BEHAVIORAL TREATMENT OF AUTOMATICALLY REINFORCED SIB:1982 – 2015

GRIFFIN W. ROOKER

JOHNS HOPKINS UNIVERSITY SCHOOL OF MEDICINE AND THE KENNEDY KRIEGER INSTITUTE

ANDREW C. BONNER

UNIVERSITY OF FLORIDA

CHRISTOPHER M. DILLON

THE KENNEDY KRIEGER INSTITUTE AND UNIVERSITY OF MARYLAND, BALTIMORE COUNTY

JENNIFER R. ZARCONE

JOHNS HOPKINS UNIVERSITY SCHOOL OF MEDICINE AND THE KENNEDY KRIEGER INSTITUTE

Some individuals diagnosed with intellectual and developmental disabilities engage in automati-cally reinforced self-injurious behavior (SIB). For these individuals, identifying effective treat-ments may be difficult due to the nature of the reinforcement contingency. The purpose of thisstudy was to review the literature on the treatment of automatically reinforced SIB to determinecommonalities in procedures that produced effective and ineffective treatment outcomes, as wellas historical trends in the treatment of this class of SIB. Results of this review indicated thatthere were many high-quality studies on this topic, but also a wide range in the quality of stud-ies. As for effective treatments, noncontingent reinforcement (the most common treatment com-ponent) was found to be more effective when informed by a competing stimulus assessmentrather than a preference assessment. Suggestions to improve the quality of the published recordand areas in which additional research is needed are discussed.Key words: automatic reinforcement, treatment, self-injury

Many individuals with intellectual and devel-opmental disabilities (IDD) engage in self-injurious behavior (SIB; Baghdadli, Pascal,Grisi, & Aussilloux, 2003; Bodfish, Symons,Parker, & Lewis, 2000; Soke et al., 2016). Thenature of SIB as a severe problem behavior is

unique because SIB produces wounds to one’sown person (Tate & Baroff, 1966). In addi-tion, specific topographies and forms of SIB areone of the defining features of a number ofbehavioral phenotypes, such as Smith-Magenis,Lesch–Nyhan, and Prader-Willi syndromes(Didden, Korzilius, & Curfs, 2007; Elsea &Girirajan, 2008; Hustyi, Hammond,Rezvani, & Hall, 2013, Lee, Berkowitz, &Choi, 2002; Nyhan, 1976; Poisson et al.,2015; Symons, Butler, Sanders, Feurer, &Thompson, 1999; Thompson & Caruso,2002). For example, in a study by Symonset al. (1999), 81% of their sample of individ-uals with Prader-Willi syndrome engaged in aspecific form of SIB (skin picking) and often inthe same body locations.Functional analysis (FA; Iwata, Dorsey, Sli-

fer, Bauman, & Richman, 1982/1994) research

Griffin W. Rooker, Department of Behavioral Psychol-ogy, Johns Hopkins University School of Medicine,Baltimore, MD.

Manuscript preparation was supported by Grant R01HD076653 from the Eunice K. Shriver National Instituteof Child Health and Human Development (NICHD). Thecontents are solely the responsibility of the authors and donot necessarily represent the official views of NICHD.

The authors wish to thank Louis P. Hagopian and ClareJ. Liddon for their comments on earlier versions of thismanuscript.

Address correspondence to: Andrew C. Bonner,Department of Psychology, University of Florida,945 Center Dr., Gainesville, FL, 32611 (email:acbonner@ufl.edu).

doi: 10.1002/jaba.492

JOURNAL OF APPLIED BEHAVIOR ANALYSIS 2018, 51, 974–997 NUMBER 4 (FALL)

© 2018 Society for the Experimental Analysis of Behavior

974

indicates that SIB was maintained by automaticreinforcement (henceforth referred to as auto-matically reinforced SIB) in approximately aquarter of cases. For example, Iwata, Pace, Dorsey,et al. (1994) found that SIB of individuals hospi-talized for treatment was automatically reinforcedin 25.7% of cases. Similarly, multiple reviews ofFAs of SIB in the published record have found asimilar prevalence of automatically reinforced SIB(Beavers, Iwata, and Lerman, 2013; Hanley,Iwata, & McCord, 2003; Kahng, Iwata, & Lewin,2002b). Iwata, Pace, Dorsey, et al. identified twopatterns of behavior observed in an FA that wereindicative of an automatic function. The first ofthese patterns was SIB occurring more frequentlyin the alone condition of the FA relative to otherconditions (i.e., when there are no materials,demands, or people present). The second patternwas SIB occurring across all conditions of the FA,including the control condition. Both of theseresponse patterns indicate that SIB occurs indepen-dent of social consequences, suggesting that there isa direct relation between SIB and reinforcementthat is not mediated by another person(Vaughan &Michael, 1982; Vollmer, 1994).

