5th Edition Task List Section C: Measurement, Data Display and Interpretation

Operational Definition

Operational Definition

An observable, measurable description of a target behavior.

An example of operational definition

You are having a disagreement with your significant other. They tell you that you are "emotionally unavailable". You ask, "What does that mean?" They define "emotionally unavailable" as "not responding to them when they need to talk, ignoring their attempts for affection, and not providing them with any reassurance".

Why do operational definitions matter?

They describe exactly what we are measuring. Without operational definitions, our ability to collect and analyze data would be extremely hindered. We need our definitions to be objective enough that multiple people could take the same data on the same behavior.

Direct Measures

A way of taking data on a behavior of interest by observing the behavior itself and recording observable and measurable information about it.

An example of direct measures

A directly observes, and takes frequency data on, a client's self injurious behavior in a classroom setting.

Why do direct measures matter?

They allow an observer to directly assess a client's behavior in the environment, which may yield the most accurate and useful information about the target behavior.

Indirect Assessment

Data that are obtained by interviews, checklists and rating scales which include an individual's subjective experience of target behavior. Indirect measures still gather information about the behavior of interest using interactions with people, but not through direct observation. Depending on the case, information could be gathered from the primary client themselves as well as other stakeholders.

An example of indirect Assessment

A behavior analyst is conducting a functional behavior assessment on a new client who demonstrates a significant amount of aggression towards others. The behavior analyst has direct support professional staff members complete a behavior checklist that identifies the time, location, activity and severity rating of the problem behavior throughout the day.

Why do indirect assessments matter?

They provide information from the individual, family members and other caregivers that may not be available when conducting a direct assessment. Indirect assessment measures should not be used alone; rather, they make sense as part of an overall clinical case conceptualization process which also included direct observation and assessment measures.

Product Measures

Measuring a behavior after it occurred by examining the effects the behavior produced on the environment. Unlike direct and indirect measures, product measures sometime do not involve people at all.

An example of product measures

A student is expected to complete math work before accessing a video game. The teacher examines the work and determines whether the behavior was satisfactory based on predetermined criteria regarding the product measurement (at least 10 separate math problems completed at 80% or more accuracy overall).

Why do product measures matter?

Permanent product measures allow for accurate, complete and continuous data when warranted.

Frequency

How often a behavior occurs

An example of frequency

A BCBA is observing a student in their classroom and observes them call out 17 times while sitting at their desk.

Why does frequency matter?

Frequency is used when you want to count how many times a behavior is occurring. Behaviors may occur too often or too little and may need to be targeting for intervention.

Rate

A measure of how often a behavior occurs over an amount of time. Rate is like frequency, expect with a time component added.

Example of rate

You eat 30 potato chips in 15 minutes.

Why does rate matter?

Rate is used when you want to know how much a behavior is occurring over time. In certain instances, rate may be targeted to increase or decrease depending on the behavior.

Percentage

A measurement expressed as a portion of each hundred.

An example of percentage

A behavior analyst calculates that their client engaged in aggression in 20% of data intervals over the course of the day.

Why does percentage matter?

Percentage can be useful to get a snapshot of how a person's behavior reduction or skill acquisition is progressing. Percentages have limitation because they do not tell us all necessary clinical information. For example, if a client is consistently getting 90% on letter identification, we don't know if he is making "random" errors or if there is a pattern of skill deficit.

Duration

The amount of time during which a behavior happens; how long the behavior takes

An example of duration

A student engages in tantrum behavior for eight minutes during music class.

Why does duration matter?

How long a behavior lasts matters a lot in determining whether it is harmful or beneficial. For example, taking a 20 minute nap in the afternoon is different from sleeping 6 hours during the day. Crying for a few minutes once on a while is different from crying in every class or meeting. Exercising for two minutes twice a week is unlikely to have a major impact on health or fitness, but exercising two hours per day might lead to increased susceptibility to injury.

Latency

The time between an opportunity to emit a behavior and when the behavior is initiated.

An example of latency

Your phone beeps because you received a text message. You reach over to check your phone 30 seconds later.

Why latency matters

Decreasing latency can generally increase a person's ability to both contact reinforcement more quickly and develop more adaptive social repertoires. For example, taking two hours to do something your boss asked you todo will probably be problematic, and taking 5 minutes to respond to a social greeting will likely not make you many new friends.

Interresponse time (IRT)

The amount of time that elapses between two consecutive instances of a behavior. IRT is measures from the end of the first response to the beginning of the second response (and so forth if there are more than two responses).

An example of Interresponse time (IRT)

A patient is engaging in vocal outbursts which the behavior analyst is tracking. The record the time that elapses between the end of each vocal outburst and the beginning of the following one.

Why inter response time (IRT) matters

Interresponse time is a frequent measurement used to examine the effects of schedules of reinforcement guided by rate of response.

Topography

What a behavior looks like

An example of topography

A behavior analyst describes a client's aggression toward property behavior in the following manner: "Tearing items off of walls, knocking over furniture, throwing or swiping items off surfaces"

Why does topography matter?

