Measurement Basics
What is measurement?
Process of applying quantitative labels to describe and differentiate objects and natural events
Why do we measure socially significant behaviors?
–Help determine which behaviors need to be changed
–Detect and compare the effects of various interventions on behaviors targeted for change
–Evaluate the acquisition, maintenance, and generalization of behavior changes
Describe the x and y axis
•Horizontal Axis: Also called x-axis, a straight horizontal line that often represents time, presence, absence, and/or value of the independent variable
•Vertical Axis: Also called y-axis, a vertical line drawn upward from the left-hand end of the horizontal axis. Most often represents a range of value of the dependent variable
What is experimental control?
–Achieved when predictable change in behavior is produced by the systematic manipulation of the environment
Describe reversal design
•Entails measures of behavior in a given setting that requires at least three consecutive phase
–Initial baseline phase in which the independent variable is absent
–An intervention phase during which the independent variable is introduced and remains in contact with the behavior
–A return to baseline conditions accomplished by withdrawal of the independent variable
•An experiment entailing one reversal is described as an A-B-A design
–Capital letter A and B denote the first and second conditions
–Baseline (A) data are collected until steady state responding is achieved
–An intervention (B) condition is applied that signifies the presence of a treatment- the independent variable
–This design entails on reversal
Why is frequent measurement important?
Frequent measurement helps practitioners:
•Optimize their effectiveness
•Verify the legitimacy of treatments
•Identify and terminate treatments based on pseudoscience, fad, fashion, or ideology
•Be accountable to clients, consumers, employers, and society
•Achieve ethical standards
Describe Validity
–Measurement yields data directly relevant to the phenomenon measured and to the reason(s) for measuring it
–Determining the validity of measurement revolves around this basic question:
▪Was a relevant dimension of the behavior that is the focus of the investigation measured directly and legitimately?
–A direct measure of the actual behavior of interest will always possess more validity than an indirect measure
Describe a cumulative record
•Skinner invented the cumulative recorder, a device that automatically draws a graph of a subject’s rate of behavior
•A cumulative record is constructed by adding the number of responses recorded during each observation period to the total number of responses recorded during all previous observation periods
•The steeper the slope, the higher the response
•Overall response rate is the average rate of response over a given time period
•Local response rate refers to the rate of response during periods smaller than that for which an overall rate has been given.
Describe internal and external validity
•Internal Validity
–The degree to which an experiment shows that changes in the dependent variable are a function of the independent variable and not the result of uncontrolled or unknown variables
•External validity
–The degree to which a study’s results are generalizable to other subjects/settings/behaviors
Describe A-B-A-B design
–Preferred because reintroducing the B condition enables replication of treatment effects
–Strengthens the demonstration of experimental control
–Generally most powerful within-subject design for demonstrating a functional relation between environmental manipulation and a behavior
–Also called withdrawal design
Describe repeatability, temporal extent, and temporal locus
–Repeatability: Instances of a behavior can occur repeatedly through time
–Temporal extent: Every instance of behavior occurs during some amount of time
–Temporal locus: Every instance of behavior occurs at a certain point in time with respect to other events
What are the elements necessary to ensure valid measurement?
1.Measuring directly a socially significant target behavior
2.Measuring a dimension of the target behavior relevant to the question or concern about the behavior
3.Ensuring that the data are representative of the behavior’s occurrence under conditions and during times that are most relevant to the question or concern about the behavior
What two questions must we answer when analyzing data and graphs?
–Did behavior change in a meaningful way?
–If so, to what extent can that behavior change be attributed to the independent variable?
What is a confounding variable?
–Uncontrolled variable known or suspected to exert an influence on the dependent variable
When would you use B-A-B design
•Begins with the application of the independent variable: the treatment
•After stable responding has been achieved during the initial treatment phase (B), the independent variable is withdrawn
–If the behavior worsens in the absence of the independent variable (A condition) the treatment variable is reintroduced in attempt to recapture the level of responding during the first treatment phase.
•Practitioners may prefer B-A-B to the A-B-A design which ends with treatment variable in effect
•In terms of demonstrating a functional relation, the B-A-B design is the weaker
•May be appropriate with target behaviors that result in physical harm or danger to the participant or others
What measures can be used to evaluate repeatability?
Count and Rate
Describe accuracy and reliability
•Accuracy
–The extent to which the observed value (i.e., the quantitative labels produced by measuring an event) match the true value of the event
–Measurement is accurate to the degree that it corresponds to the true value of the thing measured
•Reliability
–The consistency of measurement, specifically, the extent to which repeated measurement of the same event yields the same value
–Reliable measurement is consistent measurement
–The more observed value from repeated measurement of the same event differ from one another, the less the reliability
Describe Variability, level, and trend
•Variability
–The extent to which repeated measures of behavior yield different outcomes
–A high degree of variability indicates insufficient control over the factors influencing behavior
•Level
–The value on the vertical axis scale around which a set of behavioral measures converge
•Trend
–The overall direction taken by a data path, described in terms of their direction (increasing, decreasing, or zero trend)
▪Direction and degree of trend can be visually represented by a straight line drawn through the data
Describe the dependent and independent variables
•Dependent variable: the measurable dimensional quantity of the target behavior
–It is so labeled because the experiment is designed to determine whether the behavior is dependent on the independent variable
•Independent variable
–The particular aspect of the environment that the experimenter manipulates to find out whether it affects the subjects behavior
–Sometimes called the experimental variable, intervention, or treatment
–The researcher controls or manipulates it independent of the subject’s behavior
Describe Multiple Baseline Design and the three versions
•Multiple baseline design
-After collecting initial baseline data simultaneously across two or more behaviors, settings, or people, apply the treatment variable sequentially across these behaviors, settings, or people and note the effects
•The multiple baseline design takes three primary forms:
1.Across behaviors
2.Across settings
3.Across subjects
describe discrete trial measurement
•Response rates of behaviors that occur within discrete trials are controlled by a given opportunity to emit the response
•Rate of response is controlled by the presentation of a discriminative stimulus
Describe threats to valid measurement
–Measurement is indirect
–The wrong dimension of the target behavior is measured
–When measurement is conducted in such a way that the data it produces are an artifact of the actual events
Describe the three levels of scientific understanding
•Description
–Systematic observation enables scientists to describe natural phenomena accurately
–Yields a collection of facts that can be quantified and classified
•Prediction
–Achieved when repeated observation shows two events consistently covary – correlation
–Can be used to predict the probability that one event will occur based on the presence of the other event
•Control
–Highest level of scientific understanding
–Comes from establishing experimental control
What is the purpose of baseline data?
•Baseline: data collected in the absence of the independent variable
–Serves as a control condition
–Does not necessarily mean the absence of instruction or treatment, only the absence of the independent variable
Describe component analysis
•A component analysis is any experiment designed to identify the active elements of a treatment package, the relative contributions of different components in a treatment package, and/or the necessity and sufficiency of treatment components
•Two methods for conducting component analyses
–Drop-out component analysis: the investigator presents the treatment package and then systematically removes components
▪If the treatment’s effectiveness wanes when a component is removed, the researcher has identified a necessary component
–Add-in component analysis assesses components individually or in combination before the complete treatment package is presented
▪Can identify sufficient components