An Overview of Experimental Design Experimental Control Definition

An Overview of Experimental Design

Experimental Control Definition Is a predictable change in behavior (dependent variable) that can be reliably produced by the systematic manipulation of some variable (independent variable)

Goal of Behavior Analysis Experimentally determine the effects of environmental manipulation on behavior 2. and demonstrate the effects can be reliably produced 1.

Internal Validity The extent to which an experiment shows that changes in behavior are a function of the independent variable and not the result of uncontrolled or unknown variables

Problem: Confounding Variables known or suspected to exert an uncontrolled influence on the dependent variable Effects must be evaluated and eliminated to demonstrate experimental control Problem: May not know all of the confounding variables Time of day Barometric pressure Other unknown

Solution Eliminate as many uncontrolled variables as possible OR Hold constant the influence of all other variables except the independent variable

Behavior Can be an individual phenomenon Can be a continuous phenomenon Assumes That: Behavior is determined Behavioral variability is extrinsic to the organism

Behavior as an Individual Phenomenon Examine a person’s interaction with the environment With ABA Strategy often based on within-subject (single-subject) designs

Contrast: Dynamic / Continuous Behavior Ø Behavior changes over time Ø Requires continuous measurement over time • Complete record of behavior as it occurs in context • Often uses cumulative recorder types of methodologies for recording • Systematic repeated measurement is the “hallmark” of ABA

Assumes 1. The occurrence of any event is determined by the functional relations it holds to other events 2. Behavior is a natural phenomenon 3. Behavioral variability is the result of some environmental influence • In ABA • Try to find out what it is • Try to find interventions to change the behavior

Components of Experiments in ABA • Examine at least one – – Subject or participant Behavior (dependent variable) Setting Treatment or intervention condition (independent variable)

Research Question(s) Ø Are statements about what the person wants to learn about a behavior by conducting an experiment. Ø Specifically: What are the effects of the independent variable on the dependent variable Ø Are always written down: Ø Nothing exists until it is on paper (Brigham) Ø Allows you to identify what you are examining Ø Allows you to identify problems Ø Allows you to create solutions

Many Different Designs Group Designs Single Subject Multiple Baseline Simultaneous Treatment Others Each has its advantages and disadvantages. Past research and experience helps you decide on the type of design to use

Experiment using Groups Randomly selected pool of subjects from relevant population Divided into experimental and control groups Pretest, application of independent variable to experimental group, and posttest

Problem Group data may not be representative of individual performance Some individuals within a group may stay the same or decrease, while the improvement of others could make it appear as overall average improvement To be most useful, treatment must be understood at an individual level Individual can mean a group in an organization

Group Data Masks Variability Hides variability that occurs within and between subjects Use statistical control to compensate Should not be a substitute for experimental control To control effects of any variable, must either hold it constant or manipulate it as an independent variable

ABA q Often uses single-subject designs q Subject becomes its own control Ø Measures of the subject’s behavior during each phase of the study Ø Allows you to compare variables as they are presented or withdrawn during various conditions

May have Single or Multiple Dependent Variables • Single • Only manipulation one variable • Can often see clear results • Multiple • Still manipulate one variable • Have multiple dependent variables

Reasons Gives data patterns that can serve as controls when evaluating & replicating the effects of an independent variable Allows you to assess the presence or extent an independent variable may effect behaviors other than the response class. Allows you to determine whether changes in behavior of others occur during the course of an experiment Changes may explain changes in the subject’s behavior

Measurement Systems and Visual Analysis • Observation & recording procedures must be conducted in a standardized manner • Standardization involves every aspect of the measurement system – Definition of the target behavior to scheduling of observations • Behavior analysts must develop skills in the detection of changes in the level, trend, and degree of variability in behavioral data

Independent and Dependent Variables – Independent Variable – The aspect of the environment that the experimenter manipulates – The researcher controls or manipulates this variable independent of the subject’s behavior or any other event – Dependent Variable – The behavior that changes as a result of changes in the independent variable – Is the thing you are observing

Experimental design • Nonparametric study – Independent variable is either presented or absent during a time period or phase of the study • Parametric study – The value of the independent variable is manipulated – Seeks to discover the differential effects of a range of values

Fundamental Rule in Experimental Designs • Change only one variable at a time – Experimenter can attribute any measured changes to a specific independent variable – If investigating the effects of a “treatment package” • Ensure that the entire package is presented or withdrawn each time a manipulation occurs • Allows you to make causal statements

Some Additional Rules • Do not get locked into textbook “designs” • Select & combine experimental tactics that best fit the research questions

Steady State Strategies – Repeated exposure of a given subject to a given condition – Try to eliminate or control extraneous influences on behavior & obtaining a stable pattern of responding before introducing the next condition

Baseline Data in ABA • Serves as a control condition • Does not imply the absence of intervention – Absence of a specific independent variable • Why examine? – Allows you to observe a level of responding in the absence of an independent variable – Allows you to make comparisons when you introduce an intervention

Benefits • You can obtain descriptions of antecedent-behaviorconsequent correlations for the planning of an effective treatment – Valuable guidance in setting initial criteria for reinforcement – Baseline data may reveal the behavior targeted for change does not warrant intervention

Types of Baseline Data Patterns • Stable baseline (A) • Ascending baseline (B and C) • Variable baseline (D)

Types of Baseline Data Patterns

Example

Problem May be due to other variables

Need Verification • Done by termination or withdrawal of the treatment variable

Verification

Can use Replication

Conclusions Lots of aspects to experimental designs Need to make sure you know the strengths and weaknesses of the designs you plan to use. Want to ensure your conclusions are based on solid data.