Differences Between Automatically andSocially Reinforced SIBApplied behavior analytic research suggests

some differences in the clinical presentationand treatment of automatically and sociallyreinforced SIB. When the results of an FA indi-cate a socially mediated function, the nature ofthe reinforcement contingency is also indicated(social positive, social negative reinforcement,or both) by increased rates of behavior as com-pared to the control condition. Although therate of responding in a specific condition(e.g., the attention condition) may be affectedby subtle aspects of how the reinforcer is deliv-ered (e.g., Piazza et al., 1999), the general classof the reinforcer (e.g., social positive reinforce-ment) can be identified. This is not the case

when the result of an FA suggests SIB is auto-matically reinforced.Automatically reinforced SIB has been hypo-

thetically linked to a number of controlling vari-ables, including access to positive reinforcement(e.g., sensory stimulation or production of opi-oids) and negative reinforcement (e.g., painattenuation). However, for a given individual, itis unclear whether their automatically reinforcedbehavior is a function of positive or negativereinforcement, or perhaps both. For this reason,the conclusion of automatic reinforcement ismerely a useful hypothetical construct that setsthe stage for additional analysis to identify thesources of that reinforcement (Kennedy, 1994).Thus with an automatic function, the class ofthe reinforcer remains unknown (Iwata, Dorsey,et al., 1982/1994; Kennedy & Thompson,2000; Thompson, Symons, Delaney, &England, 1995).Research suggests that treatment of automati-

cally reinforced SIB may be more difficult thansocially reinforced SIB because of the uncertaintyin specific variables maintaining automaticallyreinforced SIB. Iwata, Pace, Dorsey, et al. (1994)summarized the FA outcomes for 152 individualswho engaged in SIB and treatment for 121 ofthese individuals. They reported some differencesin the effectiveness of reinforcement-based treat-ments (noncontingent reinforcement and differen-tial reinforcement) for SIB maintained byautomatic reinforcement and social reinforcement.Both differential reinforcement and noncontin-gent reinforcement treatments were successful inmore than 90% of applications for socially rein-forced SIB, whereas these treatments were effectivein approximately 65% of applications for auto-matically reinforced SIB (see Figure 6 in Iwata,Pace, Dorsey, et al., 1994). These apparent differ-ences in treatment outcomes suggest that auto-matically reinforced SIB may be more resistant totreatment than socially reinforced SIB.Hagopian, Rooker, and Zarcone (2015) con-

ducted a consecutive case series of 38 clinicalcases to describe a model for subtyping

975TREATMENT OF AUTOMATICALLY REINFORCED SIB

automatically reinforced SIB based on the pat-tern of responding in the FA. Automaticallyreinforced SIB was categorized as Subtype 1 if itoccurred at a higher level in the no-interactionconditions (i.e., alone or ignore) relative to thecontrol condition. It was categorized as Subtype2 if it occurred at a comparable level across theno- interaction and control condition. A thirdsubtype (Subtype 3) was characterized by thepresence of self-restraint. Hagopian et al. (2015)found that Subtype 2 SIB was highly resistant tothe least restrictive treatment (i.e., reinforcementalone) relative to socially reinforced SIB;whereas, Subtype 1 SIB was responsive to treat-ment at similar levels to SIB maintained by socialreinforcement.Hagopian, Rooker, Zarcone, Bonner, and

Arevalo (2017) conducted a review of pub-lished research on automatically reinforced SIBto further examine SIB subtypes. They foundsimilar differential treatment outcomes basedon Subtypes 1 and 2. The findings of Hago-pian et al. (2015) and Hagopian et al. (2017)in combination with Iwata, Pace, Dorsey,et al. (1994) add support to the notion that atleast some forms of automatically reinforcedSIB (Subtype 2) may be more resistant to treat-ment than socially reinforced SIB.

Differences Between Automatically ReinforcedSIB and Other BehaviorAny behavior maintained by automatic rein-

forcement is reinforced by stimulation pro-duced by the response itself. However, it canbe argued that the stimulation produced bySIB is fundamentally different than that pro-duced by other types of behaviors maintainedby automatic reinforcement because one of thesensory consequences of SIB should be pain.For example, some evidence suggests that indi-viduals with IDD who engage in SIB haveincreased sensitivity to tactile stimulation.Symons, Byiers, Raspa, Bishop, & Bailey(2010) examined the facial expressions of

individuals undergoing a test of sensory evokedpain (light touch, deep pressure, pin prick,warm, and cold stimuli) in 44 adults withIDD. Of these individuals, 29 engaged in SIB,and those individuals were more facially expres-sive during these tests, suggesting that individ-uals with IDD who engage in SIB may bemore sensitive to tactile stimulation. Courte-manche, Schroeder, Sheldon, Sherman, andFowler (2012) evaluated the temporal correla-tion between indices of distress (thought to becorrelates of painful experience) and SIB forfour individuals using a lag sequential analysis.For two individuals, the authors found that dis-tress was highly likely subsequent to SIB sug-gesting that distress related to pain may berelated to SIB. Thus, SIB may be unique fromother behaviors maintained by automatic rein-forcement because both appetitive (the rein-forcer) and aversive (painful stimulation)consequences may be present, although addi-tional research is needed on this topic.

