consistently alter a dimension of a response class |
Experimental Design Goals |
|
|
* Systematic manipulation includes - Repeated and systematic presentation & removal of the IV - While measuring the DV (some aspect of the behavior of interest) - And holding other factors constant |
Experimental Design Qualities |
|
|
*Relative to group designs they permit investigation of behavior change as a dynamic process *Repeated measurement & stability criteria mean you need to keep observing *Permits you to see patterns of behavior change *Taking single measurement obscures that process *Allows examination of intersubject variability *Group data may not reveal anything about the performance of any individual *Participants serve as their own control |
Experimental Design Advantages of Single Case Designs |
|
|
*BL phase followed by a tx phase *An effect is demonstrated by showing that bx changes from one phase to the next *By itself, supports only weak conclusions - Changes in bx may be the result of extraneous variables - Recommended only when other, more compelling designs are untenable *A-B arrangements form the basis for all other common single case designs |
A-B Design |
|
|
*Focuses on the implementation of basic principles to change behavior based on the significance of clients *Socially significant of immediate importance to clients |
Dimensions of ABA: Applied |
|
|
*ABA focuses on behavior in its own right as a target for change *Behavior is directly observed and measured in real-life environments |
Dimensions of ABA: Behavioral |
|
|
*ABA seeks to identify functional relations between manipulated environmental events and behavior through systematic & controlled manipulations |
Dimensions of ABA: |
|
|
*In ABA procedures are completely identified & precisely described & defined |
Dimensions of ABA: |
|
|
*Procedures are linked to & described in terms of the basic principles of behavior |
Dimensions of ABA: Conceptually Systematic |
|
|
*An accountable discipline & changes in procedure are data-based *Treatment protocols are revised based on data. *If the plan is not effective it is revised until we get the desired effect |
Dimensions of ABA: |
|
|
*Generality of behavior change *Behavior changes achieved should maintain, transfer to other settings & spread to other behaviors. *Procedural generality *Discovers procedures that can be applied to many persons & in many settings |
Dimensions of ABA: Generality |
|
|
*Experimental design in which the researcher attempts to verify the effect of the independent variable by "reversing"responding to a level obtained in a previous condition *Following BL the IV is introduced and then withdrawn at least once (ABA design), introduction and removal of the IV 2 times (ABAB design) may be more common *If bx changes systematically as a function of the introduction and withdrawal of the IV the likelihood is small that an extraneous variable produced the bx change - The likelihood decreases with each subsequent withdrawal and introduction of the IV *If intervention is immediately critical, analysis may begin with an intervention phase - Does not alter the logic of the design |
Reversal/Withdrawal Design |
|
|
*Most straightforward single-case arrangement *Most powerful demonstration of functional relations |
Advantages of Reversal Design |
|
|
*Reversibility - Some behavior changes are not reversible i.e. Skill acquisition Bx after initial change makes contact with other variables that make reversal unlikely *Ethics of intervention reversal (it may be unethical to reverse) - Must be balanced with the right to effective tx *May require considerable time b/c stability required in all phases *Dangers in the comparison of multiple tx due to sequence effects |
Disadvantages of Reversal Design |
|
|
*Best to use: - If target bx is reversible - If withdrawal of the intervention is not a concern - If stability/time is not a concern |
Uses of Reversal Design |
|
|
*Rapid, sequential application & removal of 2 or more IVs *Repeated measurement of bx while the 2 conditions alternate rapidly E.g. Every session is a different condition |
Multielement Design |
|
|
*Shares logical properties with reversal design - Each data point serves to verify previous predictions - Each data point permits comparison against the predictions made by data in the other conditions - Experimental control is demonstrated when bx is appreciably and consistently different in one condition relative to others |
Multielement Design Logic |
|
|
*Comparison of 2 or more treatments *Rapid comparison of treatment to BL *Comparison of 2 or more assessment conditions (as in Functional Analysis of Bx) |
Multielement Design Common Uses |
|
|
