BEH 5043 Unit 6 Slides

Rev 03/22/23 CLG Copyright © ABA Technologies, Inc. 2020 BEH 5043 Unit 6 Slides p. 3 Objectives q Experimentation a. Define b. Identify two variables q Define independent variable q Define dependent variable q Given a scenario of an experiment, identify the independent and dependent variables 13 Experimentation q Basic strategy to collect and test information about the world 14 Two Variables in Experimentation q Independent variable (IV) q Dependent variable (DV) 15 Independent Variable (IV) Revisited q The variable that is manipulated § An environmental event/condition or stimulus class q AKA experimental variable 16 IVs in Behavior Analysis q Treatment q Intervention q Experimental condition 17 Independent Variable Example 1 q EAB researcher looking at effects of different schedules of reinforcement on the rate of pigeon key pecking in an operant chamber q Independent variable: Changes in the different schedules of reinforcement 18

Rev 03/22/23 CLG Copyright © ABA Technologies, Inc. 2020 BEH 5043 Unit 6 Slides p. 6 ASR q A behavior analyst is measuring the effects of feedback on the number of errors made when employees handle shipments. The DV is _____ and the IV is _____. 31 ASR q Dawson is a BCBA evaluating the effects of an intervention to increase students’ exercise. He gives students tokens based on the total distance they walk each day during gym class. The longer the distance, the more tokens they receive. The tokens can be exchanged for fun items at the school store. Initially, Dawson has the students walk without receiving any tokens or specific instructions; after a few days, he initiates the token system. After several days of running the experiment, Dawson learns that a student’s parents are also giving the student extra allowance for each day the student earns at least five tokens. 32 ASR (continued) q Which of the following is the dependent variable in this scenario? a. Money provided by the parents b. Total distance walked each session c. Tokens provided contingent on distance walked d. Fun items in the school store 33 ASR q Eli is a student who only willingly eats three foods: cake, French fries, and watermelon. The BCBA wishes to increase the variety of foods Eli will eat willingly. The BCBA implements a procedure where a very small amount of non- preferred food is presented to Eli. If he consumes the food, he immediately receives two bites of one of his preferred foods. Ten presentations per day of a non-preferred food occur. If Eli does not accept the bite after five seconds, the trial is ended. The BCBA monitors the percentage of successful bites out of the total number of trials each session. During this intervention Mom takes Eli to the dentist and Eli has three cavities filled. During the next several sessions, Eli refuses all foods and then again begins to initiate trying bites of non-preferred foods. 34 ASR (continued) q Which of the following is the independent variable in this scenario? a. The three cavities that were filled at the dentist b. The percentage of successful bites of non-preferred food consumed c. Access to preferred food contingent on consumption of non-preferred food d. The three preferred food items 35 Objectives q Experimental design a. Define b. Identify the goal q Define experimental control 36 The DV is number of errors and the IV is feedback

Rev 03/22/23 CLG Copyright © ABA Technologies, Inc. 2020 BEH 5043 Unit 6 Slides p. 9 Two Major Types of Validity q Internal validity q External validity 49 Internal Validity q Extent to which measured changes in the DV (behavior) are attributable to the independent variable manipulation and not some other factor 50 Internal Validity Example q Token system to reduce calling out and increase hand raising implemented with Greggory q If experimental design demonstrated IV changed DV, then good internal validity 51 External Validity q Extent to which the results of a study extend to other individuals, settings, or behaviors 52 External Validity Example q Token system to reduce calling out and increase hand raising now implemented with Greggory in a new classroom OR with other children 53 Importance of Internal Validity q Regarded as a priority over external validity q Generality of the effect (external validity) is unimportant if the effect wasn’t due to the independent variable, but rather some uncontrolled factor 54

Rev 03/22/23 CLG Copyright © ABA Technologies, Inc. 2020 BEH 5043 Unit 6 Slides p. 12 Testing q When changes in the dependent variable may have come about as a function of repeated exposure to the experimental arrangements 67 Testing Example 1 q “Practice effects” § Accuracy on a task improving as a function of repeated exposure before the intervention is introduced 68 Testing Example 2 q Wild mouse and Ts65Dn mouse and exposure to the novel object day after day § Is behavior change a result of the drug (“smart pill”) or repeated exposure to the novel object (learning)? 69 ASR q Other events or changes that coincide with the IV and could affect behavior are known as what kind of threat to internal validity? a. History b. Maturation c. Testing 70 ASR q Describe maturation: 71 ASR q What best describes testing as a threat to internal validity? a. A biased test b. Repeated performance on the same task c. Only having a small sample size d. Measurement systems are inaccurate 72 Growing older/stronger/healthier at the same time as IV is introduced Maturation involves physiological changes or naturally-occurring learning experiences that could coincide with the implementation of treatment.

Rev 03/22/23 CLG Copyright © ABA Technologies, Inc. 2020 BEH 5043 Unit 6 Slides p. 15 Selection Bias Example 1 q “Smart pill” evaluation § “Smart” group A (ceiling effect) § “Regular” group B (more improvement possible) 85 Selection Bias Example 2 q Recycling behavior study q Self-selection bias § Individuals who are prone to show greater improvement may also be more likely to participate in the study 86 ASR q Changes occurred because baseline measures were not representative of natural state of events = _____ q Outcome reflects a skewed assignment of participants into groups = ____ 87 Attrition q Loss of participants over time that may influence the effects of a study 88 Attrition Example q “Smart pill” evaluation and loss of participants due to improved performance 89 ASR q Attrition refers to which of the following? a. Seepage of the treatment to control conditions b. Loss of subjects over time c. Levels of the dependent variable returning to their natural state d. Skewed selection of subjects 90 regression to the mean selection bias

Rev 03/22/23 CLG Copyright © ABA Technologies, Inc. 2020 BEH 5043 Unit 6 Slides p. 18 ASR q A teacher has noticed that delivering attention for appropriate behavior was effective for one child in a study, and they begin to apply it to all children before intended. Which type of threat to internal validity is this? a. Maturity b. Testing c. Instrumentation d. Diffusion of treatment 103 Another Example of Selection Bias q “Opting in” to an experimental group 104 Another Example of Attrition q A company implements an intervention over year-long period q The study includes both permanent and seasonal employees 105 Threats to Internal Validity List Revisited 1. History 2. Maturation 3. Testing 4. Instrumentation 5. Diffusion of treatment 6. Regression towards the mean 7. Selection bias 8. Attrition 106 ASR q A company implements a new training program with the hopes of increasing employee sales. However, the company specifically only selects employees who are the top sellers with the hopes of making these employees’ sales skills even better. Specifically, picking employees who are only the top sellers is which potential type of threat? a. Attrition b. Selection bias c. Regression towards the mean d. History 107 ASR q Which of the following is the best example of attrition? a. A group study includes more experienced employees in the test condition b. Initial measures include data from inexperienced interns who are not working when post-intervention data are collected 108

Rev 03/22/23 CLG Copyright © ABA Technologies, Inc. 2020 BEH 5043 Unit 6 Slides p. 21 ASR (continued) q Which of the following is most likely to be an extraneous variable in this scenario? a. The use of a DRO procedure b. The frequency of items thrown c. The change from frequency of individual items thrown to frequency of throwing episodes d. The presence of the principal in the room 121 Objectives q Confounds a. Define b. Identify three types related to experimental design q Define multiple-treatment interference q Define carryover effects q Define sequence effects 122 Confounds q Uncontrolled variable known or suspected to exert influence on a dependent variable § Extraneous variables that affect the results of analysis become confounds q AKA confounding variables 123 Confounds Related to Experimental Designs 1. Multiple-treatment interference 2. Sequence effects 3. Carryover effects 124 Multiple-Treatment Interference q When results of an analysis are affected by the specific combination of two or more treatments 125 Multiple-Treatment Interference Example q Evaluate treatment A + treatment B = results C q Cannot determine the affect of A alone or B alone 126