Topography tells us what the behavior looks, sounds, and feels like. Topographical operational definitions should include objective descriptions of the behavior of interest.

Magnitude

The force, intensity and/or severity of a behavior

An example of magnitude

You are at a concert with your friends. Your friend begins screaming and shouting that your ears start hurting. This is a high magnitude of screaming behavior.

Why does magnitude matter?

Magnitude measures how intense and/or how severe a behavior can be. The magnitude of a behavior should be described in tandem with the topography of the behavior.

Trails to Criterion

A measure of the number of response opportunities (trials) needed to achieve a predetermined level of performance (the mastery criterion)

An example of trials to criterion

A student is learning how to tie their shoes in occupational therapy. The OT collects trials to criterion data on the steps required to complete the shoe typing routine during their daily sessions. It took 11 trials for the client to complete all the steps of shoe tying independently and accurately.

Why does trials to criterion matter?

Trials to criterion can be used for assessing a learner's competence in acquiring new skills as well as comparing efficiency of different treatment methods. If intervention A involves 8 trials to criterion, while intervention B involved only 2 trials to criterion, that information can be used to tailor instruction to the individual client in question.

Validity

The extent to which we are measuring what we intend to measure. In other words, do out data points actually represent what we think/say they do?

An example of validity

A behavior analyst wants to collect data on how long a behavior of interest lasts. They collect data on duration. This is a valid measure because the behavior analyst wants to determine the duration of the behavior and uses an appropriate measure. An invalid measure would have involved, for example, taking frequency sound data, which would not have indicated how long the behavior lasts.

Why does validity matter?

When data are relevant to the phenomenon of interest, we can begin to use other scientific processes to better understand and intervene on improving socially significant behaviors. To ensure the best outcomes, we must be certain that we are treating the behavior that we want to treat and not some other behavior.

Reliability

The extent to which a measurement procedure produces the same value repeatedly. In other words, can you rely on it?

An example of reliability

Two behavior analyst are conducting a functional analysis on a client who exhibits self-injurious behavior. Each condition lasts for five minutes each and is repeated over the course of four consecutive days. Both behavior analysts use the same measurement tool to collect data during the functional analysis and their results are nearly identical over repeated measures. This measurement was reliable.

Why does reliability matter?

Highly reliable means that change in data can be attributed to other variables, such as the intervention, rather problems within the measurement system itself. This is crucially important when evaluating the effectiveness of the intervention

NOTE on reliability

You can have reliability without validity. For example, a small child might step on a scale multiple times, and get a weight reading of 482 pounds each time. This is a reliable, but clearly not valid, measure.

Whole Interval Recording

An observation time that is divided into smaller series of brief time intervals where at the end of each interval, the observer records whether the target behavior occurred throughout the entire interval or not.

An example of whole interval recording

A BCBA is collecting data on a child's on task behavior in 10 second intervals. They only circle "yes" if on task behavior occurred in the interval if it occurred for the entire 10 seconds.

Why does whole interval recording matter?

Whole interval recording is best used for measuring behaviors you want to see increase, since it provides a conservative rate of the behavior of interest. Whole interval recording tends to underestimate the rate of a behavior. Though not precise, it is useful when recording each occurrence of (a) behavior(s) is impractical.

Partial Interval Recording

An observation time that is divided into smaller series of brief time intervals where at the end of each interval, the observer records whether the target behavior occurred at any point during the interval.

An example of partial interval recording

A behavior analyst is conducting an observation on a client and is collecting partial interval data in five-minute intervals on the target behavior. The client engages in problem behavior two minutes into the interval and not again for the rest of the interval. The behavior analyst circles "yes", because the target behavior occurred during part of the interval.

Why does partial interval matter?

Partial interval data is best used for measuring behaviors that you want to see decrease, since it provides a conservative measure of the behavior of interest. Partial interval data tends to overestimate a behavior of interest. Though not precise, it is useful when recording each occurrence of (a) behavior(s) is impractical.

Momentary time sampling

An observation time that is divided into smaller series of brief time intervals where at the end of each interval, the observer records whether the target behavior occurred at the end of the interval only.

An example of momentary time sampling

A behavior analyst is observing a client's pacing behavior using momentary time sampling with an interval of 10 seconds. The behavior analyst indicates whether pacing behavior occurred only if the client was pacing at the very end of the 10 second intervals.

Why does momentary time sampling matters?

Momentary time sampling can overestimate or underestimate the rate of behavior. Since you do not continuously measure throughout the entire interval as you do with partial and whole interval, it's smart to ensure that interval are short enough to observe the target behavior frequently. Though not precise, it is useful when recording each occurrence of (a) behavior(s) is impractical.

Momentary Time Sampling

An observation time that is divided into smaller series of brief time intervals where at the end of each interval, the observer records whether the target behavior occurred at the end of the interval only in a group setting.

An example of Planned Activity Check

A behavioral consultant is collecting data on classroom of student's on task behavior during their scheduled rotating activities using PLACHECK intervals of 30 seconds. At the end of the 30 second interval, the behavioral consultant observes if all the students in the classroom were in their assigned areas of the room. If not all students were, there behavioral consultants would indicate that the target behavior was not observed.