Current Knowledge of Treatment forAutomatically Reinforced SIBCurrent knowledge about behavioral treat-

ment of automatically reinforced SIB is informedby a number of sources. These include:(a) individual studies in the behavior analytic lit-erature (e.g., Tiger, Fisher, & Bouxsein, 2009);(b) review studies describing published datasetson the treatment of SIB, but not specificallyautomatically reinforced SIB (e.g., Kahng,Iwata, & Lewin, 2002a); and (c) review studiesof the treatment of automatically reinforcedproblem behavior (e.g., pica, stereotypies), butnot specifically SIB (e.g., LeBlanc, Patel, &Carr, 2000).Individual studies, which use single-case

experimental designs to determine the effects ofbehavioral treatments on the occurrence of SIBfor a single or handful of individuals, comprisethe majority of the research on the behavioraltreatments of SIB in general, and automatically

GRIFFIN W. ROOKER et al.976

reinforced SIB in particular (Kahng, Iwata, &Lewin, 2002b). For example, Ringdahl, Voll-mer, Marcus, and Roane (1997) examined thepredictive validity of competing stimulus assess-ments (CSA) for the treatment of automaticallyreinforced SIB. Procedurally, a CSA concur-rently measures stimulus interaction or engage-ment and problem behavior, as compared tothe control condition where no stimulus is pre-sent (Piazza, Fisher, Hanley, Hilker, & Derby,1996). Ringdahl et al. used CSAs to identifyitems to be included in a differential reinforce-ment of alternative behavior (DRA), a differen-tial reinforcement of other behavior (DRO), ora noncontingent reinforcement (NCR) treat-ment (i.e., environmental enrichment). Theauthors found that the CSA correctly predictedthe effectiveness of NCR to treat SIB for allthree participants.Individual studies such as that of Ringdahl

et al. (1997) provide strong support for effec-tive treatments of automatically reinforced SIB;however, individual studies are not withouttheir weaknesses. These studies are commonlyconducted with only a few individuals and,when detailing clinical care, the application ofassessment and treatment procedures is (appro-priately) driven by the needs of the patient.Thus, literature reviews often summarize theresearch in a particular area to highlight com-monalities across these studies.Two relevant literature reviews were con-

ducted by Kahng et al. (2002a) and LeBlancet al. (2000). Kahng et al. reviewed 35 years ofpublished research on the behavioral treatmentof SIB that was maintained by social and/orautomatic reinforcement. The authors detailedthe effectiveness of several behavioral treat-ments and reported on the relative effectivenessof these treatments over time. Specifically, theauthors found that SIB was reduced by a meanof 83.7% across all published behavioral treat-ments. Whereas Kahng et al. focused on thetreatment of SIB regardless of function,Leblanc et al. focused specifically on the

effectiveness of treatments for problem behav-iors maintained by automatic reinforcement(e.g., pica, SIB, stereotypies) but did not focusexclusively on SIB (i.e., less than 50% of theindividuals’ data reviewed in that study engagedin SIB). In their analysis, the authors notedthat CSAs have advanced our understanding oftreatment for automatically reinforced behavior.In both of these reviews, the authors did not

focus solely on automatically reinforced SIB.Kahng et al. (2002a) may not have been able toconduct this analysis because the function ofSIB was likely unknown for more than half theirsample (particularly for articles reviewed before1982). This type of analysis would have beenoutside the scope of Leblanc et al. (2000), but asa result the study did not examine how theunique properties of SIB may have differentiallyaffected treatment outcomes for this behavior.Thus, although these articles and other reviewpapers on SIB (e.g., Favell, McGimsey, &Schell, 1982; Matson & LuVullo, 2008) andbehaviors maintained by automatic reinforce-ment (e.g., Lanovaz, & Sladeczek, 2012) arevery informative in demonstrating the effective-ness of behavioral treatments, they are less infor-mative about the treatment of automaticallyreinforced SIB in particular. In addition, morethan a decade has passed since these reviewswere conducted, and additional summary ofmore recent studies has not been conducted,particularly in light of additional evidence sug-gesting the differential effectiveness of treatmentfor socially maintained SIB versus automaticSIB (Hagopian et al., 2017).Hagopian et al. (2017) provides a review of

the literature on the treatment of automaticallymaintained SIB. However, because the focus ofthe Hagopian et al. study was to subtype FAsand determine the effectiveness of treatmentacross subtypes, the authors excluded fromtheir review a large portion of the published lit-erature on the treatment of automatically rein-forced SIB. This resulted in limitedinformation on the breadth of research on the

977TREATMENT OF AUTOMATICALLY REINFORCED SIB

treatment of automatically reinforced SIB. Forexample, the data of 27 individuals, the major-ity of whose data was published in this century,were excluded from their study (see Hagopianet al., supplemental table). Further, the numberof excluded datasets suggested that there mightbe issues with the quality of research publishedon the topic. Thus, the purpose of this studywas to critically review studies published overthe last 34 years on the treatment of automati-cally reinforced SIB. Analyses were performedto identify trends and commonalities in proce-dures that produced effective and ineffectivetreatment outcomes. In addition, we evaluatedthe methodological quality of the studies, sum-marized current knowledge, and identifieddirections for future research.