Also known as *Alternating treatments design: not necessarily "treatments" *Simultaneous treatment design: but not really "simultaneous" *Concurrent schedule design: but unlike a concurrent schedule of reinforcement *Multiple schedule design: not like a multiple schedule as commonly employed in basic research |
Multielement/Alternating Treatments Design |
|
|
*Multielement w/no BL - Often includes a BL or control as one of the alternating conditions - Should not be regarded as the same as a pre-treatment BL due to multiple treatment interference *Multielement w/BL - Preferable unless contraindicated by severity of bx *Multielement w/BL + a final treatment phase - Clinically prudent in applied research - Permits explanation of multi treatment interference |
Multielement Design Common Variations |
|
|
*Best if each condition associated with a distinct set of stimuli to promote discriminability (mixed vs multiple schedules) *Conditions are counterbalanced across experimental contexts (e.g. time of day, therapist) to neutralize confounding factors *Order of interventions - Strict alternation- not recommended as it does not neutralize sequence effects - Randomization: randomly arranged sequence - Randomization w/restriction: quasi-random, never more than 3 in a row *"Yoked" elements- rapid alternation makes yoking meaningful - Yoking: retain elements of 1 condition in another condition as control |
Multielement Designs Additional Design Considerations |
|
|
*Ideal for comparison of treatments *Can compare treatments while minimizing sequence effects *Useful for highly variable bx that fluctuates as a function of non-experimental variables - can "absorb" the influences of extraneous variables as long as clear & consistent differences remain btwn conditions - can be more efficient (in terms of # of sessions) than other designs |
Multielement Design Advantages |
|
|
*Notoriously subject to multiple treatment interference - Would the effects on any one tx be different if it wasn't being simultaneously compared with another? - Can be examined/controlled by including a final best treatment phase *Unsuitable for individuals that have problems forming discriminations *Enhancing discriminability - less of a problem if IV are quite different - Providing & "anchoring" additional stimuli (e.g. therapist ambient stimuli) to facilitate discrimination - Reducing # of conditions - Instr. control when appropriate - Reverting to other designs *Unsuitable for interventions that produce change slowly or require continuous implementation to produce effects *Limited to situations in which bx is reversible or at least pliable *May require considerable care in doing the necessary conterbalancing |
Multielement Design Limitations |
|
|
*An experimental design in which an initial BL phase is followed by a series of treatment phases consisting of successive and gradually changing criteria for reinforcement or punishment. - Each subphase more closely resembles the terminal target bx or goal *Experimental control is evidenced by the extent the level of responding changes to conform to each new criterion. |
Changing Criterion Designs |
|
|
*Each subphase provides a "BL" for the following phase (description & prediction) *Experimental control is demonstrated when performance closely matches the specified criterion - Probability is small that an extraneous variable produced change across conditions if bx changes when & only when a new criterion is introduced |
Changing Criterion Design Logic |
|
|
*Number of criterion changes - Minimum of 2 (otherwise an AB design) - Too many changes too rapidly may obscure orderly effects *Phase duration (i.e. sessions per phase) - Determined by stability (each subphase is BL) - Phases can be shorter if bx changes rapidly - Length of phases should vary (additional demonstration of control: you change or keep bx at a given level for as long or briefly as you plan) *Small initial criterion changes maximize probability of success, but if too small may not be large enough to ensure change in bx are detectable *If too large - Might not be able to have enough phases to demonstrate control - Requiring drastic change may make it difficult for the subject to meet criterion *Rule of thumb: small change for very stable bx, large change for variable bx |
Changing Criterion Design Additional Design Considerations |
|
|
*Treatments do not have to be withdrawn *Does not require multiple bx, subjects or settings *All subjects can receive tx after the same length of BL |
Changing Criterion Design Advantages |
|
|
*Difficult to interpret when bx does not closely match criterion *Useful only when it is meaningful to measure bx change in stepwise increments *Requires considerable time & effort in planning |
Changing Criterion Design Limitations |
|
|
