Part 3 Flashcards
Why do behavior analysts use graphs?
graphs are the device BAs use to organize, store, interpret, and communicate results of their work
What is the medium with which behavior analysts work?
data
What do graphs show about a behavior?
behavior change properties: level, trend and variability
What are the parts of a line graph/frequency polygon?
x-axis/abscissa, y-axis/ordinate, intersection/origin
What questions do graphs attempt to answer?
Did a socially-meaningful change in behavior occur? Can it be attributed to the independent variable?
What is considered in a visual analysis?
number of data points, variability and level of performance, direction and degree of any trend (mean and median level lines)
What is internal validity?
when experiments convincingly show that changes in behavior are a function of the independent variable and not the result of unknown/uncontrolled factors (i.e., confounding variables)
What are the experimental methods of behavior analysis guided by?
behavior being an individual phenomenon as well as behavior being determined and behavioral variability being extrinsic to the organism
What are the components of any experiment?
1+ subjects, 1+ behaviors/dependent variables, 1+ settings, system to measure behavior, ongoing visual analysis of data, 1+ treatments/interventions/independent variable conditions, manipulations of independent variable/s (experimental design), experimental question
What is experimental control?
predictable change in behavior (dependent variable) can be reliably/repeatedly produced by the systematic manipulation of the individual’s environment (independent variable)
What are extraneous variables?
all environmental variables outside the independent variable
What is the difference between nonparametric and parametric experiments?
nonparametric studies have independent variables either present or absent while parametric analysis explores a range of independent variable values
What is steady/stable state responding?
little variability in data
What is steady state strategy?
elimination or control of extraneous influences on behavior of interest (requires stable responding)
What are practice effects?
improved performance from repeated opportunities to emit behavior
What is affirmation of the consequent?
behavior will not change from baseline rates if the independent variable isn’t introduced
What are sequence effects?
effects on subject’s behavior resulting from experience from a prior condition
What is DRO reversal technique?
reversal to reinforcement for any non-target behavior (DRA and DRI reversals are similar)
What is irreversibility?
level of behavior in an earlier phase can’t be reproduced despite a return to those earlier conditions
What are alternating treatments variations?
single-phase: alternating treatment with or without no-treatment control condition
2-phase: alternating treatments with initial baseline
3-phase: alternating treatments with initial baseline and final best treatment phase
What are advantages of alternating treatments designs?
treatment withdrawal is not required, speed of comparison, minimizes irreversibility problem, minimizes sequence effects, can be used with unstable data, can be used to assess generalization of effects, intervention can begin immediately
What are alternating treatments designs disadvantages?
multiple treatment interference (confounding effects of treatments), unnatural nature of rapid alternation of treatments, limited capacity for conditions, selection of treatments
Which design is the most straight-forward and generally the most powerful design for demonstrating functional relations?
A-B-A-B design
When should multiple baseline designs be used?
target behavior is likely irreversible, it’s undesirable/unethical/impractical to reverse conditions
What are the three most common variations for multiple baseline designs?
subjects, settings, behaviors
What is a multiple probe design?
method of analyzing the relation between the acquisition of a successive approximation or task sequence [and the application of a treatment(?)]
-intermittent measures to determine whether a behavior change has occurred prior to the intervention
What are the 3 key features of a multiple probe design?
initial probe to determine performance level on each behavior in the sequence, series of baseline measures on each step before training, probe for all steps after current criterion is met
What is a delayed MBL design?
initial baseline and intervention are begun with subsequent baselines added in a staggered/delayed fashion
When are delayed MBLs used?
reversal design is no longer desirable/possible; limited resources, ethical concerns, or practical difficulties preclude a full-scale MBL design; a “new” behavior/setting/participant becomes available
What are guidelines for designing/conducting MBL designs?
select independent, yet functionally similar baselines; select concurrent and plausibly related multiple baselines; do not apply the IV to the next behavior too soon; vary significantly the lengths of the multiple baselines; intervene on the most stable baseline first
What are the advantages of MBL designs?
treatment withdrawal is not required; suited for evaluation of progressive multiple baseline changes, assessing generalization, relatively easy to conceptualize
What are limitations of MBL designs?
may not allow demonstration of experimental control; weaker design for showing experimental control than reversal; provides more information regarding the effectiveness of treatment variable than it does regarding the function of any particular target behavior; requires treatment to be delayed; can be costly
What are guidelines for changing criterion designs?
phase lengths, change magnitudes, number of changes
What are limitations of changing criterion designs?
target behavior must already be in individual’s repertoire and lend itself to stepwise modification; may impede optimal learning rates
What are issues with group data?
may not represent that performance of individual subjects; masks variability of data; intrasubject replication is absent from group designs
Which experiments have a high degree of internal validity?
those demonstrating a functional relation
What is experimental control?
extent to which researcher controls all relevant variables in a given experiment
What are confounding variables?
uncontrolled factors known/suspected to have exerted influence on DV (subjects, setting, measurement of DV or IV)
What is maturation?
subject changes over course the course of an experiment
What are control procedures?
placebo control, double-blind control
In which 4 dimensions should treatment operational definitions be described?
verbal, physical, spatial, and temporal
What is social validity?
importance of behavior change goals, social acceptance of interventions, social importance of behavior changes (normative sample, consumer opinion, expert evaluation, real-world test, standardized tests)
What is systematic replication necessary for?
to determine external validity
How is ABA research evaluated?
- internal validity - definition and measurement of DV, graphic display, meaningfulness of baseline conditions, experimental design, visual analysis, interpretation
- social validity - maintenance and generality of behavior change
- external validity
- theoretical significance and conceptual sense