Why does Planned Activity Check (PLACHECK) matter?

It allows the observer to collect data on groups as a whole

Equal-Interval Graph

Graphs where the distance between two consecutive points on both the X and Y axis represent the same value (e.g., the X axis and Y axis are both intervals of five)

An example of Equal-Interval Graph

Line graphs, bar graphs, cumulative records, scatterplots.

Line graph

A graph based on the Cartesian plane where a two-dimensional area is formed by the intersection of two or more lines forming a data path.

An example of a line graph

A behavior analyst creates a line graph that displays baseline data and intervention data on the target behavior of hitting others.

Bar graph

A graph based on the Cartesian plane where there are no distinct data points representing responses through time

An example of bar graph

A behavior analyst displays data from a preference assessment that was conducted with a child

Why do line graphs matter?

Line graphs should be used when you want your data to communicate quantitative relations such as time or the other order of responses in a sequence. Line graphs can include multiple data paths for different behaviors

Why do bar graphs matter?

Bar graphs are used to display a set of data that are not related to each other.

Cumulative Records

A graph that displays the cumulative number of responses emitted are represented on the vertical axis, where the steeper the slope of the response the greater the response rate

An example of cumulative records

A teacher is conducting a Manding program with a student and uses a cumulative record to display that student's mastered mands.

Why does a cumulative record matter?

Cumulative records demonstrate response frequency over a period of time, which helps to quickly conceptualize the rate at which a learner is gaining new skills.

Scatterplot

a graph that shows the relative distribution of individual measures data set (e.g., aggression, sleep, etc.)

An example of scatterplot

A behavior analyst creates a scatterplot of a client's aggressive behavior to determine any temporal patterned between aggression and time of day

Why do scatterplots matter?

Scatterplots communicates temporal relations of a behavior of interest if one exists.

Level

The value of a data point along the x-axis of a graph

An example of level

A behavior analyst is conducting visual analysis of a client's target behavior of head to wall self-injury. The behavior analyst determines the level by loading the number along the x-axis to the data points within the graph. The behavior analyst observes that the level of data point are located around the 10% interval along the x-axis.

Why do levels matter?

Examining the level of a data point is a skill in visual analysis that allows the behavior analyst to determine how much or little a behavior has changed.

Variability

The extent to which the data move around on the graph.

An example of variability

A behavior analyst is conducting visual analysis of a client's target behavior of dropping to the floor. The data path is scattered all around the graph. This shows a high degree of variability in the client's dropping behavior.

Why does variability matter?

Variability demonstrates the consistency to which change is taking place. A high variability may demonstrate a low degree of control of an intervention condition, whereas a low variability may demonstrate a high degree of control of an intervention condition. (In other words, if data points are all over the place, there is probably something else going on that has not been accounted for yet.)

Trend

The overall direction of the data path

An example of trend

A behavior analyst is conducting visual analysis on a client's hitting behavior. They observe that the data path is increasing in trend.

Why does trend matter?

Examining trend provides us with information about the "bigger picture" of where a behavior is heading based on past responding. It is helpful as part of intervention planning and evaluation.

Dependent Variable

The target behavior which the intervention is designed to change. It depends on the environment to change it.

An example of dependent variable

A client's eloping behavior which is targeted for intervention

Why do dependent variables matter?

The dependent variable must be identified if the goal is to produce change in behavior.

Independent variable

The intervention designed to have an effect on the dependent variable. The independent variable intervenes on the phenomenon of interest.

An example of independent variable

Response blocking as a means to prevent elopement.

Why do independent variables matter?

In order to accurately understand behavior change, all change targets and treatment conditions must be identified. Though we may not be actively involved in publishing research, our work with clients must be driven by an empirical process.

Internal Validity

An experiment shows convincingly that changes in a behavior are a function of the intervention/treatment and NOT the result of uncontrolled or unknown factors.

An example of internal validity

A behavior analyst implements a DRA procedure to support a client who engages in skin picking. The skin picking behavior responds favorably to the intervention. When the DRA is removed, the target behavior returns. When the intervention is put in place a second time, the behavior returns to low rates. This fact pattern strongly suggests that the DRA intervention (and nothing else) was primarily responsible for the reduction in skin picking behavior.

Why does internal validity matter?

Without high internal validity, cause and effect relationships cannot be discovered.

External Validity

The degree which a study's results are generalizable to other subjects, settings and/or behaviors not include in original study.

An example of external validity

A behavior analyst is implementing a new intervention from a study that they read in a peer reviewed journal. The individual participant variables (developmental level, topography and function of behavior, for example) are a good match with the behavior analyst's current client. The analyst replicates the intervention steps with their client and achieves similar favorable results. This supports the study's external validity, since the results from the study have been replicated with a different subject

Why does external validity matter?

Research findings are clinically useless unless they can convincingly demonstrate (1) that the methods were responsible for the observer changes, and (2) that the methods can work across participants and contexts not included in the original study. Science is always building and correcting itself, and replication is a vital-if unglamorous- part of the scientific process!