Rev 03/22/23 CLG Copyright © ABA Technologies, Inc. 2020 BEH 5043 Unit 6 Slides p. 22 Sequence Effect q When changes in the dependent variable are attributable to the order of the experimental conditions q AKA order effects 127 Sequence Effect Example q Tx A, Tx B, Tx C q Would you get the same effect if the order was: Tx C, Tx A, Tx B? 128 Carryover Effect q When patterns of behavior established in one condition extend into a second condition even if the independent variables are very different 129 Carryover Example q Condition A = prob Bx high q Followed by Condition B = prob Bx high § Changes in the DV may be attributable to exposure to a previous condition, not the current IV 130 Cyclical Variability q Repeated patterns of responding over time unrelated to the independent variable q AKA cyclic variation 131 Cyclical Variability Examples q Hormonal fluctuations q Staff changes q Parent schedules q Institutional schedules (i.e., school or facility) (Barlow et al., 2009) 132

Rev 03/22/23 CLG Copyright © ABA Technologies, Inc. 2020 BEH 5043 Unit 6 Slides p. 23 ASR q What confound is identified when changes in the dependent variable are attributable to the order of the experimental conditions? a. Carryover effect b. Sequence effect c. Multiple-treatment interference d. Cyclical variability 133 ASR q An uncontrolled variable suspected to exert influence on a dependent variable is known as what? a. Confound b. Sequence effect c. Carryover effect d. Multiple-treatment interference 134 ASR q Which best describes multiple-treatment interference? a. Repeated patterns of responding over time unrelated to the independent variable b. When patterns of behavior established in one condition extend into a second condition even if the independent variables are very different c. When results of an analysis are affected by the specific combination of two or more treatments 135 ASR q Which best describes carryover effects? a. Repeated patterns of responding over time unrelated to the independent variable b. When patterns of behavior established in one condition extend into a second condition even if the independent variables are very different c. When results of an analysis are affected by the specific combination of two or more treatments 136 ASR q Which confounds are due to the manner in which an experiment is designed? q Multiple-treatment interference q Cyclical variability q Sequence effects q Carryover effects 137 Outline A. Introduction to Experimentation B. Internal and External Validity C.The Experimental Question D.Single-Case Designs E. The A-B Design F. Reversal Designs 138

Rev 03/22/23 CLG Copyright © ABA Technologies, Inc. 2020 BEH 5043 Unit 6 Slides p. 24 BACB Tasks 5 th ed TL D-6 Describe rationales for conducting comparative, component, and parametric analyses. 6 th ed TCO BCBA D.8. Identify rationales for conducting comparative, component, and parametric analyses. 6 th ed TCO BCaBA D.6. Distinguish among and implement comparative, component, and parametric analyses. (BACB, 2017, 2022a, 2022b) 139 “We conduct experiments to find out something we do not know” (Sidman, 1960, p. 214). 140 Research and Practice in ABA q No distinction between the two § Research: Systematic manipulations used to answer a research question § Practice: Systematic manipulations used to evaluate the effects of a treatment 141 Selecting the IV and DV q Basic researcher: Simple behaviors to observe and measure in controlled settings q Applied researcher: To answer a specific question about an IV q Applied clinician: A treatment that is likely to be effective with a behavior that has social significance 142 The Research Question q Specifies what the experimenter wants to know 143 Four Types of Analyses (Questions) 1. Demonstrative (AKA demonstration ) 2. Parametric 3. Component 4. Comparative 144

Rev 03/22/23 CLG Copyright © ABA Technologies, Inc. 2020 BEH 5043 Unit 6 Slides p. 25 Objective q Demonstrative analysis a. Define b. Identify examples and non- examples 145 Demonstrative Analysis q Analysis that seeks to determine the extent to which an intervention is effective § Question: “To what extent will this intervention work (i.e., change the behavior of interest)?” (Cooper et al., 2020, p. 158) 146 Demonstrative Analysis Examples q Will a specific error correction procedure improve student performance? q Will specific praise increase cleaning-up behavior? 147 ASR q How does experimental design for research differ from designs for practice? a. Designs options are more limited in research b. Design options are more limited in practice c. The essential goals are the same 148 ASR q What are the four types of analyses? q Demonstrative q Transitive q Parametric q Component q Collateral q Comparative 149 Demonstrative Analysis in the Literature Example 1 q Purpose: To evaluate ( demonstrate ) the effect of in situ training on gun safety q IV: In situ/behavioral skills training (BST) q DV: Rating scale 1–3 of behavior § 0 - touches the gun § 1 - doesn’t touch the gun § 2 - doesn’t touch the gun & leaves area § 3 - doesn’t touch the gun, leaves area, & tells an adult (Miltenberger et al., 2005) 150

Rev 03/22/23 CLG Copyright © ABA Technologies, Inc. 2020 BEH 5043 Unit 6 Slides p. 26 Demonstrative Analysis in the Literature Example 2 q Purpose: To evaluate (demonstrate) the effects of differential reinforcement procedures on food- guarding behavior q IV: Three separate differential reinforcement procedures q DV: “Resource guarding” (Mehrkam et al., 2020) 151 Demonstrative Analysis Rationale q To evaluate the effects of an intervention on behavior 152 Objective q Parametric analysis a. Define b. Identify examples and non- examples 153 Parametric Analysis q Analysis of the effects of various levels of an independent variable on behavior § Question: “Does more or less of the intervention work better?” (Cooper et al., 2020, p. 158) 154 Parametric Analysis Examples q The effects of differing values of a reinforcement schedule (e.g., FR 1 vs. FR 10) q Comparison of treatments at different “strengths” (e.g., brief time-out vs. long time-out) 155 Parametric Analysis Rationale q To determine effective parametric values of contingencies 156

Rev 03/22/23 CLG Copyright © ABA Technologies, Inc. 2020 BEH 5043 Unit 6 Slides p. 27 Parametric Analysis in the Literature Example 1 q Purpose: To evaluate the number of demands in a response interruption and redirection (RIRD) procedure on stereotypy q IV: RIRD procedure with three vs. one demand q DV: Stereotypy (Saini et al., 2015) 157 Parametric Analysis in the Literature Example 2 q Purpose: To evaluate the effects of varying lengths of delay to reinforcement on task performance q IV: Reinforcement provided immediately, or after one of six delays q DV: Completion of academic tasks (Leon et al., 2016) 158 ASR q Analysis of the effects of various levels of an independent variable on behavior = _____ q Analysis that seeks to determine the extent to which an intervention is effective = _____ 159 ASR q Bridget is testing out how different durations of break from task (15 s vs. 30 s) affect the rate of acquisition. Which type of analysis is she likely conducting? a. Demonstrative b. Parametric 160 ASR q Which is an example of a demonstrative analysis? a. Examining whether or not popcorn serves as a reinforcer b. Examining how different amounts of popcorn affect performance 161 Objective q Component analysis a. Define b. Identify examples and non- examples 162 parametric demonstrative