METHOD

Article Search, Inclusion/Exclusion, andCoding of ArticlesStudies were identified for inclusion by con-

ducting a comprehensive search of GoogleScholar, Pubmed, and Web of Science forarticles published between 1982-2015 thatcontained the following terms: “self-injur*AND automatic reinforcement”, “SIB ANDautomatic reinforcement”, “self-injur* ANDsensory”, “SIB AND sensory”, “self-injur*AND nonsocial”, “SIB AND nonsocial”, “self-injur* AND undifferentiated”, “SIB ANDundifferentiated”, “self-restrain* AND self-injur*”, and “self-restrain* AND SIB.” Weused 1982 as a search parameter because thepurpose of the study was related to a particularfunction of SIB and a conclusive demonstrationof the function of SIB was unlikely to be pub-lished prior to the Iwata, Dorsey,et al. (1994/1982) publication.In order to try to capture all of the possible

treatments attempted for automatically rein-forced SIB, we applied inclusion criteria thatwere as broad as possible, while still ensuringthe behavior being treated was thought to be

automatically reinforced SIB. Thus, all articlesthat met the following criteria were included:(a) the article contained original research(i.e., not a review of previously published data),(b) the targeted behavior met the definition ofSIB (i.e., was injurious) and was measuredindependently from other behaviors (i.e., thetargeted problem behavior did not contain SIBand another non-self-injurious behavior),(c) SIB was reported or suggested by theauthors to be maintained by automatic rein-forcement (or an analogous term, such as sen-sory stimulation or self-stimulatory), (d) abehavioral treatment of SIB was reported, and(e) individualized treatment data were depictedin session-by-session format. In addition, allstudies with individuals with SIB maintainedby automatic reinforcement that Hagopianet al. (2017) had identified were included, aslong as treatment data were available. Data setswere excluded if the authors concluded thatSIB was maintained by social reinforcement, abehavioral treatment of SIB was not conducted,the specific type of behavioral treatment(e.g., NCR) was not indicated, or individualdata on the behavioral treatment were not pre-sented. We defined a behavioral treatment asany treatment that manipulated the externalenvironment in an effort to reduce automati-cally reinforced SIB. Studies targeting hand-mouthing were excluded, as published researchindicates this behavior may sometimes be inju-rious but is sometimes considered noninjurious(i.e., a stereotypy). For example, Roscoe, Iwata,and Zhou (2013) found that at least 20% ofindividuals in their study had no evidence ofinjury related to hand-mouthing and did notwear equipment or restraint that limited thedamage produced by this behavior. In addition,a number of published reports directly state thisbehavior is stereotypic or proto-SIB rather thanSIB (e.g., Rapp & Vollmer, 2005; Richman &Lindauer, 2005). Thus, to avoid includingbehaviors in this review that may be noninjur-ious, we chose to be conservative and exclude

GRIFFIN W. ROOKER et al.978

any studies on hand mouthing. Among thosearticles that met our inclusion criteria, a dem-onstration of experimental control in the FA ortreatment was not required; however, we scoredthese articles on the degree to which experi-mental control was demonstrated (see below).After identifying articles for inclusion, a

15-item Research Article Rating Scale (RARS)was developed to judge the quality of each arti-cle based on distinct relevant features (see Sup-porting Material for a copy of the RARS). TheRARS was composed of five subscales: (a) thedescription of the participant, setting, anddependent variables; (b) measurement of thedependent variable, and interobserver agree-ment; (c) the presence of an FA, the length ofthis FA, and the inclusion of test and controlconditions in the FA; (d) the degree of baselinestability (based on visual analysis), complete-ness/quality of experimental design, and degreeof level change between phases; and(e) description of the baseline and treatmentconditions, as well as experimenter control ofthe independent variable (during introductionand withdrawal). Each of the subscales con-tained three questions and each question hadequal weight toward the final RARS score. Foreach question scorers assigned a value of fullcredit (1 point) for complete information,partial-credit (0.5 points) for limited or incom-plete information, or no credit (zero points) forno information being provided. By totaling thenumber of points earned and dividing by thetotal number of possible points (15 points),each article received a score, and each subscalewas worth a maximum of 20% of the total pos-sible score. To ensure scoring of articles wasaccurate, the second and third authors weretrained to score the articles using five quasiran-domly selected articles. The raters then codedeach article and interrater agreement was calcu-lated using exact agreement for each item onthe datasheet, and the level of agreement wasdetermined for each article. Interrater agree-ment was assessed for 30.6% of articles and the

mean exact agreement score across these articleswas 95% (range 87.5% - 100%).Descriptive data. For each participant in a

