UNIT 6 TEST

When behavior in one condition is affected by the person's experience in the prior condition, this is a potential confound known as A)A sequence effect B)Extraneous variable C)History D)Multiple treatment interference

A)A sequence effect When one specific intervention phase regularly follows another specific intervention phase, the sequencing of these phases can be studied in isolation. If the data in the second phase is reliably affected by the fact that it was preceded by the first phase, regardless of the data in that first phase, this is a sequence effect. The presence of the first phase has a reliable effect on the data in the second phase.

Which of the following is NOT a consideration in the visual analysis of a single subject design A)All of these are considerations in visual analysis B)Trend and level You Answered C)Phase duration and latency to change D)Variability and overlap of between phases

A)All of these are considerations in visual analysis Trends, mean shifts in level, variability of data and the overlap of that variability between phases, latency of change once a phase has changed, and overall phase duration (to see if the effect maintains) are all key components to a visual analysis of data.

What is an effective way to enhance discrimination across conditions when using a multi-element design? A)All of these are effective ways to increase discriminability B)Provide instructions before sessions C)Reduce the number of distinct conditions being compared D)Include supplemental stimuli

A)All of these are effective ways to increase discriminability Including supplemental stimuli, providing instructions before sessions, and reducing the number of conditions are all effective ways to enhance discriminability when using a multi-element design.

All of the following are important considerations when using a multi-element design, except: A)All of these are important considerations B)Discriminability between conditions C)Order of the conditions D)Counterbalancing environmental variations

A)All of these are important considerations Counterbalancing, discriminability, and order of conditions are all important considerations when using a multi-element design. Keep in mind that these are distinct considerations, and may require different phases to be implemented in a functional analysis protocol in order for each one to be adequately addressed. For example, counter-balancing (e.g., making sure that not all sessions occur in the morning, or in the same room) may require tactics that are nearly the opposite of the tactics used to enhance discriminability (e.g., distinct, but specific rooms are always used for distinct conditions - so a specific room begins to exert discriminative control related to that particular condition). The point of this question is that all of these considerations must be kept in mind in a multi-element (alternating treatment) design.

A multi-element design would be appropriate when: A)All of these would be appropriate when using a multi-element design B)There are time constraints for assessment C)Behavior is highly variable D)Trying to compare multiple interventions

A)All of these would be appropriate when using a multi-element design Multi-element designs do not require the stability of behavior to demonstrate functional relations, as long as consistent and clear difference are observed across conditions. Therefore, it would be appropriate for highly variable behavior. Since the stability of behavior is not necessary to demonstrated functional elations, a fewer sessions may be required, and it would be appropriate when there are time constraints. In addition, it would be appropriate to compare different interventions because this is how a multi-element design is conducted.

What design technique does this chart represent A)Bi-directionality of change B)Taking multiple baselines on a series of behaviors C)Alternating treatments within the design D)Decreasing stability to improve criterion matching

A)Bi-directionality of change This technique is seen in treatment phase 4 after baseline (the criterion line in sub-phase 3 is at level 16, the criterion line in sub-phase 4 is at a level of 13, and the criterion line goes back up in sub-phase 5 to a value of 16). This technique is used to establish experimental control by demonstrating a behavior rate clearly tracking the criterion as set by the experimenter. Note the mild carry-over effect between treatment sub-phase 4 and the beginning of sub-phase 5, where the behavior rate temporarily continues to be low, but then rises to match the criterion which has been once again increased.

What experimental design combination is represented by the graph below? A)Combined withdrawal and multiple-baseline B)Combined multiple-baseline and multi-element C)Combined withdrawal and multi-element D)Combined changing criterion and reversal

A)Combined withdrawal and multiple-baseline This one may seem very difficult at first, given the limited information on the chart, but this is clearly, at least in part, a multiple baseline design—given the staggered introduction of the IV. The other design is not a changing criterion (there are no level lines setting criteria in any sub-phases) and is not a multi-element, since there is only one data path in every phase. Thus, only the correct answer, withdrawal combined with multiple baseline, is left as a possibility -and this makes graphical sense.

What type of treatment evaluation is being conducted in the graph below? A)Component analysis B)Parametric analysis C)Sequence analysis D)Probes

A)Component analysis There is a treatment package consisting of three components on this graph (components A, B, and C). After demonstrating the effectiveness of this intervention in the first four phases, subsequent evaluations of different combinations of these components are carried out. It is demonstrated that as long as component A is in effect, this intervention is successful, but when component A is removed (B+C combination), this effect is lost. Therefore, this is an example of a component analysis.

Chris' mom expressed concerned over her son's weight gain. She is worried about the "heavy use" of edible reinforcement in his program. She asked the behavior analyst to stop using edibles to reinforce Chris' performance of self-care skills, which are currently maintained by contingent praise, plus choice-of-edible for completing each skill. Current data indicate a high rate of self-care task completion. The behavior analyst wants to evaluate if praise alone will maintain high rates of self-care skills completed on a daily basis. What should the behavior analyst do next? A)Conduct a component analysis B)Conduct a parametric analysis C)Conduct a structural analysis D)Use a multi-element design with final treatment phase

A)Conduct a component analysis A component analysis involves evaluating components (individual treatments) contained in a treatment package to determine the effects of each component alone or in conjunction with one or more other components in the treatment package. In this example, the treatment package contains praise plus edibles. A component analysis allows us to see if we get similar effects with praise alone and if we do we can clinically conclude to discontinue the use of the edibles component.