Rev 03/22/23 CLG Copyright © ABA Technologies, Inc. 2020 BEH 5043 Unit 6 Slides p. 28 Component Analysis q Analysis to identify what part(s) of a multi-part independent variable are necessary to produce behavior change § Question: “How effective is the intervention when various components are added or subtracted?” (Cooper et al., 2020, p. 158) 163 Component Analysis Example q Enzo and the many treatments: § ABA services § Special diet § Speech therapy § Floor time § Essential oils 164 Component Analysis Rationale q To “pull apart” an IV and identify necessary components of the intervention 165 Component Analysis in the Literature Example 1 q “Good-behavior game” package: 1) rule statements, 2) light box signal 3) differential reinforcement q Package = effective q Evaluated (component analysis) § Rules alone = ineffective § Rules + light box feedback = effective (Medland & Stachnik, 1972) 166 Component Analysis in the Literature Example 2 q Purpose: To evaluate components added to an electronic data collection (EDC) system on data collection timeliness q Conditions: § EDC alone § EDC + automated prompts § EDC + overall session feedback § EDC + specific interval feedback § EDC + guided selection (Morris & Peterson, 2020) 167 ASR q Which one describes a component analysis? a. Identifying what parts of a multipart IV are necessary to produce change b. Examining the effects of different IV values on DV 168

Rev 03/22/23 CLG Copyright © ABA Technologies, Inc. 2020 BEH 5043 Unit 6 Slides p. 29 ASR q Which is the best example of a component analysis? a. Evaluating how different lengths of delay to reinforcement affect performance b. Evaluating the use of reinforcement for replacement behavior, extinction, and prompting vs. reinforcement and prompting alone 169 Objective q Comparative analysis a. Define b. Identify examples and non- examples 170 Comparative Analysis q Analysis of the differential effects of two or more independent variables on the dependent variable § Question: “Do the two different interventions produce differential effects?” (Kennedy, 2005, p. 68) 171 Comparative Example q Traditional approach vs. error correction approach to learning sight words 172 Comparative Analysis Rationale q To determine which of a given number of procedures is most effective 173 Comparative Analysis in the Literature Example 1 q Purpose: To compare two treatments , DRO and DRA, for treating behavior maintained by automatic reinforcement q IV: DRO vs. DRA q DVs: Stereotypy, engagement, productivity (Hedquist & Roscoe, 2020) 174

Rev 03/22/23 CLG Copyright © ABA Technologies, Inc. 2020 BEH 5043 Unit 6 Slides p. 30 Comparative Analysis in the Literature Example 2 q Purpose: To compare the effects of two interventions on child compliance q IVs: Guided compliance procedure vs. time-out procedure q DV: Compliance (Handen et al., 1992) 175 ASR q ______ examines whether or not an IV has an effect. q ______ examines if one IV has a greater effect than another. 176 ASR q Examining if it is more effective to prompt through an entire sequence of steps or use shaping would likely require which type of analysis? a. Comparative b. Demonstrative 177 ASR q Naomi would like to determine the impact self-monitoring will have on decreasing skin-picking behavior. This is an example of which type of analysis? a. Parametric analysis b. Component analysis c. Demonstrative analysis d. Comparative analysis 178 ASR q The BCBA is interested in improving her RBTs’ proficiency in data collection procedures. The BCBA is experimenting to evaluate if a token economy or a monthly bonus system will be more effective at improving the target behavior. This is an example of which type of analysis? a. Demonstrative analysis b. Comparative analysis c. Parametric analysis d. Component analysis 179 ASR q Xavier receives a packaged intervention to reduce severe self-injurious behavior. Initially the BCBA focused on teaching a functionally equivalent replacement behavior. To further reduce the self-injurious behavior, a contingent-effort consequence was introduced following each instance of the behavior. Lastly, the BCBA adds a DRO procedure, all the while monitoring the rate of self-injurious behavior. The BCBA then isolates each intervention to determine which is most effective. This is an example of which type of analysis? a. Demonstrative analysis b. Parametric analysis c. Component analysis d. Comparative analysis 180 Demonstrative Analysis Comparative Analysis

Rev 03/22/23 CLG Copyright © ABA Technologies, Inc. 2020 BEH 5043 Unit 6 Slides p. 31 ASR q The BCBA is implementing a procedure to maintain Joe’s appropriate behavior of answering teaching questions in class. The BCBA tests which schedule—a VR 2, VR 4 or VR 8—is sufficient to maintain the behavior. This is an example of which type of analysis? a. Parametric analysis b. Demonstrative analysis c. Component analysis d. Comparative analysis 181 Outline A. Introduction to Experimentation B. Internal and External Validity C.The Experimental Question D.Single-Case Designs E. The A-B Design F. Reversal Designs 182 BACB Tasks 5 th ed TL D-3 Identify the defining features of single- subject experimental designs (e.g., individuals serve as their own controls, repeated measures, prediction, verification, replication). 6 th ed TCO BCBA D.4. Identify the defining features of single- case experimental designs (e.g., individuals serve as their own controls, repeated measures, prediction, verification, replication). 6 th ed TCO BCaBA D.3. Identify the defining features of single- case experimental designs (e.g., individuals serve as their own controls, repeated measures, prediction, verification, replication). (BACB, 2017, 2022a, 2022b) 183 BACB Tasks 5 th ed TL D-4 Describe the advantages of single- subject experimental designs compared to group designs. 6 th ed TCO BCBA D.5. Identify the relative strengths of single-case experimental designs and group designs. 6 th ed TCO BCaBA D.4. Identify strengths of single-case experimental design. (BACB, 2017, 2022a, 2022b) 184 Objective q Group designs a. Define b. Identify features 185 Two Categories of Experimental Designs 1. Group 2. Single-case 186

Rev 03/22/23 CLG Copyright © ABA Technologies, Inc. 2020 BEH 5043 Unit 6 Slides p. 32 Group Design q Experimental design in which each experimental and control condition includes a different set of individuals q AKA between-subjects design , large N design 187 Control in Group Designs q Comparisons made between groups of individuals § Control group vs. experimental group 188 Exposure to Independent Variables in Group Designs q Each individual often exposed to only one value of the IV § Control (no treatment or intervention) vs. treatment/intervention 189 Number of Participants in Group Designs q Typically large number of subjects q Few (often single) observations of the DV 190 Data Analysis in Group Designs q Usually inferential statistics § Measures of variance and central tendency 191 Independent Variable Introduction in Group Designs q Changes in IV assigned according to randomized or matched designs q When to introduce is clearly defined § E.g., exactly 1 week into the experiment, Group A receives IV manipulation (drug), and Group B receives the control manipulation (placebo pill) 192

Rev 03/22/23 CLG Copyright © ABA Technologies, Inc. 2020 BEH 5043 Unit 6 Slides p. 33 Mechanism of Generalization in Group Designs q Random selection from a population q Random assignment to conditions 193 Group Design Example q Drug evaluation: § Placebo (control) § Drug A § Drug B 194 Group Design Single-Case Design Control Between groups IV Exposure 1 condition Numbers Many Data Analysis Statistics IV Introduction Randomized & matched Mechanism of Generalization Random selection & assignment 195 ASR q What are the common features of a group design? q Comparisons made between groups of individuals q Each individual observed several times q Random assignments to conditions q Use of inferential statistics q Typically small numbers of participants 196 Objective q Single-case experimental designs a. Define b. Identify features 197 Single-Case (SC) Experimental Designs q Experimental designs characterized by repeated measures of a small number of participants, within-subject comparisons, use of steady state strategy, and visual analysis of graphed data 198