study, data were extracted and retained on theindividual’s demographic information, responsetopography, the treatment setting, and treatmenttype. For demographic information, we retainedthe individual’s gender, age, level of intellectualdisability, autism spectrum disorder diagnosis,and any identified genetic or medical conditions.For SIB topography, similar to Hagopianet al. (2015), we classified each specific formwith respect to the area of the body to which SIBtargeted. That is, SIB targeting the body(e.g., leg slapping) was recorded as body-directed, SIB targeting the head (e.g., head bang-ing) was recorded as head-directed, SIB targetingthe skin (e.g., scratching) was recorded as skin-directed, and SIB targeting the mouth or lips(e.g., tongue-biting) was recorded as mouth-directed. In one case (Barry from Ringdahl et al.,1997), two different response topographies wereassessed and scored separately. For treatment set-tings, we identified if interventions were con-ducted in an outpatient day program, inpatienthospital, school, vocational setting, residentialsetting, or community setting; or as unknown, ifthe location could not be identified. For treat-ment types, we categorized them with respect tothe components embedded within the interven-tion as follows: type of differential and/or non-contingent reinforcement (e.g., DRA, DRO,NCR), antecedent manipulations (e.g., the useof wrist weights as a means to increase theresponse effort associated with SIB), protectiveequipment (e.g., a helmet), blocking (e.g., a ther-apist using any part of their own body to physi-cally block or interfere with the occurrence ofSIB), punishment (e.g., contingent hands downfor 30 s), mechanical restraint (e.g., devices usedto limit the movement of extremities or preventthe occurrence of SIB, such as arm splints), andfinally, multiple components if reinforcementand two of the components listed previously were

979TREATMENT OF AUTOMATICALLY REINFORCED SIB

present. Throughout, the term “additional com-ponent” refers only to the nonreinforcement-based procedures described above. Individualtreatments were distinguished from each other bychanges in fundamental components of treat-ment (e.g., changing from noncontingent to con-tingent reinforcement) or the addition of newtreatment components (e.g., adding blocking toreinforcement). For example, if a single partici-pant experienced a reinforcement treatment in anABAB design, this would count as one imple-mentation of reinforcement treatment; whereas ifboth reinforcement and reinforcement with anadditional component (e.g., punishment) wereexperienced (e.g., ABACAC), two treatmentimplementations would have been included inthe analysis.For all treatments utilizing reinforcement,

the specific type of reinforcement (e.g., NCR,DRO, DRA) was identified. For NCR (becausethis was the most common treatment), we alsoidentified how items were identified for inclu-sion in treatment. Specifically, we identified if apreference assessment (Fisher et al., 1992; Pace,Ivancic, Edwards, Iwata, & Page 1985) or aCSA (Piazza et al., 1996) was used to identifythe item(s) included during NCR. The primarydifference between these procedures was thecollection of data on the occurrence of SIB dur-ing periods in which a stimulus was present.Similarly, all additional procedures werereviewed in detail to determine the types ofadditional components used.Treatment effectiveness data. For all data sets

included in the analysis, data point values wereextracted using GetData Graph Digitizer (ver-sion 2.26; www.getdata-graph-digitizer.com).This computer software allows digital conver-sion of graphed data points into actual values.This program has been used in prior behavior-analytic (Bowman-Perrot, Burke, Zaini,Zhang, & Vannest, 2015; Hagopian et al.,2017; Levy et al., 2017) and medical research(Ma et al., 2015; Prieto, Vazquez, & Murado,2014; Zhou et al., 2015).

To extract published data point values, theportable document format of the publishedgraph was converted to a picture file anduploaded to the GetData program. The axesparameters were provided to the program basedon the graph and the value of individual datapoints was determined by selecting them usinga point and click method with the mouse.Once these data were obtained, we examinedthe procedures of each individual study to iden-tify both the session duration and measurementtype. For frequency and percentage of intervaldata collection, we rounded the GetDataobtained value to the closet possible value giventhe session duration and method of data collec-tion. For example, if the session duration was10 min and data were presented as responsesper minute (RPM), an obtained GetData valuewas rounded to the nearest 10th decimal place(e.g., 2.31 RPM was rounded to 2.3 RPM). Ifthe authors did not specify the length of thesession or the interval size, the unaltered Get-Data value was used.After we extracted the data values for each

graph using GetData, we calculated the meanoccurrence (rate, percentage of intervals, or per-centage of session) of SIB in baseline and treat-ment conditions using the last five data pointsin each of these conditions. For cases in whicha given condition contained fewer than fivedata points, all data points in that phase wereused. We then calculated the percent reductionin SIB during the final implementation of agiven treatment condition compared to the ini-tial baseline condition to determine the effec-tiveness of the intervention (mean baselineoccurrence of SIB minus mean treatmentoccurrence of SIB divided by mean baselineoccurrence of SIB multiplied by 100). Effectivetreatments were defined as those that producedat least an 80% reduction of target responding.