Which experimental design technique can vary non-essential condition components (e.g., times of day, therapists, rooms, etc.) to minimize potential confounding variables? A)Counterbalance B)Use strict alternation C)Use randomized order of conditions D)Enhance discriminability

A)Counterbalance Enhancing discriminability, using randomized order of conditions, and using strict alternation do not involve "switching" condition components.

These are "events not related to the IV that may affect the dependent variable". A)Extraneous Variables B)Sequence effects C)Confounds D)Explanatory fictions

A)Extraneous Variables This is definitional. Extraneous variables are not related to the independent variable. Confounds are usually related to the structure of the independent variable (the way the design is implemented). Moreover, extraneous variables may be controlled prior to their occurrence, while confounds usually are defined by their appearance in data already collected (such as a carry-over effect or multiple treatment interference). This is not the definition for explanatory fictions - which are hypothetical constructs or teleological explanations for behavior.

Which of the following options is NOT an advantage of a changing criterion design? A)Highly useful for the rapid reduction of dangerous topographies B)All subjects can receive treatment after the same length of baseline (assuming stable baseline for all) C)Does not require multiple behaviors, subjects, or settings to demonstrate a functional relation D)Treatment does not have to be withdrawn to demonstrate a functional relation

A)Highly useful for the rapid reduction of dangerous topographies No—not for dangerous behavior! This is a design for incremental change, and is typically used for behavior which is already in a person's repertoire, but needs to be strengthened. This design can be used for reduction targets, but ones which can be reduced slowly, without continuing to pose a threat to the individual or others. When using a changing criterion design, all subjects can receive treatment after the same length of baseline, as long as those baselines are stable of course. Multiple behaviors, subjects, or settings are not required to demonstrate a functional elation, and treatment does not have to be withdrawn to demonstrate a functional relation.

What characteristic of the following chart provides the most convincing evidence of a functional relationship? A)Latency to change B)Overlap C)Variability D)Trend

A)Latency to change The trend change is not compelling; the variability is very similar, perhaps a bit greater in the treatment phase, but not in a way that suggests anything in particular; there is no overlap, which argues for a possible effect, but not strongly. However, the latency to change was rapid and significant.

Which of the following is a limitation of multi-element designs? A)May be prone to multiple treatment interference B)May require multiple behaviors, participants, or settings C)May require a long time to complete D)Requires behavioral stability before other conditions are introduced

A)May be prone to multiple treatment interference Multi-elements designs do not require multiple behaviors, participants, or settings. They do not require a long time to complete because behavioral stability is not necessary to move from one phase to the next, and functional relationships will be exhibited by consistent differences between data paths across distinct conditions. Multi-element designs may be prone to multiple treatment interference because different conditions are rapidly alternated.

During Allison's treatment evaluation, her program is randomly alternated with a baseline condition and a non-contingent reinforcement condition. Her program consists of a time-out procedure for a problem behavior along with a reinforcement procedure for a functionally equivalent alternative response. What type of confound is potentially hidden in the way this evaluation is arranged? A)Multiple-treatment interference B)Carryover effects C)Potential of extraneous variables influencing the DV D)Sequence effects

A)Multiple-treatment interference Notice that her two treatment components are always compared as a package, and neither treatment being tested in isolation. A sequence effect or carry-over effect could occur in some of the randomized pairings, but this would be evident in the data. But if the two treatments are never tested in isolation, any potential multiple treatment interference will go unevaluated, and will threaten the internal validity of any results as a hidden confound.

The graph below represents which design? A)Parametric Analysis B)Component Analysis C)Changing Criterion design D)Multielement design

A)Parametric Analysis Note that we are examining the effects of a single drug at different levels of dosing (parameters). The graph indicates that 70 mg has the greatest reductive effect on rate of aggression.

The graph below represents which design A)Parametric Analysis B)Multi-element design C)Component Analysis D)Changing Criterion design

A)Parametric Analysis Note that we are examining the effects of a single intervention at different values specifically duration of access. 30 minutes of video game time produces the greatest increase in the target response.

Which of the following is NOT a consideration in the visual analysis of a single subject design A)Range of data and mode shift B)Trend and level C)Variability and overlap of between phases D)Phase duration and latency to change

A)Range of data and mode shift Trends, mean shifts in level, and variability of data are the most common considerations in visual analysis. Overlap of variability between phases, latency of change once a phase has changed, and overall phase duration (to see if the effect maintains) are also key components to visual analysis. However, range of data simply refers to the difference between the high data point and the low data point, while the mode is simply the most common data value within a data set. These pieces of information are not particularly useful for visual analysis.

What characteristic of the following chart provides the most convincing evidence of a functional relationship? (Assume this is in the midst of an A-B-A-B-A-B design showing similar results between all baseline versus treatment phases). (The Y-axis label is "# of responses") A)Trend B)Variability and overlap C)Latency to change D)Mean shift

A)Trend Mean shift, latency to change, and variability/overlap would all actually argue against an effect, but the change in trend is so significant, that a functional relationship is suggested.