Rev 03/22/23 CLG Copyright © ABA Technologies, Inc. 2020 BEH 5043 Unit 6 Slides p. 34 Single-Case Designs AKA q Single-subject designs q Within-subject designs q Small-n or N=1 designs q Time series designs q Intensive designs q Intra-subject replication designs 199 History of Single-Case Designs q Wundt (late 1800s): Perception via introspection q Ebbinghaus (late 1800s): Memory using himself q Pavlov (late 1800s–early 1900s): Respondent conditioning in dogs q Watson (early 1900s): Development of phobias in individual children 200 History of Single-Case Designs (continued) q Thorndike (early 1900s): Operant conditioning using a few animals at a time q Bekhterev (early 1900s): Respondent conditioning using a few human participants q Skinner and others develop EAB (mid- 1900s): Investigate nonhumans under highly controlled conditions q Formalization of ABA (1968): JABA 201 ASR q What might be a rationale for using the term single-case design as opposed to single-subject or within- subject? a. The term “case” allows for non- human participants b. These designs can be used with more than one subject 202 Some Identifying Features of Single-Case Designs q Individuals serve as their own control q Repeated measures with a small number of subjects q Replication q Visual analysis q Stability of DV before IV changes § Prediction and verification 203 Some Identifying Features of Single-Case Designs q Individuals serve as their own control q Repeated measures with a small number of subjects q Replication q Visual analysis q Stability of DV before IV changes § Prediction and verification 204

Rev 03/22/23 CLG Copyright © ABA Technologies, Inc. 2020 BEH 5043 Unit 6 Slides p. 35 Individuals Serve as Their Own Control q Each individual experiences every value of the independent variable, and comparisons are made within each individual 205 Control in Single-Case Designs q Comparisons made within individuals § Individuals serve as their own control § Before-IV implementation compared to after-IV implementation with the same individual 206 Exposure to Independent Variables in Single-Case Designs q Each individual is exposed to each level of the IV § E.g., baseline and treatment(s) 207 Some Identifying Features of Single-Case Designs q Individuals serve as their own control q Repeated measures with a small number of subjects q Replication q Visual analysis q Stability of DV before IV changes § Prediction and verification 208 Number of Participants in Single-Case Designs q Small number of subjects q Multiple, repeated observations 209 Repeated Measures Minimize Threats to Internal Validity q Data are collected on the DV over an extended period of time q Helps to rule out “other” factors that may influence behavior (e.g., testing, regression, maturation) 210

Rev 03/22/23 CLG Copyright © ABA Technologies, Inc. 2020 BEH 5043 Unit 6 Slides p. 36 ASR q Whereas group designs compare a group of individuals receiving treatment to a control group without treatment, single- case designs typically measure an individual receiving treatment and compare to what? a. A statistical average as a control b. A measure of the same individual at a point under different conditions 211 ASR q How is internal validity established in single-case design? a. Taking the same measure from many individuals b. Repeated measures of the same individual c. Random assignment to interventions to reduce selection bias 212 Some Identifying Features of Single-Case Designs q Individuals serve as their own control q Repeated measures with a small number of subjects q Replication q Visual analysis q Stability of DV before IV changes § Prediction and verification 213 Replication q Demonstration using multiple cases § Within individuals § Across individuals 214 Minimizing Threats to Internal Validity: Replication q If the IV affects the same individual in the same way each time or affects many subjects in the same manner, a stronger case can be made that the IV produced the change 215 Mechanism of Generalization in Single-Case Designs q Replication allows for evaluation of generality of effects 216

Rev 03/22/23 CLG Copyright © ABA Technologies, Inc. 2020 BEH 5043 Unit 6 Slides p. 37 Some Identifying Features of Single-Case Designs q Individuals serve as their own control q Repeated measures with a small number of subjects q Replication q Visual analysis q Stability of DV before IV changes § Prediction and verification 217 Data Analysis in Single-Case Designs q Usually visual analysis of graphed data q Changes in level, trend, or variability as a function of the change in the independent variable 218 Some Identifying Features of Single-Case Designs q Individuals serve as their own control q Repeated measures with a small number of subjects q Replication q Visual analysis q Stability of DV before IV changes § Prediction and verification 219 ASR q Which best describes replication? a. Number of sessions exposed to the IV within a single condition b. When multiple journals publish the same study c. Demonstrating the change either across multiple individuals or multiple times with the same individual 220 ASR q Which are advantages of replication? q More time-efficient q Can promote generalization q Minimizes threats to internal validity q Helps ensure social validity 221 ASR q Which characteristics are we commonly looking for when visually analyzing graphed data? a. IV consistently applied across conditions b. Changes in level, trend, and variability attributable to the IV c. Changes in measure of the DV spreading to untreated conditions d. Clear statistical difference in measures of the DV 222

Rev 03/22/23 CLG Copyright © ABA Technologies, Inc. 2020 BEH 5043 Unit 6 Slides p. 38 Independent Variable Introduction in Single-Case Designs q Changes in independent variable made once dependent variable displays stability q Known as baseline logic or steady state strategy 223 Steady State q When a pattern of responding shows little variation within a given condition over time 224 Minimizing Threats to Internal Validity: Stability q Stability prior to IV and change with IV implementation increases confidence that Bx change can be attributed to the IV 225 Group Design Single-Case Design Control Between groups Within individuals IV Exposure 1 condition All conditions Numbers Many Few people, many observations Data Analysis Statistics Visual analysis IV Introduction Randomized & matched When DV is stable Mechanism of Generalization Random selection & assignment Replication 226 ASR q What is the advantage of ensuring a steady state before making changes to the IV? a. Strengthens the confidence that effects on the DV are due to the IV b. Increases the likelihood that the DV impacts the IV c. It can expedite the process of demonstrating control 227 Some Advantages of Single-Case Designs q Repeated measurement q Inter-subject variability q Intra-subject variability q Clinical accountability 228

Rev 03/22/23 CLG Copyright © ABA Technologies, Inc. 2020 BEH 5043 Unit 6 Slides p. 39 SC Design Advantage: Repeated Measurement q Repeated measurement § Increases confidence that IV caused change in DV § Steady state § Permits investigation of behavior as dynamic process (identification of patterns of behavior/behavior change) 229 SC Design Advantage: Inter- Subject Variability q Examination of differences between individuals 230 SC Design Advantage: Intra- Subject Variability q Examination of the differences in responding within the same individual § Idiosyncratic effects § Serendipitous findings 231 SC Design Advantage: Clinical Accountability q Clinical accountability § Subject serves as own control § Visual analysis reveals socially significant changes 232 ASR q Which are advantages to single-case design? q Ensures changes of statistical significance are identified q Increases likelihood the change is socially significant in size q Repeated observations of the individual of interest q Can verify that the intervention was applied as intended 233 ASR q Which design is more likely to identify idiosyncratic effects? a. Group b. Single-case 234

Rev 03/22/23 CLG Copyright © ABA Technologies, Inc. 2020 BEH 5043 Unit 6 Slides p. 40 Objective q Define baseline 235 Baseline (BL) q Measurement of the dependent variable prior to the implementation of the independent variable q AKA baseline condition 236 Baseline and Treatment q Baseline can mean the absence of treatment q Baseline does not necessarily mean the absence of treatment 237 1 2 3 4 5 6 7 8 9 10 Response per Unit time 0 1 2 3 4 5 6 7 8 9 10 11 12 Horizontal axis (x-axis, abscissa): Usually represents passage of time and value of the IV Graph Review Sampling Units (e.g., Sessions, Days, etc.) 238 1 2 3 4 5 6 7 8 9 10 Response per Unit time 0 1 2 3 4 5 6 7 8 9 10 11 12 Horizontal axis (x-axis, abscissa): Usually represents passage of time and value of the IV Vertical axis (y-axis, ordinate): Usually represents values of the DV Graph Review Sampling Units (e.g., Sessions, Days, etc.) 239 1 2 3 4 5 6 7 8 9 10 Response per Unit time 0 1 2 3 4 5 6 7 8 9 10 11 12 Baseline Example Sampling Units (e.g., Sessions, Days, etc.) 240