RESULTS

Results of the review identified 49 publishedresearch articles that met the inclusions criteria,

GRIFFIN W. ROOKER et al.980

detailing the treatment of 69 individuals (1.4individuals on average per study). Table 1shows the demographic information for theseindividuals. Slightly more than half of the indi-viduals were younger than 18 years of age. Adiagnosis of autism was reported for approxi-mately one third of individuals. The majorityof individuals had either profound or severeintellectual disability (62.3%). Additional

genetic or medical conditions were reported for39.1% of individuals, with the presence of somegenetic or natal syndrome (n = 14) and sensorydeficits (n = 8) being the most common condi-tions. Across the 69 individuals, 90 categories ofSIB were reported (1.3 categories on average perindividual), with individuals most commonlyengaging in head-directed (68.1%) and/or skin-directed SIB (44.9%).

Quality Analysis of the Published RecordThe quality of each article was assessed to

demonstrate the relative strength (or lackthereof ) of literature in this area, and to iden-tify aspects of the published record that couldbe improved. The quality of each article wasdetermined using the RARS (see SupportingInformation), and the mean overall article scorefor all papers reviewed was 86.2% (range 53%to 100%). Table 2 shows the results across allRARS subscales.In general, scores were lowest for the descrip-

tion of and presence of experimental control inthe FA. Figure 1 provides additional detail onthese data. The top panel shows the percentageof points obtained in each subscale and totalpercentage of points for each article. These datasuggest that articles lost the most percentagepoints for the FA and for experimental controlduring treatment. Additionally, the majority ofthe articles scored were of a high quality (scored>80%); however, 11 articles (22.4% of thesample) scored less than 80%. In the bottompanel, the percentage of available points scoredfor each RARS question across all articles is

Table 1Participant Characteristics (n = 69)

Participant VariableNumber ofParticipants

Percent ofParticipants

AgeChildren (3 to 12 years) 30 43.5Adolescents (13 to 18 years) 10 14.5Adults (> 18 years) 29 42.0

AutismYes 21 30.4No/not reported 48 69.6

Level of IntellectualDisability (n = 69)None 2 2.9Mild/Borderline 1 1.4Moderate 7 10.1Severe 11 15.9Profound 32 46.4Unspecified 8 11.6

Intellectual Disability not reported 8 11.6Genetic/Medical Conditions (n = 27)Sensory deficit 8 11.6Cerebral palsy 5 7.2Down syndrome 3 4.3Other genetic or natal syndromes 6 8.7Disorders 5 7.2Hydro/microcephalus 3 4.3

Response CategoryHead-directed 47 68.1Body-directed 9 13.0Skin-directed 31 44.9Mouth -directed 3 4.3

Table 2Summary of All RARS Scores

Number of articleswith a perfect score

Percent of articleswith perfect scores Mean Score Range

Participants and Setting 28 57.1% 17.1 13.3 - 20Dependent Variables and Agreement 36 73.5% 18.2 13.3 - 20Functional Analysis 24 49.0% 14.3 0 - 20Experimental Control 32 65.3% 17.5 10 - 20Treatment Description 43 87.8% 19.2 6.7 - 20

981TREATMENT OF AUTOMATICALLY REINFORCED SIB

shown. For example, all 49 articles providedcomplete information on first RARS question.These data indicated that in the minority ofstudies, which were not scored as high quality,the missing information in these articles tended

to be related to the similar aspects. Thisincluded an incomplete description of the set-ting, IOA reported at less than the conven-tional standard or not at all, not conductingand presenting FA data, and a failure to

Figure 1. Results of the Research Article Rating Scale (RARS) analysis.. The top panel shows the percentage ofpoints obtained in each subscale and total percentage of pointsfor each article. The bottom panel shows the percentageof available points obtained across all articles for each question on the RARS. See the RARS for the specific questions ineach subscale.

GRIFFIN W. ROOKER et al.982

produce a change in level (which is required todemonstrate experimental control).Because experimental control was directly

related to the purpose of the study (i.e., examiningthe effectiveness of treatment), additional analyseswere conducted on articles relating to three aspectsof the experimental control subscale. Specifically,we compared results of studies: (a) determined tohave scored perfectly (20%) or not (< 20%) in thissubsection; (b) determined to have a complete(e.g., ABAB), partial (e.g., ABA or BAB), orincomplete (e.g., AB) experimental design; and(c) determined to have produced a strong(i.e., ~ 80%) reduction based on visual analysis,rates were similar across baselines, and less than25% of data points in baseline and treatment wereoverlapping), moderate (i.e., ~50% - 80% reduc-tion based on visual analysis, or baseline rates werenot recaptured in second baseline, or 25% - 50%of data points were overlapping) and weak(i.e., < ~50% reduction based on visual analysis)level change.When comparing studies that scored per-

fectly (32) on the experimental control sub-section to those that didn’t (17), we noted nodifference in the distributions of treatmenttypes employed. In addition, although themean percent reduction (82.7%) across treat-ments was greater for those articles having aperfect score in comparison to those scoringless than 20% (55.2%), approximately 60% oftreatments in articles scoring less than 20%were effective (producing an 80% reduction inSIB from baseline) and there was no significantdifference in the percent reduction as deter-mined by a Mann–Whitney test of statisticalsignificance (p = .44; U = 1730). Thus, thereappeared to be no relation between overallscore on the experimental control subscale foran article and treatment effectiveness.When comparing studies that used a com-