When a pattern of behavior is established in one session and it extends into the next session, this X is known as Y . A)X = confound Y = carryover effect B)X = confound Y = multiple treatment effect C)X = extraneous variable Y = sequence effect D)X = extraneous variable Y = diffusion of treatment

A)X = confound Y = carryover effect An extraneous variable is an event that is not related to the IV that may influence the DV, and its potential can be controlled with careful planning and execution. A confound is an uncontrolled factor which often is a result of the implementation of the IV, which is likely to have already influenced experimental results. Diffusion of treatment is an extraneous variable, but occurs when components of treatment are implemented during baseline, for whatever reason, which spoils the data comparison between baseline and treatment. This question gives no information about multiple treatments, so that is not salient. Sequence effects are confounds, which can sometimes present as another semi-independent type of confound, the carry-over effect: When the behavior effect of an IV, implemented in one session, literally carries over into the next session, this is called the carry-over effect, appropriately enough. Carry-over cannot be reliably predicted or prevented in advance - it is a phenomenon that arises from the sequence of the IV itself. You can however take steps to clarify your data with further experiments, which include such techniques as use of extended alone conditions.

A component analysis is carried out by: A)systematically withdrawing treatment components to see if behavior change is maintained B)summarizing A-B-C data to identify sequences of events that are correlated with a target behavior C)systematically examining the effects of a range of values of the independent variable D)assessing behavior on occasions when the contingencies arranged in the analysis are not in effect

A)systematically withdrawing treatment components to see if behavior change is maintained Assessing behavior on occasions when the contingencies arranged in the analysis are not in effect describes probes, not a component analysis. Summarizing A-B-C data to identify sequences of events that are correlated with a target behavior describes a sequence analysis, not a component analysis. Systematically examining the effects of a range of values of the independent variable describes a parametric analysis, not a component analysis.

What is the approximate level of the independent variable in the final sub-phase? A)27 B)17 C)15 D)13

B)17 The final independent variable value on this chart (the final targeted behavior criterion for "# of sit ups in one minute") is 17. The final data point is also 17. The first data point in the final phase is 15. The final phase begins in session #27. The level of the IV in the second to last sub-phase (sub-phase #4 - the last one run to completion) is 13.

ooking at the chart below, what appears to be the rule for when to implement the next change in criterion (i.e., begin the next sub-phase) for "sit ups in one minute"? A)4 data points at or above criterion B)3 consecutive data points at or above criterion C)5 or more data points within phase D)3 data points at or above criterion

B)3 consecutive data points at or above criterion In treatment sub-phases 1 and 3, there are only 3 data points at or above the criterion line before the change to the next sub-phase; in treatment sub-phases 2 and 4, there are 4 data points above criterion before the change—but the data path falls below criterion after the first data point above the line in each case. In all four cases, the criterion changes after 3 consecutive data points at or above the criterion line.

When using a changing criterion design, bi-directional change refers to the strategy of: A)Selecting behavior that is reversible so it can show an increase or decrease under different conditions B)Changing the next criterion to a value seen in a prior sub-phase - to bolster a demonstration of experimental control C)Changing the next criterion to the value of a later planned sub-phase value to demonstrate experimental control D)Reversing a treatment contingency up and down across phases

B)Changing the next criterion to a value seen in a prior sub-phase - to bolster a demonstration of experimental control Bi-directional change occurs when a string of criteria changes occur in one direction, and then a criterion is inserted that moves against that trend. Behavior which matches this criterion, bucking the trend established in prior criteria levels, demonstrates that the criterion level (the IV) is controlling the behavior (the DV), thus demonstrating experimental control. Changing a criterion to a value seen in a later planned phase does not assist in the establishment of experimental control, since in essence, ALL criteria changes meet this definition, including a "uni-directional" design change. Bi-directional change refers to the change in the target level for the behavior, not the reversing of a treatment contingency across phases (which might be a reversal design, depending on the particulars). Bi-directional change does not consist of selecting behavior that is reversible - such a selection process would occur prior to the collection of any data.

When the conclusions of an experimental design are tenuous, which option makes the most sense for the experimenter to consider? A)Change the calibration of the horizontal and/or vertical axes to enhance apparent treatment effects B)Combine elements from two or more research designs to enhance the overall internal validity C)Question the external validity of the intervention D)Repeat the exact same study to see if the conclusions would be replicated

B)Combine elements from two or more research designs to enhance the overall internal validity When the conclusions of an experimental design are tenuous, one can consider combining elements from two or more designs to enhance the certainty of experimental control. If for example a changing criterion design led to questionable conclusions due to behavioral variability during sub-phases, a treatment withdrawal might be included as a control phase between changing criterion phases, to see if the effect of the reinforcement in sub-phases could be reversed when withdrawn.

What is one advantage of the changing criterion design? A)Very efficient: almost no pre-planning is required, and data are collected quickly B)Effective treatment need not be fully withdrawn to demonstrate a functional relation C)It can easily compare multiple behaviors, subjects, or settings on a single chart D)Highly variable behavior does not need to match changes in criterion in order for a functional relation to be demonstrated

B)Effective treatment need not be fully withdrawn to demonstrate a functional relation In a changing criterion design, target behavior levels must meet or exceed each criterion, in sequence, but do so in a relatively stable fashion to demonstrate a functional relation. Highly variable data weaken the internal validity of the results. It can be quite time consuming, both in the data collection period, and in the pre-intervention planning phase, when setting criterion jumps, variations in the length of sub-phases, determining when to insert a bi-directionality test, etc. This design does not compare multiple behaviors, subjects, or settings within a single experimental protocol (as is done in the multiple baseline design) -- only one behavior is evaluated on a chart, for one person at a time.