Rev 03/22/23 CLG Copyright © ABA Technologies, Inc. 2020 BEH 5043 Unit 6 Slides p. 41 Functions of Baseline q Description q Prediction 241 Descriptive Function of Baseline q Information about extent of the problem q Indication of whether services are necessary q Insight into relevant environmental events q Useful for setting target outcomes 242 Predictive Function of Baseline q Predicts future level of behavior in the absence of the IV or if the IV has no effect q Serves as a criterion to evaluate whether the intervention produces change 243 ASR q Baseline always refers to no treatment in place. a. True b. False 244 ASR q Which can be seen as functions of baseline measurement? q Gathering information about the extent of the problem q Ensuring the treatment is applied as intended q Maximizing the effectiveness of the intervention q Indicating if services are necessary 245 Objectives q Phase change a. Define b. Identify examples of when to change phases q Baseline logic a. Define b. Use to make predictions of future levels of behavior 246

Rev 03/22/23 CLG Copyright © ABA Technologies, Inc. 2020 BEH 5043 Unit 6 Slides p. 42 1 2 3 4 5 6 7 8 9 10 Response per Unit time 0 1 2 3 4 5 6 7 8 9 10 11 12 Baseline Example Sampling Units (e.g., Sessions, Days, etc.) 247 Phase Change q A change from one set of environmental conditions to another 248 1 2 3 4 5 6 7 8 9 10 Response per Unit time 0 1 2 3 4 5 6 7 8 9 10 11 12 Baseline Treatment Phase change line Sampling Units (e.g., Sessions, Days, etc.) 249 1 2 3 4 5 6 7 8 9 10 Response per Unit time 0 1 2 3 4 5 6 7 8 9 10 11 12 Baseline Treatment Predicted level of the DV if no IV change or if IV has no effect Sampling Units (e.g., Sessions, Days, etc.) 250 1 2 3 4 5 6 7 8 9 10 Response per Unit time 0 1 2 3 4 5 6 7 8 9 10 11 12 Baseline Treatment Observed level of the DV when IV is changed Sampling Units (e.g., Sessions, Days, etc.) 251 Verification q Compare predicted path with actual path 252

Rev 03/22/23 CLG Copyright © ABA Technologies, Inc. 2020 BEH 5043 Unit 6 Slides p. 43 1 2 3 4 5 6 7 8 9 10 Response per Unit time 0 1 2 3 4 5 6 7 8 9 10 11 12 Baseline Treatment Difference between prediction and actual effects Sampling Units (e.g., Sessions, Days, etc.) 253 Baseline Logic q The difference between predicted and actual effects § Prediction § Verification § Replication 254 q In consideration of baseline data illustrated below, which continuing data path (see below) best predicts future problem behavior in the absence of intervention? a. = b. = c. = Problem Behavior 0 2 4 6 8 10 12 1 2 3 4 5 6 7 8 9 10 Session Problem Behaviors / Min Baseline Problem Behavior 0 6 12 1 2 3 4 5 6 7 8 9 10 Session Problem Behaviors / Min Problem Behavior 0 6 12 1 2 3 4 5 6 7 8 9 10 Session Problem Behaviors / Min Problem Behavior 0 6 12 1 2 3 4 5 6 7 8 9 10 Session Problem Behaviors / Min Baseline Baseline ASR 255 ASR 2 4 3 1 256 ASR (continued) q Considering the baseline data shown, which continuing data path best predicts future problem behavior in the absence of an intervention? a. Graph 1 b. Graph 2 c. Graph 3 d. Graph 4 257 Steady State Strategy q During an experimental analysis, the requirement that behavior must reach a steady state prior to making changes in the independent variable 258

Rev 03/22/23 CLG Copyright © ABA Technologies, Inc. 2020 BEH 5043 Unit 6 Slides p. 44 Steady State Within Phases q Phase change logic q Guides our decision about when to change phases 259 Steady State Responding q Level : Behavior is high or low enough that you will be able to detect a change if one occurs q Stability : Levels of behavior do NOT vary greatly from one measurement to the next q Trend : The behavior is NOT already changing in the direction predicted for treatment 260 0 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 Reduction Target: Inadequate Level Example Weeks Mean Responses per Day Baseline Treatment 261 1 2 3 4 5 6 7 8 9 10 Response per Unit time 0 1 2 3 4 5 6 7 8 9 10 11 12 Baseline Reduction Target: Adequate Level Example Sampling Units (e.g., Sessions, Days, etc.) 262 Steady State Responding q Level : Behavior is high or low enough that you will be able to detect a change if one occurs q Stability : Levels of behavior do NOT vary greatly from one measurement to the next q Trend : The behavior is NOT already changing in the direction predicted for treatment 263 Weeks 0 2 4 6 8 10 12 14 16 18 Mean Responses per Day 0 10 20 30 40 50 Baseline Treatment Stable BL Data Example 264

Rev 03/22/23 CLG Copyright © ABA Technologies, Inc. 2020 BEH 5043 Unit 6 Slides p. 45 Weeks 0 2 4 6 8 10 12 14 16 18 Mean Responses per Day 0 10 20 30 40 50 Baseline Treatment Unstable BL Data Example 265 ASR q What features indicate steady state responding? q Variable data that include measures near post-treatment goals q Levels near post-treatment goals q Levels of behavior show little change from one point to the next q Behavior is different enough from post- treatment goal to identify change 266 Steady State Responding q Level : Behavior is high or low enough that you will be able to detect a change if one occurs q Stability : Levels of behavior do NOT vary greatly from one measurement to the next q Trend : The behavior is NOT already changing in the direction predicted for treatment 267 Weeks 0 2 4 6 8 10 12 14 16 18 Mean Responses per Day 0 10 20 30 40 50 Baseline Treatment Goal is to decrease levels of the behavior Usable Trend Example 268 Weeks 0 2 4 6 8 10 12 14 16 18 Mean Responses per Day 0 10 20 30 40 50 Baseline Treatment Goal is to decrease levels of the behavior Unusable Trend Example 269 Basic Rules of Thumb for Trending Data q Upward trend: § Okay to change phases if it is a behavior targeted for decrease § Not okay if it is a behavior targeted for increase q Downward trend: § Okay to change phases if it is a behavior targeted for increase § Note okay if it is a behavior targeted for decrease 270

Rev 03/22/23 CLG Copyright © ABA Technologies, Inc. 2020 BEH 5043 Unit 6 Slides p. 46 Length of Baseline q The longer the baseline, the greater the predictive power q Avoid set numbers q Minimums are okay and shoot for at least three q As long as necessary, as short as possible 271 ASR q What is generally advisable regarding data that are trending in one direction during baseline? a. Never intervene if baseline data show a trend b. Intervention is acceptable regardless of baseline trends c. Only intervene if the trend is the opposite direction of intended change 272 ASR q How long should one record baseline measures? 273 ASR 1 2 3 4 274 ASR (continued) q For which of the baseline graphs shown would it be most appropriate to introduce the independent variable when our goal is to reduce the behavior? a. Graph 1 b. Graph 2 c. Graph 3 d. Graph 4 275 Uses of Steady State Strategy q Evaluate measurement decisions q Reveal the effects of conditions q Evaluate extraneous influences q Facilitate comparisons 276 As long as necessary to establish trend, level and variability, but as short as possible so as not to delay treatment”