*An experimental design that begins with the concurrent measure of two or more bx in a BL condition, followed by the application of the tx variable to one of the bx while BL conditions remain in effect for the other bx(s). After maximum change has been noted in the first bx (and is stable), the tx variable is applied in sequential fashion to each of the other bx in the design. *Experimental control is demonstrated if each bx shows similar changes when, and only when, the tx variable is introduced |
Multiple BL Design |
|
|
*Experimental control is demonstrated by showing that bx changes when, and only when the IV is introduced to each BL - The plausibility of extraneous variables causing the change is highly unlikely under the circumstances |
Multiple BL Design Logic |
|
|
*Multiple BL Across Subjects *Multiple BL Across Behaviors *Multiple BL Across Settings |
Multiple BL Design Variations |
|
|
*Useful when bx change is not reversible *Does not require counter-therapeutic bx change to demonstrate experimental control *Experimenter can "test" methods & interventions before applying on a larger scale |
Multiple BL Design Advantages |
|
|
*How many baselines? - The larger the # the more convincing - Lends internal & external validity to the analysis - Using only 2 BL in a multiple BL design can be a risk *If 1 does not change, the conclusions are questionable (essentially an AB design) *If 3 or more used & 1 doesn't change, still a reasonable demonstration of exp. control *How long a BL? - Same rules apply (as long as necessary but...) dictated by stability level & trend of first BL - Problems arising from prolonged BL *Ethical considerations (can we wait to treat 2nd subj?) * The longer the BL, the greater opportunity for influence of ext. variables, generalization effects, practice effects, history effects, maturation effects *Avoid prolonged BL through - Very short (e.g. 1 session) lags or staggering (practical but not highly recommended) - Combine BL into clusters & implement multiple BL across clusters - Multiple probe technique |
Multiple BL Design Additional Design Considerations |
|
|
*Select independent but functionally similar BL * Select concurrent & plausibly related BL *Intervene on the most stable BL first *Vary the length of the multiple BL significantly |
Multiple BL Design Procedural Guidelines |
|
|
*AKA delayed multiple BL *Separate BL are taken & staggered but not conducted at the same time *Logic works on the same principle *Advantages - Permits greater flexibility in the analysis not constrained by having to have all subjects concurrently present *Disadvantages - Presents a greater interpretive challenge than concurrent multiple BL if bx changes on subsequent BL before IV is introduced - Not recommended across bx or settings - Data may be "contaminated" by exp. manipulation of the first bx and may produce a false positive |
Multiple BL Design Variation Non-Concurrent Multiple BL Across Individuals |
|
|
*Intermittent measures (probes) are taken rather than continuous measurement on each BL *Often, first BL is continuous, but subsequent BL data collection is conducted on an intermittent basis relative to the first BL |
Multiple BL Design Variation Multiple Probe Technique |
|
|
*Avoids "ritualistic" gathering of BL data bx is so stable (non-existent) it is unlikely to change *Avoids various threats (e.g. extended practice) & is easier to implement *Useful when extended BL is impractical or costly or possibly detrimental (e.g. repeated exposure to non-treatment is potentially punitive) |
Multiple BL Design Variation Multiple Probe Technique Advantages |
|
|
*Risks stability (e.g. infrequent probes could be outliers) |
Multiple BL Design Variation Multiple Probe Technique Disadvantages |
|
|
*The inclusion of features from 2 or more designs w/in the same experiment *Especially if the conclusion of the "planned" design are tenuous (e.g. discriminability problem in multielement or reversal) *Are not usually planned, rather used to make judgments about experimental control as the data evolve *Uses & advantages - Enhances clarity of the results if it meets the requirement of more than 1 design |
Design Cominations |
|
|
*Gradually withdrawing tx components to see if bx is maintained Uses *Evaluate maintenance of tx effects in the absence of intervention *can be a fading process or can be a component analysis Examples *Implementing isolated components of a tx package |
Component Analyses Sequential Withdrawal |
|
|
*Systematic examination of the differential effects of a range of values of the IV Uses *Determining the effective parametric values of consequences such as duration or magnitude Examples *Examination of differential effects of varying reinforcement schedule values *Comparison of tx @ different strengths (brief vs long time out, reinforce every 3rd mand) |
Parametric Analysis |