What is the most effective way to minimize the effect of multiple treatment interference as a confound? A)Enhance discriminability B)End with a single treatment in isolation C)Counterbalance the design D)Use an extended alone condition

B)End with a single treatment in isolation These are all good ways to reduce confounds in various experimental conditions. However, regardless of the counterbalancing of a design, or providing enhanced discriminability between phases, the multiple treatment phase will be unaffected per se. Only by separating out the treatments into phases where a single one of those treatments is tested in isolation can this confound be weeded out. We have not yet discussed the extended alone condition, but this device is used in functional analysis to separate out a possible carry-over/sequence effect.

Which graph illustrates the use of bi-directional change? A)Graph 1, below. B)Graph 3, below C)Graph 4 below D)Graph 2, below

B)Graph 3, below Only graph 3 exhibits horizontal level lines, indicators of changing behavior target levels, i.e., changing criteria, which represent the independent variable in this design. Be clear on this point: The IV is not the level of the behavior - the amount of the behavior is the dependent variable, right? Rather it is the changing criterion lines that represent changes in the independent variable: Each one defines a behavioral criterion, which must be obtained in order for the programmed consequence to be delivered. Then, when this criterion is reached a certain number of times, the next criterion in the sequence is introduced, as the new trigger for the consequence (typically, reinforcement) to be delivered.

What type of confound is exemplified by the graph, below A)Diffusion of treatment B)Multiple treatment interference C)Carry-over D)Counterbalancing

B)Multiple treatment interference Note how the behavior is very low for both of two treatments in phase 2, while in phase 3, the behavior is much higher for one of those treatment in isolation. This indicates the strong possibility of multiple treatment interference in phase 2. This is not carry-over. Counterbalancing is a way to control some confounds, and diffusion of treatment is an extraneous variable.

Identify what design characteristic is exemplified within the two ovals on the chart below A)Pre-treatment baseline B)Pre- and post-treatment probes C)Data points known as "outliers" D)Pre- and post-treatment baseline

B)Pre- and post-treatment probes The graph is of a multiple-probe design. The single data points in such a design are the probes, in this case after the treatment on bites, to track maintenance, and prior to the treatment on slaps, to loosely track the behavior prior to focusing on it for a full baseline (which begins on day 13).

"Assessment of behavior on occasions when the contingencies arranged in the analysis are not in effect" is a good definition of A)Parametric analysis B)Probes C)Sequential withdrawal D)A component analysis

B)Probes Probes can be conducted either before treatment, to explore aspects of the occurrence of the behavior prior to a full baseline, or after the treatment has been implemented, to see if treatment must be continued (e.g., in the midst of a fade out, can you simply terminate treatment rather than doing a full step-by-step fade). Component analysis evaluates the necessity for specific treatment components. Sequential withdrawal is a type of component analysis, where treatment components are sequentially withdrawn. Parametric analysis evaluates the effectiveness of different levels of a single treatment component.

The extent to which the independent variables are implemented as dictated by the research plan" is otherwise defined as: A)All of these are correct B)Procedural integrity C)Inter observer agreement D)Procedural drift

B)Procedural integrity The opposite of procedural integrity is procedural drift. "The extent to which the IV is implemented as written" would not be called "drift", despite their close relationship, since excessive drift reduces the extent to which the IV is implemented as written, namely, it reduces treatment integrity. Inter-observer agreement is a data collection measure, regarding how much two different observers of the same phenomenon agree as to the data collected on that phenomenon. Data collection is not a program procedure—but it is used to track the effectiveness of program procedures.

The most common types of confound in ABA design are: A)Final treatment phase, staggered introduction of the IV, multiple treatment interference B)Sequence effects, carryover effects, multiple treatment interference C)Counterbalancing, yoking, enhancing discriminability D)Extraneous variables, intrinsic variables, yoking

B)Sequence effects, carryover effects, multiple treatment interference Extraneous variables, intrinsic variables, final treatment phase, staggered introduction of the IV, counterbalancing, yoking, and enhancing discriminability are not confounds

This type of treatment evaluation process is used to assess maintenance of treatment effects in the absence of an increasing number of components in a treatment package: A)Parametric analysis B)Sequential withdrawal C)Probes D)Sequence analysis

B)Sequential withdrawal When a multiple component treatment is sequentially withdrawn, it can lead to the eventual elimination of the treatment - but in the meantime it identifies which components are not necessary and which still are. A parametric analysis would be used to determine the effective parametric values of contingencies (how much of a particular intervention is needed), not as a way ot assess the maintenance of treatment effects in the absence of the intervention. Post-treatment probes are measurements taken on the behavior of interest following the total discontinuation of formal treatment. A sequence analysis is used to identify sequences of events that are correlated with a target behavior, not to evaluate maintenance of treatment effects in the absence of the intervention.

When you find no functional relationship, but there actually is a functional relationship between the IV and the DV, this is considered a X , otherwise known as a Y . A)X = false positive Y = type 2 error B)X = false negative Y = type 2 error C)X = false positive Y = type 1 error D)X = false negative Y = type 1 error

B)X = false negative Y = type 2 error You find a functional relation when there is not one = false positive: type 1 error. You find NO functional relationship when there really is a functional relationship = false negative: type 2 error.