Rev 03/22/23 CLG Copyright © ABA Technologies, Inc. 2020 BEH 5043 Unit 6 Slides p. 47 Evaluate Measurement Decisions q Evaluate operational definitions q Evaluate measurement practices 277 Reveal the Effects of Conditions q Reveal changes in behavior that are characteristic of the condition 278 Evaluate Extraneous Influences q Identify variability uncharacteristic of the condition q Evaluate excessive variability within a condition 279 Facilitate Comparisons q Between BL and intervention conditions q Understanding patterns within conditions facilitates comparisons between them 280 ASR q How do steady states help evaluate extraneous factors? a. By identifying uncharacteristic data within a condition b. By showing a difference between DV measures across different conditions c. By ensuring the measurement practice is the same 281 ASR q How does the steady state strategy help in determining response definitions? a. Unclear definitions often result in unstable data which are detected when looking for a steady state b. Unclear definitions provide more stable data by accepting more variations in responding 282

Rev 03/22/23 CLG Copyright © ABA Technologies, Inc. 2020 BEH 5043 Unit 6 Slides p. 48 Outline A. Introduction to Experimentation B. Internal and External Validity C.The Experimental Question D.Single-Case Designs E. The A-B Design F. Reversal Designs 283 BACB Tasks 5 th ed TL D-5 Use single-subject experimental designs (e.g., reversal, multiple baseline, multielement, changing criterion). 6 th ed TCO BCBA D.7. Distinguish among reversal, multiple- baseline, multielement, and changing-criterion designs. D.9. Apply single-case experimental designs. 6 th ed TCO BCaBA D.5. Distinguish among and implement single- case experimental designs (e.g., reversal, multiple baseline, multielement, changing criterion). (BACB, 2017, 2022a, 2022b) 284 Objective q A-B design a. Define b. Identify relation to all other experimental designs c. Identify examples and non- examples 285 Letter Conventions q Different phases get different letters q Letters typically indicate a sequence q Baseline is usually “A” § Intervention 1, “B,” intervention 2, “C”… 286 A-B Experimental Design q Experimental design in which a single baseline phase is followed by a single intervention phase 287 A-B Design q Least complex experimental design q Serves as a basis for all other designs 288

Rev 03/22/23 CLG Copyright © ABA Technologies, Inc. 2020 BEH 5043 Unit 6 Slides p. 49 A-B Design Phases q Two phases 1. A: Baseline 2. B: Intervention 289 A-B Design Procedure 1. A: Baseline § DV measured repeatedly § Looking for steady state in DV § Predict data path 2. B: Intervention § Introduce IV & continue to measure DV § Compare predicted path with actual path 290 A-B Design Logic q Effect is demonstrated when behavior changes from baseline (A) to intervention (B) 291 A-B Design: Depiction Weeks 0 5 10 15 20 Mean Responses per Day 0 10 20 30 40 50 Baseline Treatment 292 A-B Design Example q Target: Rate of hand raising and callouts in morning block q Baseline: No intervention q Treatment: Token system 293 ASR q Which best describes an A-B design? a. A complex design intended to strongly verify even slight differences b. The simplest design on which other strategies are based c. Uses two groups, one receiving treatment and one without d. Does not require a baseline measure 294

Rev 03/22/23 CLG Copyright © ABA Technologies, Inc. 2020 BEH 5043 Unit 6 Slides p. 50 When to Change Phases: Number q Avoid set number of data points q Minimums are okay § At least three data points needed to identify a trend q Regardless of length, always consider the last three data points as a small trend 295 When to Change Phases: Stability Criteria q A rule for determining if the trend is sufficiently stable 296 ASR q Why is it not advisable to solely base the introduction of the intervention on data that will show the greatest change? a. We want to demonstrate the truth of the change, not simply verify our hypotheses b. Change might actually be more difficult to demonstrate c. The demonstrated change might be lower than it actually is 297 ASR q How do we ensure we are intervening in a way that shows accuracy of the intervention rather than showing bias of our hypothesis? a. Adjust stability criteria as data are being collected b. Use a fixed number of data points for each phase c. Set stability criteria before taking baseline data 298 Stability Criteria Example q No more than 25% deviation in last three points q No trend in any direction in last three points 299 Visual Analysis of A-B Data q Changes in: § Level § Trend § Variability q Immediacy of change § More immediate the effect, stronger the case the IV produced it 300

Rev 03/22/23 CLG Copyright © ABA Technologies, Inc. 2020 BEH 5043 Unit 6 Slides p. 51 A-B Design Advantages q Serves as a basis for all other designs q Useful in practice when more complex designs are not possible 301 A-B Design Limitations q Weak internal validity q Causal conclusions are not possible § Correlational only § Unable to rule out extraneous factors 302 ASR q What is more desirable in terms of how quickly the DV changes after the intervention is in place? a. Longer latency to change strengthens confidence that the IV produced that change b. Shorter latency to change strengthens confidence that the IV produced that change 303 ASR q Why are longer changes in the DV following the IV less desirable? a. If behavior does not change immediately, the IV cannot be responsible for the change b. Longer changes make it harder to eliminate extraneous variables c. Longer experimental designs are most costly 304 ASR q When would it be most appropriate to select an A-B design? a. Other designs are not possible b. You must eliminate as many extraneous variables as possible c. The most convincing design is required 305 JOBM Reports From the Field q “Data-based case studies that describe the application of OBM principles in organizational settings are valuable to the ongoing development of the field” (Houmanfar, 2014, pp. 3–4). 306

Rev 03/22/23 CLG Copyright © ABA Technologies, Inc. 2020 BEH 5043 Unit 6 Slides p. 52 A-B in the Literature q Purpose: Evaluate the effect of a process design intervention on setup time of welding employees q IV: Process design q DV: Welder setup time q Design: A-B (Blasingame et al., 2014) 307 (p. 216) Blasingame et al. (2014) Results 308 A-B Design Use q Recommended only when other, more compelling, designs are untenable 309 ASR q Since A-B designs do not rule out threats to internal validity, these designs are not used in behavior- analytic practice. a. True b. False 310 Outline A. Introduction to Experimentation B. Internal and External Validity C.The Experimental Question D.Single-Case Designs E. The A-B Design F. Reversal Designs 311 Objective q Reversal design a. Define b. Identify examples and non- examples c. Identify examples of when to use 312

Rev 03/22/23 CLG Copyright © ABA Technologies, Inc. 2020 BEH 5043 Unit 6 Slides p. 53 A-B Design Revisited q Experimental design in which a single baseline phase is followed by a single intervention phase 313 Reversal Design q Experimental design in which baseline and experimental conditions alternate q AKA withdrawal design , A-B-A design , A-B-A-B design 314 Phases in Reversal Design q At least three phases 1. A: Baseline 2. B: Intervention 3. A: Baseline 4. B: Intervention 315 Reversal Design Basic Procedure (A-B-A) 1. A : Collect baseline (BL) data (A) until stability is demonstrated 2. B : Introduce IV and continue to collect data until stability is demonstrated 3. A : Return to BL with withdrawal of the IV, continue to collect data 316 ASR q What is the minimum number of phases needed for a reversal or withdrawal design? a. One b. Two c. Three d. Four 317 Reversal Design Basic Procedure (A-B-A-B) 1. A : Collect baseline (BL) data (A) until stability is demonstrated 2. B : Introduce IV and continue to collect data until stability is demonstrated 3. A : Return to BL with withdrawal of the IV, continue to collect data 4. B : Reintroduce IV and continue to collect data until stability is demonstrated 318