plete (40), partial (3), or incomplete (6) design,similar to the prior results, there was no differ-ence in the distribution of treatment types inthese articles. The mean percent reduction

across treatments was greatest in those studiesscored as having a partial design (80.3%), fol-lowed by those scored as having an incompletedesign (78.1%), and by those scored as havinga complete design (66.3%) although all wererelatively similar. Across these design groups, asimilar proportion of treatments produced an80% reduction in SIB from baseline (range66.3% - 77.8%). There was no significant dif-ference between groups when compared using aKruskal-Wallis test of statistical significance(p = .06, H = 5.6). Thus, there appeared to beno relation between overall design and treat-ment effectiveness.When comparing articles scored as having a

strong (37), moderate (8), or weak (4) levelchange, we noted no difference in the distribu-tions of treatment types employed in thesestudies. Mean percent reduction was greatestfor those treatments from articles scored as hav-ing a strong level change (82.8%), followed byarticles scored as having moderate level change(78.4%) and finally, articles scored as having aweak level change (-9.5%). Additionally, a dif-ference was found across groups comparingtreatment percent reduction using a Kruskal-Wallis test of statistical significance(p < 0.0001. H = 21.12). Specifically, differ-ences were found between the percent reduc-tion of treatments in the strong and weak levelchange groups and the strong and moderatelevel change groups. This finding was expected,as level change is directly related to treatmenteffectiveness. That is, a weak change in level isthe definition of ineffective treatment.In addition to these analyses, comparisons

were conducted based on treatment types(described below) relating to the experimentalcontrol subscale when sufficient numbers forcomparison were available; however, none ofthese comparisons produced differences sug-gesting that treatment outcomes differed basedon score or experimental design. Based on thisinformation, all of the treatment data we iden-tified were retained for continued analysis. This

983TREATMENT OF AUTOMATICALLY REINFORCED SIB

was because: (a) a high proportion of treat-ments were effective despite coming from stud-ies that scored less than 20% on theexperimental control subscale or had a partialor incomplete design and (b) when a differencewas obtained based on a quality measure ofexperimental design, it was related to the lackof a treatment effect and identifying whentreatments were ineffective was part of the pur-pose of the study.Types and outcomes of treatments of automati-

cally reinforced SIB. For the 69 participants,71 designs were used to evaluate 123 attemptedtreatments (1.8 treatments per individual onaverage). For one individual, a differentexperimental design was used to evaluate twodifferent behaviors and for another individualtwo independent designs were used to assesstreatment for one behavior. Across studies, acomplete (e.g., ABAB) or partial reversal(e.g., ABA) was used to evaluate treatment in55% of cases, a multielement or multiple base-line was used in 24% of cases, a combineddesign (e.g., reversal plus multielement) wasused in 12% of cases, and a single baseline andexposure (i.e., AB) was used in 8% of thesecases. In addition, for 62 of the 69 participants(90%), authors provided information on thesetting in which treatments were assessed.Assessments and treatments were conducted inoutpatient clinics for 20 individuals, inpatient

programs for 19 individuals, schools for 14 indi-viduals, residential care facilities for 7 individ-uals, in the community for 1 individual, and ina vocational setting for 1 individual.Table 3 shows the number of times each

type of treatment was attempted. The first col-umn provides information on the number ofreinforcement procedures used alone. The fol-lowing columns show the number of timeseach additional procedure was used. The firstrow provides information on the number oftimes procedures were used without reinforce-ment. The following rows provide informationon the type and number of reinforcement pro-cedures used. Of the treatments evaluated,57 (46.3%) used reinforcement alone(i.e., with no other component), 52 (42.3%)used reinforcement plus some other compo-nent, and 14 (11.4%) did not include a rein-forcement component. Of the reinforcementtreatment components evaluated, NCR wasevaluated 75 times (either alone or in combina-tion with an additional procedure), differentialreinforcement procedures were evaluated27 times (alone or in combination with anadditional procedure), and combined reinforce-ment treatments (e.g., DRA with DRO) wereevaluated 7 times. In addition, punishment wasthe most commonly added procedure, and infact, was used without reinforcement on sixoccasions. Given the difficulty in treating

Table 3Attempted Treatments for Automatically Reinforced SIB

Reinforcement None AntecedentProtectiveEquipment Blocking Punishment Restraint

MultipleComponents

None n/a 4 3 1 6 0 0 14DRA 3 2 0 3 0 1 0 9DRO 10 0 0 1 7 0 0 18DRA/DRO 3 0 0 0 0 0 0 3NCR 37 7 3 6 17 1 4 75DRA/NCR 3 0 0 0 0 0 0 3DRA/DRO/NCR 1 0 0 0 0 0 0 1

Total 57 13 6 11 30 2 4 123

Note. The columns depict the components of treatment and the rows depict the type of reinforcement used. DRA = dif-ferential reinforcement of alternative behavior; DRO = differential reinforcement of other behavior; NCR = noncontin-gent reinforcement; Antecedent = antecedent other than NCR.