Which of the following graphic displays corresponds to a multi-element design? A)Graph 3, below B)Graph 1 below C)Graph 4 below D)Graph 2 below

C)Graph 4 below Note the multiple data paths on the single graph, and the tell-tale connection of data points separated by several sessions (during which other data points appear, but which are parts of distinct data paths). Also note the absence of a baseline (which is optional in this design)

Which of the following is not an advantage of multi-element designs? A)A baseline phase is optional B)Highly variable behavior can still be evaluated C)It is useful for behavior with a long latency to change D)It can quickly compare treatments or environmental manipulations

C)It is useful for behavior with a long latency to change Multi-element designs do not require the use of a baseline phase (although baselines may be used). This speeds up the assessment process. Data stability is not necessary prior to making a phase change, so the design can be used with highly variable behavior, and it can thus be used to rapidly compare treatments or other environmental manipulations. One limitation is that behavior must change quickly (e.g., from condition 1 in session 1 to condition 2 in session 2), because different conditions are rapidly alternated. Long latencies to change in the target behavior would cause severe data confusion across and within phases.

Which multi-element design variation has a baseline (i.e., no treatment) condition included only as one of the alternating conditions throughout the experiment: A)Multi-element with baseline plus a final treatment phase B)Multi-element with no alternating conditions C)Multi-element with no BL D)Multi-element with pre-treatment BL

C)Multi-element with no BL The multi-element with no baseline design can include as one of its conditions a "no treatment" condition, which is essentially a baseline condition that is not the initial distinct phase, but rather is alternated randomly with all the other conditions under consideration. Multi-element with pre-treatment BL and multi-element with baseline plus a final treatment phase are variations in which a distinct baseline phase is implemented prior to treatment, to which subsequent alternating treatment conditions are compared. The multi-element with no alternating conditions is not an experimental design.

To reduce threats to procedural integrity, one should do any or all of the following, except: A)Provide contingencies for program implementers relating to program fidelity B)Provide training and practice for implementers on a regular basis C)Simplify the dependent variable by shortening the target behavior definition D)Simplify the independent var

C)Simplify the dependent variable by shortening the target behavior definition The target behaviors are the target behaviors! You do not change those—except by arranging the environment in the proper way. Provide training to caregivers in procedures and data collection—but also provide feedback and contingencies for their maintaining procedural integrity and carrying out the program procedures as written. Simplify the independent variable as much as possible so the program procedures can be clearly understood and carried out easily, especially when the difficulties of the target behavior make program fidelity the most difficult to maintain.

All of the following could be effective ways to enhance discriminability when using an alternating treatment design, except: A)Provide instructions before sessions B)Include supplemental stimuli C)Use a strict sequence of alternating phases D)Reduce the number of conditions

C)Use a strict sequence of alternating phases Use of strict alternation does not assist in the discrimination as to which condition is in effect - but does potentially set up a very nice sequence effect! However, including supplemental stimuli, providing instructions before sessions, and reducing the number of conditions are all effective ways to enhance discriminability when using a multi-element (alternating treatment) design.

Which of the following statements are true about confounds? X = in ABA, confounds are usually related to the experimental design itself Y = the terms, "confounds" and "extraneous variables" mean exactly the same thing Z = confounds are typically discovered in data which have already been collected A)Y and Z only B)X and Y only C)X and Z only D)X, Y, and Z

C)X and Z only Confounds are NOT the same as extraneous variables. Extraneous variables are not related to the independent variable, but are external variables which can affect the DV, independent of the experimental design. Confounds are usually related to the structure of the independent variable (the way the design is implemented). Moreover, extraneous variables may be controlled prior to their occurrence, while confounds usually are defined by their appearance in data already collected (such as a carry-over effect or multiple treatment interference).

When structuring the sequence of conditions to be presented in a multi-element design, the experimenter should order those conditions based on... X = Randomization Y = Randomization with restrictions Z = Strict alternation of conditions A)Y or Z but not X B)Z only C)X or Y but not Z D)X only

C)X or Y but not Z Strict alternation can easily lead to a sequence effect, and even if it does not, a sequence effect cannot be ruled out if the same sequence of conditions is always presented. Randomization is pure, and can be used but can have a downside in that certain conditions may, by chance, not be tested very often, while other conditions may be repeatedly tested over and over. Randomization with restrictions is less pure, but insures that all conditions are tested equally, and that sequence effects are reduced, or at least exposed.

Which of the following are types of social validity? Social validity of... X = Goals and targets Y = Procedures Z = Results A)Y and Z only B)X and Y only C)X, Y, and Z D)X and Z only

C)X, Y, and Z Social validity is a broad concept which applies to the acceptability of procedures and outcomes to the individual and to those around the individual, as well as society at large. It also applies to the behaviors likely to be supported in the natural environment, so target behaviors must be those exhibited by typically performing individuals in the natural environment in order to be socially valid.

Experimental control is demonstrated in a changing criterion design when the performance (behavior)... A)changes whenever treatment is introduced, at different points in time, across different baselines B)reliably changes between phases when a specific treatment is repeatedly introduced or removed C)closely matches sequential changes in behavioral targets during sub-phases D)reliably changes, depending on different independent variables, which are rapidly manipulated across successive sessions

C)closely matches sequential changes in behavioral targets during sub-phases Rapid manipulation of different independent variables describes a multi-element (aka alternating treatment) design. Introducing treatment at different points of time across different baseline is the multiple baseline design. Repeatedly introducing and removing a treatment while taking repeated measures of behavior describes the withdrawal/reversal design.