Rev 03/22/23 CLG Copyright © ABA Technologies, Inc. 2020 BEH 5043 Unit 6 Slides p. 54 A-B-A vs. A-B-A-B q A-B-A: Only one reversal § Not preferred q A-B-A-B: Two reversals § Preferred § Strengthens demonstration of experimental control § Allows for ending on Tx condition 319 Reversal Design Phase Changes q Avoid rules for number of data points § Minimum number okay to define q Stability in previous phase q Look at last three data points as a small trend q Intervene only when no trend or trend in opposite direction of intended change § E.g., If Bx intended to decrease, should not intervene if decreasing trend in BL. Could intervene if increasing trend in BL 320 Reversal Design Logic q If behavior changes systematically as a function of the introduction & withdrawal of the IV this demonstrates experimental control § Unlikely some extraneous variable produced the change § Becomes more unlikely with each subsequent withdrawal and introduction of the IV 321 ASR q An A-B-A-B design is generally less preferred than the more efficient A-B- A design. a. True b. False 322 ASR q Why is an A-B-A-B design more preferable than an A-B-A design? a. A-B-A-B does not require withdrawing effective treatment b. A-B-A-B shows stronger conclusions regarding IV affecting DV c. A-B-A does not ensure treatment is applied correctly 323 ASR q How does the reversal design demonstrate experimental control? a. By applying the IV simultaneously with an extraneous variable b. By increasing the number of extraneous variables c. When the DV changes when and only when the IV changes 324

Rev 03/22/23 CLG Copyright © ABA Technologies, Inc. 2020 BEH 5043 Unit 6 Slides p. 55 1 2 3 4 5 6 7 8 9 10 Response per Unit time 0 1 2 3 4 5 6 7 8 9 10 11 12 Baseline Treatment Predicted level of the DV if no IV change or if IV has no effect Sampling Units (e.g., Sessions, Days, etc.) 325 1 2 3 4 5 6 7 8 9 10 Response per Unit time 0 1 2 3 4 5 6 7 8 9 10 11 12 Baseline Treatment Difference between prediction and actual effects Sampling Units (e.g., Sessions, Days, etc.) 326 Example of A-B-A Reversal 0 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Responses per Unit time Sampling Units (e.g. Sessions, Days, etc) Baseline Treatment Baseline 327 Time Behavior Baseline 1 Prediction, Verification, and Replication in Reversal Designs 328 Time Behavior Prediction if IV has no effect Baseline 1 Prediction 329 Time Behavior Actual change with IV change Baseline 1 Treatment 1 Verification 330

Rev 03/22/23 CLG Copyright © ABA Technologies, Inc. 2020 BEH 5043 Unit 6 Slides p. 56 Time Behavior Prediction if IV did not produce behavior change Baseline 1 Treatment 1 Baseline 2 Prediction 331 Time Behavior Verifies IV effect & prediction of BL1 Baseline 1 Treatment 1 Baseline 2 Verification 332 Time Behavior Prediction if no IV change Baseline 1 Treatment 1 Baseline 2 Treatment 2 Prediction 333 Time Behavior Replicates initial IV effects Baseline 1 Treatment 1 Baseline 2 Treatment 2 Replication 334 ASR q Prediction of a continuing data path applies at which point of the reversal design? a. Only following initial baseline b. Only following initial treatment phase c. At each phase change 335 ASR q Which design lacks replication? a. A-B design b. Reversal design c. Withdrawal design 336

Rev 03/22/23 CLG Copyright © ABA Technologies, Inc. 2020 BEH 5043 Unit 6 Slides p. 57 Visual Analysis of Reversal Designs q BL phases resemble each other, and treatment phases resemble each other q Change in behavior from one condition to the next q Looking for immediate change 337 ASR q What is desirable when visually inspecting data in a reversal design? Select all that apply: q Similar levels in all baseline phases q Different levels at each baseline phase q Differences from baseline and treatment q Gradual change in each phase q Similar levels at each treatment phase 338 ASR q How does immediacy of change bolster conclusions drawn for reversal designs? a. Ensures the fewest number of data points necessary b. The more quickly responding matches the new contingency, the more likely the change was due to treatment and not something else c. Quick changes show the IV was applied as designed 339 ABAB Reversal Design in the Literature (Vergason & Gravina, 2020, p. 427) 340 ABAB Reversal Design in the Literature (Vergason & Gravina, 2020, p. 427) 341 ABAB Reversal Design in the Literature (Vergason & Gravina, 2020, p. 427) 342

Rev 03/22/23 CLG Copyright © ABA Technologies, Inc. 2020 BEH 5043 Unit 6 Slides p. 58 ABAB Reversal Design in the Literature (Vergason & Gravina, 2020, p. 427) 343 ABAB Reversal Design in the Literature (Vergason & Gravina, 2020, p. 427) 344 ASR q In order to demonstrate experimental control using a reversal design, the data path produced in the return to baseline phase must be identical to the data path produced in the initial baseline phase. a. True b. False 345 Internal Validity in Reversal Designs: Maturation q Minimizing maturation threats: § Condition lengths as long as necessary, as short as possible § Target behaviors that are unlikely to change over time without intervention § Look for immediate changes when IV is withdrawn in second BL • Immediate change is more believable (Ledford & Gast, 2018) 346 Internal Validity in Reversal Designs: Attrition q Minimizing attrition threats: § Explain the procedure, including withdrawal phase during consent process (Ledford & Gast, 2018) 347 Internal Validity in Reversal Designs: Selection Bias q Minimizing selection bias threats: § Randomly select participants (Ledford & Gast, 2018) 348

Rev 03/22/23 CLG Copyright © ABA Technologies, Inc. 2020 BEH 5043 Unit 6 Slides p. 59 Internal Validity in Reversal Designs: Carryover q Minimizing carryover threats: § Collect data in each condition until stability is established § Include stimuli to “signal” conditions (Ledford & Gast, 2018) 349 ASR q Explaining procedures thoroughly so participants are less likely to withdraw at a later point helps prevent which threat to internal validity? a. Maturation b. Attrition c. Carryover effects d. Selection bias 350 ASR q Keeping each phase as short as possible helps address which threat to internal validity? a. Maturation b. Attrition c. Carryover effects d. Selection bias 351 ASR q Which helps minimize carryover effects as threats to internal validity? a. Randomly select participants b. Target behavior unlikely to change without intervention c. Include stimuli to signal conditions 352 Reversal Design Variations q Start with the intervention condition § Phase 1: Intervention (B) § Phase 2: Baseline (A) § Phase 3: Intervention (B) q Reversal with more than one intervention § E.g., ABAC 353 Reversal Starting With Treatment Example 1 (Dixon et al., 2001, p. 362) 354

Rev 03/22/23 CLG Copyright © ABA Technologies, Inc. 2020 BEH 5043 Unit 6 Slides p. 60 Reversal Starting With Treatment Example 2 (Ottenbacher & Hinderer, 2001, p. 791) 355 Note on Terms q A-B-A starting with intervention or B-A-B? q Functionally the same q Same logic applies 356 ASR q When using variations of a reversal design: a. Ensure you are familiar with the separate logic and analysis of each design b. The same logic and analysis applies to each variation 357 ASR q When might it be appropriate to start an experimental design with treatment? a. When behavior requires immediate intervention b. When a treatment is already in place c. Either of these would be appropriate 358 Reversal Design Variations q Start with the intervention condition § Phase 1: Intervention § Phase 2: Baseline § Phase 3: Intervention q Reversal with more than one intervention § E.g., ABAC 359 A-B-C-B-C Reversal Example (Jorgenson et al., 2020, p. 1425) A C B B C 360