Experimental design combinations are used for the following reason: Design combinations... A)offer more "exquisite" graphical displays B)typically require less time and resources C)enhances the certainty of experimental control when the conclusions of the "planned" designs are tenuous D) have a higher probability of being published in journals

C)enhances the certainty of experimental control when the conclusions of the "planned" designs are tenuous Experimental design combinations should be reserved for use when the experimenter must enhance the certainty of experimental control exhibited by the use of a single design. Publication is based on the quality of the conduct of a study and the analysis of the results, not on the complexity of the experimental protocols. Graphs from experimental design combinations may look elegant, but they are also often more complex and difficult in terms of communicating the results of ABA. These design combinations are likely to take more time and resources, rather than less.

Which of the following is not an important consideration when using a multi-element design? A)Order of conditions B)Counterbalancing C)Discriminability D)Behavioral variability

D)Behavioral variability One of the strengths of this design is that behavior variability does not delay change in conditions or spoil any analysis within and between conditions. This design can absorb the influence of (for example) extraneous variables - as long as clear and consistent differences remain between one data path, as it is differentially measured across all the phases of one condition, versus the data path that connects the data points measured in all the phases when another condition is in effect. Counterbalancing, discriminability, and order of conditions are all important considerations when using a multi-element design - all of which assist in the reduction of confounds and data confusion.

In a changing criterion design, each sub phase provides: A)A baseline for the following phase B)A means to gauge the effects of introducing or removing a different treatment C)A means to gauge the effects of changing the IV value from the previous phase D)Both a baseline for the following phase and a means to gauge the effects of changing the IV value from the previous phase

D)Both a baseline for the following phase and a means to gauge the effects of changing the IV value from the previous phase The independent variable is not introduced and removed across sub phases when using a changing criterion design. Each sub phases is used to predict patterns of behavior if a particular criterion were to be continued, and to evaluate if behavior changes when a new criterion is introduced.

What kind of design is represented by the graph below? A)Multi-element design B)Parametric Analysis C)Changing Criterion design D)Component Analysis

D)Component Analysis Note that we are examining the effects by isolating components of a treatment package which contains both edibles and praise. When presented together we see that it results in very high rates of the target behavior. Next we just use edibles alone and even though there is still some increase it is not as robust as its combination with praise. Next we just use praise alone. This component is rather ineffective when used in isolation of edibles.

Budget cutbacks require all behavior service plans to be streamlined to remove "less effective" treatment components. The funding source has determined that trimming down to minimal, but still effective programs will significantly reduce staff resource requirements (fewer program procedures require fewer staff minutes per day). Individual service plan outcomes must not change under this new plan. Which type of analysis is most likely to identify program procedures, which may be removed from the overall treatment package without compromising its effectiveness? A)Parametric analysis B)A multiple baseline design C)Structural analysis D)Component analysis

D)Component analysis A component analysis evaluates single components of a treatment package to determine its effectiveness, alone or in conjunction with other components. A parametric analysis would only look at various levels of a particular treatment, not the overall treatment itself. A structural analysis is another name for a functional analysis based on antecedent manipulations, and had not yet been introduced in this course (see later units). A multiple baseline design focuses on identifying the functional relationship between a behavior (a DV) and an intervention (an IV) but is not typically used for a component analysis since it looks at one intervention at time over different behaviors, subjects, or settings.

Advantages of using multi-element designs include all of the following, except: A)Useful for highly variable behavior B)May require a fewer number of sessions and less time C)Useful to compare treatments quickly D)Controls for multiple treatment interference

D)Controls for multiple treatment interference Multi-element designs may require a fewer number of sessions because they do not require behavior within a phase to become stable before changing to the next phase (and thereby demonstrate functional relations over time). In other words, highly variable behavior does not delay the change in phase, and this speeds up the process. This design can be used to rapidly compare different treatments. However, because of the facility of that comparison, this design is prone to multiple treatment interference

What experimental design is characterized by repeated measurement of behavior while different assessment or treatment conditions are alternated rapidly? A)Reversal B)Changing criterion design C)Multiple-baseline D)Multi-element

D)Multi-element Different assessment or treatment conditions are not alternated rapidly in a changing criterion, multiple-baseline, or reversal designs. This is done when using a multi-element design.

In this design variation, a no-treatment phase is followed by several treatments, rapidly alternated; this is followed by the introduction of an extended phase of a single one of those treatments. It is often used to evaluate the possibility of multiple treatment interference. A)Alternating treatment with no BL B)Multiple probe with pre-treatment BL C)Alternating treatment with a post-treatment BL phase D)Multi-element with baseline plus a final treatment phase

D)Multi-element with baseline plus a final treatment phase Multi-element design is another name for the alternating treatment design. Neither the alternating treatment with no BL nor the alternating treatment with a post-treatment baseline design begin with a no treatment phase, or end in treatment. The multiple probe design does not evaluate several alternating treatments at once—it only evaluates one treatment at a time, implemented in a staggered fashion across 3 different baselines.

In preparation for her meeting with the Psychiatrist, Alicia reviews Tierra's data comparing the effects of 2 mg of Haldol, 5 mg of Haldol, and 10 mg of Haldol on the rate of "agitated self-scratching" (defined in the program). Based on visual analysis, Alicia concludes that rate of "agitated self-scratching" was lowest at 2 mg of Haldol. Which type of analysis did Alicia do to evaluate the effects produced by the different dosages of the medication? A)Component Analysis B)Probe Analysis C)Sequential Withdrawal D)Parametric Analysis

D)Parametric Analysis A parametric analysis evaluates different parameters of a single intervention. In this example, the behavior analyst seeks to empirically ascertain the effects of different doses of a specific medication. This will allow decisions to be made about doses of Haldol based on its effectiveness, including titration.

Rob is on an intervention plan that includes access to his favorite music CD by Genesis contingent upon taking out the trash. His behavior analyst is trying to determine if 1 minute, 3-minutes, or 5-minutes of access will be effective in reliably maintaining Rob's taking out trash behavior. Which type of analysis should Rob's behavior analyst do to determine which amount of access to the music CD is most effective as a reinforcer? A)Sequential withdrawal B)Probe analysis C)Component analysis D)Parametric analysis

D)Parametric analysis A parametric analysis evaluates different parameters of a single intervention. In this example, the behavior analyst seeks to empirically ascertain the effects of differential durations of access to reinforcers and the effect on cooperation with taking out the garbage.

A multi-element design is characterized by: A)Introduction of an intervention at staggered points in time across behavior, participants, or settings B)Introduction and removal of an intervention across phases C)Introduction of increasing or decreasing target levels for behavior, sequentially across sub-phases D)Repeated measurement of behavior while different treatment/environmental conditions are rapidly alternated

D)Repeated measurement of behavior while different treatment/environmental conditions are rapidly alternated The key to the multi-element design (aka, alternating treatment design) is its rapid presentation of changing environmental conditions while repeated measurements of the behavior are taken. Introduction and removal of an intervention across phases is a withdrawal design. Introduction at staggered points in time across behaviors, participants, or setting represents a multiple baseline design. A changing criterion design is characterized by the introduction of increasing or decreasing target levels for behavior sequentially across sub-phases.

Which of the following sequence of conditions would not control for any potential sequence effects? A)Randomization B)All of these sequence arrangements would control for sequence effects C)Randomization with restrictions D)Strict alternation

D)Strict alternation When using a multi-element design, the experimenter should avoid using strict alternation because use of this technique can easily lead to sequence effects.

A multi-element design is least likely to be effective at rapidly establishing functional relationships between an IV and a DV when... A)trying to compare multiple interventions B)the individual has major perceptual problems C)the individual's behavior is highly variable D)time constraints demand rapid assessment results

B)the individual has major perceptual problems When individuals have problems discriminating environmental conditions (e.g., the person is deaf, or blind, or both) this design can be ineffective: when condition changes cannot be rapidly discriminated, discriminative control of those differing conditions may never occur, at least in the short duration test phases that generally characterize this design. In fact, one of the ways to IMPROVE this design is to "enhance discriminability"—so that changing conditions begin to exert discriminative control quickly, and the data reflect those rapid changes, fast. Multi-element designs do not require the stability of behavior to demonstrate functional relations, as long as consistent and clear differences can be perceived-by the subject—across conditions. It is the most appropriate design for highly variable behavior, since stability is not required between condition changes. This also means fewer sessions may be required since data stability is not necessary for phase change, so this design is most appropriate when time constraints exist. It is obviously the most appropriate for comparing different interventions -it is in the name of the design itself: alternating treatment design (aka multi-element design).

In a changing criterion design, each sub phase provides all of the following, except: A)All of these are provided by each sub phase of a changing criterion design B)A baseline for the following phase C)A means to gauge the effects of introducing or removing a completely different treatment D)A means to gauge the effects of changing the IV value from the previous phase

C)A means to gauge the effects of introducing or removing a completely different treatment The independent variable is not introduced and removed across sub phases when using a changing criterion design. Each sub phases is used to predict patterns of behavior if a particular criterion were to be continued, and to evaluate if behavior changes when a new criterion is introduced.

What type of confound, if any, does this chart most likely represent? (The Y-axis label may not be viewable on some computers. While not important to the question, the label reads: "Self hits") A)Multiple treatment interference B)Sequence effect C)Carry-over effect D)No confound, just regular data

C)Carry-over effect This MAY be a sequence effect, but there are not enough pairings of these two sequences to be sure. However, the data path appears to carry-over from the demand condition to the alone condition, so without being certain that there is a reliable sequence effect here, carry-over is evident (or at least very possible). There are not multiple treatments being evaluated here, at least by the information provided on the chart.

Which of the following is not a common experimental design combination? A)Combined reversal and multiple-baseline B)Combined reversal and multi-element C)Combined single-case and group design D)Combined multiple-baseline and multi-element

C)Combined single-case and group design Combined multiple-baseline and multi-element, combined reversal and multi-element, and combined reversal and multiple-baseline are all common experimental design combinations.

What experimental design combination is illustrated in the graph below? A)Combined multiple-baseline and multi-element B)Combined withdrawal and multiple-baseline C)Combined withdrawal and multi-element D)Combined changing criterion and reversal

C)Combined withdrawal and multi-element This is a little tricky, due to the apparently limited information included on the chart, but it clearly does not show a combination which includes either a multiple baseline design (there is no staggered introduction of the IV during multiple baseline data paths) or a changing criterion design (there are no criterion level lines). That leaves only one answer choice possible, which is a reasonable analysis of the designs apparent here.

When one intentionally varies non-essential environmental parameters within an experimental phase - e.g., time of day when the experiment is run, specific therapists, rooms, etc.— with the purpose of neutralizing potential confounds, this is known as: A)Using randomized order of conditions B)Enhancing discriminability C)Counterbalancing D)Using strict alternation

C)Counterbalancing Enhancing discriminability, using randomized order of conditions, and using strict alternation do not involve structured variation in non-essential environmental conditions.