Rev 03/22/23 CLG Copyright © ABA Technologies, Inc. 2020 BEH 5043 Unit 6 Slides p. 61 Reversal Design Advantages q Most straightforward single-case arrangement q Current standard for the demonstration of functional relations q Allows replication within the same participant 361 Reversal Design General Limitations q Irreversibility q Ethics q Time and effort 362 Reversal Limitations: Irreversibility q Some behavior change is not reversible: § Intervention effects impossible to withdraw • E.g., new skill is taught § Behavior makes contact with other variables making reversibility unlikely • E.g., natural reinforcement contingencies contacted 363 Reversal Limitations: Ethics q May be unethical to reverse an effective treatment q Counterpoint: Responsibility to show the treatment had an effect and not something else 364 Reversal Limitations: Time and Effort q May require considerable time because stability required in all phases 365 ASR q Reversal designs can be used to assess more than one treatment in comparison to baseline measures. a. True b. False 366

Rev 03/22/23 CLG Copyright © ABA Technologies, Inc. 2020 BEH 5043 Unit 6 Slides p. 62 ASR q Which are advantages of the reversal design? q Strong demonstration of functional relations q Easy to reverse all types of behavior q It is designed to demonstrate an effect quickly q Can be used with same participant 367 ASR q List some disadvantages of the reversal design: 368 Limitations of Reversal Design Variations q B-A-B does not allow assessment of the DV prior to implementing the IV q Reversals with multiple IVs assessed may suffer from sequence effects § E.g., ABAC produce same results as ACAB? 369 Use of Reversal Design q A reversal design is appropriate when: § The target behavior is reversible § Withdrawal of the intervention is not a concern § Stability, order, and/or time is not a concern 370 ASR q In which case is the reversal design most appropriate? a. Behavior appears reversible, withdrawing a treatment could be an issue, stability will likely occur quickly b. Behavior does not appear reversible, no issues withdrawing treatment, stability will likely occur quickly c. Behavior appears reversible, no issues withdrawing treatment, stability will likely occur quickly 371 ASR q If the sequence of phases is likely to impact DV measures, reversal design _____ be appropriate. a. Would b. Would not 372 Disadvantages of reversal designs can include: (a) May be unethical to remove effective treatment; (b) Some types of behavior may be difficult to reverse; (c) Making the reversals requires time and effort.

Rev 03/22/23 CLG Copyright © ABA Technologies, Inc. 2020 BEH 5043 Unit 6 Slides p. 63 ASR q A reversal design would most likely be used for which of the following? a. Learning to recite multiplication facts b. Reducing eye-gouging c. Increasing appropriate requestion for wanted items d. Implementation of a DRO procedure for nose-picking 373 Outline A. Introduction to Experimentation B. Internal and External Validity C.The Experimental Question D.Single-Case Designs E. The A-B Design F. Reversal Designs 374 References q Behavior Analyst Certification Board. (2017). BCBA task list (5th ed.). Author. q Behavior Analyst Certification Board. (2022a). BCaBA test content outline (6th ed.). Author. q Behavior Analyst Certification Board. (2022b). BCBA test content outline (6th ed.). Author. q Barlow, D. H., Nock, M. K., & Hersen, M. (2009). Single case experimental designs: Strategies for studying behavior change (3rd ed.). Pearson. 375 References (continued) q Blasingame, A., Hale, S., & Ludwig, T. D. (2014). The effects of employee-led process design on welder set-up intervals. Journal of Organizational Behavior Management, 34 (3), 207–222. q Cooper, J. O., Heron, T. E., & Heward, W. L. (2020). Applied behavior analysis (3rd ed.). Pearson. q Dixon, M. R., Benedict, H., & Larson, T. (2001). Functional analysis and treatment of inappropriate verbal behavior. Journal of Applied Behavior Analysis , 34 (3), 361–363. 376 References (continued) q Handen, B. L., Parrish, J. M., McClung, T. J., Kerwin, M. E., & Evans, L. D. (1992). Using guided compliance versus time out to promote child compliance: A preliminary comparative analysis in an analogue context. Research in Developmental Disabilities, 13, 157–170. q Hedquist, C. B., & Roscoe, E. M. (2020). A comparison of differential reinforcement procedures for treating automatically reinforced behavior. Journal of Applied Behavior Analysis, 53 (1), 284–295. 377 References (continued) q Houmanfar, R. (2014). The research report as an outlet for scholarly work [Editorial]. Journal of Organizational Behavior Management, 34 (1), 1–6. https://doi.org/10.1080/01608061.2014.873682 q Jorgenson, C. D., Clay, C. J., & Kahng, S. (2020). Evaluating preference for and reinforcing efficacy of a therapy dog to increase verbal statements. Journal of Applied Behavior Analysis, 53 (3), 1419–1431. 378

Rev 03/22/23 CLG Copyright © ABA Technologies, Inc. 2020 BEH 5043 Unit 6 Slides p. 64 References (continued) q Kennedy, C. H. (2005). Single-case designs for educational research. Pearson. q Ledford, J. R., & Gast, D. L. (Eds.). (2018). Single case research methodology: Applications in special education and behavioral sciences (3rd ed.). Routledge. q Leon, Y., Borrero, J. C., & DeLeon, I. G. (2016). Parametric analysis of delayed primary and conditioned reinforcers. Journal of Applied Behavior Analysis , 49 (3), 639–655. 379 References (continued) q Medland, M. B., & Stachnik, T. J. (1972). Good- behavior game: A replication and systematic analysis. Journal of Applied Behavior Analysis, 5 (1), 45–51. q Mehrkam, L. R., Perez, B. C., Self, V. N., Vollmer, T. R., & Dorey, N. R. (2020). Functional analysis and operant treatment of food guarding in a pet dog. Journal of Applied Behavior Analysis . Advance online publication. 380 References (continued) q Miltenberger, R. G., Gatheridge, B.J., Satterlund, M., Egemo-Helm, K. R., Johnson, B. M., Jostad, C., Kelso, P., & Flessner, C. A. (2005). Teaching safety skills to children to prevent gun play: An evaluation of in situ training. Journal of Applied Behavior Analysis, 38( 3), 395–398. q Morris, C., & Peterson, S. M. (2020). A component analysis of an electronic data collection package. Journal of Organizational Behavior Management . Advance online publication. 381 References (continued) q Ottenbacher, K. J., & Hinderer, S. R. (2001) Evidence-based practice: Methods to evaluate individual patient improvement . American Journal of Physical Medicine & Rehabilitation, 80 (10), 786–796. q Saini, V., Gregory, M. K., Uran, K. J., & Fantetti, M. A. (2015). Parametric analysis of response interruption and redirection as treatment for stereotypy. Journal of Applied Behavior Analysis, 48 (1), 96–106. 382 References (continued) q Sidman, M. (1960). Tactics of scientific research: Evaluating experimental data in psychology . Basic Books. q Vergason, C. M., & Gravina, N. E. (2020). Using a guest - and confederate - delivered token economy to increase employee–guest interactions at a zoo. Journal of Applied Behavior Analysis, 53 (1), 422–430. https://doi.org/10.1002/jaba.599 383