Select a limitation of a changing criterion design from the list below: A)Discrimination is necessary to demonstrate functional relations B)Ethical considerations must be considered when effective treatment is withdrawn C)Difficult to interpret when behavior does not closely match criteria D)Inappropriate when behaviors are interdependent

C)Difficult to interpret when behavior does not closely match criteria Treatment is not withdrawn in this design, and only one behavior is studied, so interdependence does not pose a threat. Discrimination is not necessary in any design, but it does assist in the alternating treatment design at times—but not here. However, the behavior—to—criterion match is the lynch pin of the analysis of this design, and if that match is not clear, the design loses its interpretive strength.

Which of the following graphic displays exemplifies a changing criterion design? A)Graph #1 below B)Graph #4 below C)Graph #2 below D)Graph #3 below

C)Graph #2 below These graphs show a multiple baseline design, an alternating treatment design, an actual reversal design, and a changing criterion design.

Determining the social validity of a behavior intervention plan requires that the behavior analyst look at all of these aspects of the program, except A)The social significance of the goals - will this increase or decrease in behavior represent an improvement in the person's life B)The appropriateness of the procedures—are they practical, produce minimal side effects, and acceptable to society at large C)The social importance of the results to the individual—is the person better off now that their behavior has changed D)The degree to which the program procedures can apply to other social environments and other persons in society- do these procedures generalize

D)The degree to which the program procedures can apply to other social environments and other persons in society- do these procedures generalize This is a type of validity, but not social validity: This exception describes external validity, at least in part. Social validity and external validity are often confused. External validity is an experimental concept—to what degree do the results of this experiment apply to other experimental settings? Social validity is much more focused on the acceptability and efficacy of the program in social context, whether it involves the acceptance of procedures by the individual, caregivers, other stakeholders, and society at large - or - the actual results, and how whatever behavior change occurs assists the consumer in obtaining truly socially significant improvements in their life.

A changing criterion design is most appropriate when: A)The target response is highly variable across sub-phases B)The target behavior is completely absent from the person's repertoire C)Appropriate levels of the behavior must change, based on a changing natural environment D)The goal is to gradually increase or decrease some dimension of a target response

D)The goal is to gradually increase or decrease some dimension of a target response A changing criterion design would not be appropriate when behavior is highly variable, because this will require bigger changes in criteria to demonstrate experimental control, and the bigger the changes in criteria, the less replication across sub phases is possible. A changing criterion design is not typically used when the target behavior is not in the person's repertoire, because in this design the nature of the intervention remains constant—as does the topography of the response - just the amount of the behavior that triggers the intervention changes. New topographies are more typically addressed by prompt fading (antecedent manipulation) or shaping across topographies (consequence manipulation). A changing criterion design has nothing do to with changing criteria to match a natural environmental. This design is about strengthening (and less often, weakening) behavior.

A changing criterion design would NOT be appropriate when: A)A changing criterion design would not be appropriate in any of these conditions B)It is meaningful to measure behavior change in stepwise increments/decrements C)It is meaningful to demonstrate experimental control during fading and shaping D)The target behavior is highly variable

D)The target behavior is highly variable Highly variable data do not match well to a changing criterion level line. Moreover, selecting a criterion for change at each step would pose a very significant challenge - and even if chosen well, the variability of the data would still undercut the interpretive effectiveness of the design. Overlap of data sets between sub-phases would challenge any conclusions about experimental control. A changing criterion design would be appropriate when it is meaningful to measure behavior change in stepwise increments/decrements. In applying to slowly changing behavior, this design is useful for use during fading and shaping procedures.

When you find a functional relationship, even though there actually is NO functional relationship between the IV and the DV, this is considered a X , otherwise known as a Y . A)X = false negative Y = type 2 error B)X = false positive Y = type 2 error C)X = false negative Y = type 1 error D)X = false positive Y = type 1 error

D)X = false positive Y = type 1 error You find a functional relation when there is not one = false positive: type 1 error. You find NO functional relationship when there really is a functional relationship = false negative: type 2 error.

Which of the following statements are true about probes? Probes may be conducted X = prior to the implementation of a full baseline Y = during a treatment, to evaluate generalization and expedite a fade Z = following the end of treatment to evaluate maintenance A)Y and Z only, but not X B)X only C)X and Y but not Z D)X, Y, and Z

D)X, Y, and Z Probes may be done in all three cases, with probes before full baselines being one of the salient features in the distinction between a multiple-probe design and a multiple baseline design. Although Dr. DeLeon did not mention probes following treatment to evaluate maintenance, this use of probes is inferred in his discussion, and is more fully discussed in Cooper, p. 651.

Which of the following manipulations of the changing criterion design tends to strengthen the internal validity of the specific study? X = Keep the length of the phases as similar as possible Y = Use bi-directionality Z = Use 3 or more criterion changes, but not too many A)X only B)X, Y, and Z C)X and Y only D)Y and Z only

D)Y and Z only Bi-directionality strengthens internal validity by demonstrating behavior matching the set criterion line, whether that line represents an increase or decrease in relation to the prior sub-phase. Although 2 criterion changes are the minimum number required to distinguish this design from a simple A-B design, 3 or more changes demonstrate increasing evidence via replication of the behavior—to—criterion match. Too many changes however (of too small an incremental change) may muddy the data, making behavior change across several criterion changes an undifferentiated morass. Finally, phases of variable length, especially ones that run a little longer than they should, demonstrate once again that the behavior is tracking the criterion rather than just increasing or decreasing along a relatively steady trend (through phases of very similar lengths).