GRIFFIN W. ROOKER et al.984

automatically reinforced SIB, it is not surpris-ing that additional procedures had to be com-bined with reinforcement. However, it wassomewhat surprising that in 14 applications atreatment with no reinforcement component atall was attempted.Table 4 shows how often these treatments

were successful (producing an 80% reductionin SIB from baseline) as well as the average per-cent reduction (in parentheses). Similar toTable 3, Table 4 shows reinforcement-basedtreatments with no additional components(reinforcement alone) in the first column, rein-forcement with additional treatment compo-nents in subsequent columns, and treatmentswithout reinforcement across the first row, andthe type of reinforcement procedure in subse-quent rows. For the 69 individuals, 85.5% hadat least one successful treatment (defined asreducing target behavior by at least 80% com-pared to a baseline). Of the 123 treatments,85 effective treatments were identified (69.1%)and the mean percent reduction was 69.3%.Considering each reinforcement procedure,including with and without additional compo-nents, NCR treatments were the most com-monly used and effective (74.7% applicationswere successful and a mean percent reductionof 80.2% was found). DRA and DRO bythemselves were not found to be effective, but

in combination an 89% reduction wasachieved. Interestingly, DRA and DRO werecombined with fewer additional procedures,but when they were, these treatments weremostly successful.Across the 57 reinforcement-alone treat-

ments, 36 effective treatments were identified(63.2%), and the mean percent reduction was52.5%. For each type of reinforcement treat-ment: NCR was effective in 28 applications,and the mean percent reduction was 74.2%;DRO was effective in 3 applications, and themean percent reduction was -35% (that is, tar-get behavior actually increased on average, as aresult of the other seven failed attempts); DRAcombined with DRO was effective in 2 applica-tions, and the mean percent reduction was89%; DRA combined with NCR was effectivein 2 applications, and the mean percent reduc-tion was 81.4%; DRA combined with DROand NCR was effective in 1 application, andthe mean percent reduction was 100%; andDRA was never effective (0%), the mean per-cent reduction was -4.4%.Across the 52 treatments that used reinforce-

ment with an additional component, 41 effec-tive treatments were identified (78.9%), andthe mean percent reduction was 87.8%. Whencomparing all of the additional componentsthat were combined with reinforcement,

Table 4Percentage of Effective Treatments and Mean Percentage Reduction for All Treatments

Reinforcement None AntecedentProtectiveEquipment Blocking Punishment Restraint

MultipleComponents

None n/a 75 (82.5) 33.3 (75.2) 0 (73) 66.7 (68.9) -- -- 57.1 (74.9)DRA 0 (-4.4) 100 (89.4) -- 66.7 (82.2) -- 100 (100) -- 55.6 (56.9)DRO 30 (-35) -- -- 100 (96.4) 100 (95.8) -- -- 61.1 (23.2)DRA/DRO 66.7 (89) -- -- -- -- -- -- 66.7 (89)NCR 75.7(74.2) 71.4 (88.7) 100 (97.8) 66.7 (76.3) 70.6 (85.9) 100 (97.3) 75 (86.4) 74.7 (80.2)DRA/NCR 66.7 (81.4) -- -- -- -- -- -- 66.7 (81.4)DRA/DRO/NCR 100 (100) -- -- -- -- -- -- 100 (100)Total 63.2 (52.5) 76.9 (86.9) 66.7 (86.5) 63.6 (79.4) 63.6 (84.8) 100 (98.6) 75 (86.4) 69.1 (69.3)

Note. The columns depict the components of treatment and the rows depict the type of reinforcement used. DRA = dif-ferential reinforcement of alternative behavior; DRO = differential reinforcement of other behavior; NCR = noncontin-gent reinforcement.

985TREATMENT OF AUTOMATICALLY REINFORCED SIB

punishment was the most frequently used addi-tional component, with 19 of 24 applications(79.2%) effective, and a mean percent reduc-tion of 88.8%. Additional components thatappeared to be effective were the use ofrestraints (98.6% mean percent reduction),protective equipment (97.8% mean percentreduction); antecedents (88.9% mean percentreduction), blocking (80.1% mean percentreduction), and multiple components (86.4%mean percent reduction).For the reinforcement components used with

these additional procedures: the use of NCRwas effective in 28 of 38 applications (73.7%),with a mean percent reduction of 86.2%;DRO was effective in 8 of 8 applications(100%), with a mean percent reduction of95.8%, and DRA was effective in 5 of 6 appli-cations (83.3%), with a mean percent reduc-tion of 87.6%. Of the 14 treatments withoutreinforcement: punishment was effective infour applications (66.7%) with a mean percentreduction of 68.9%; the use of an antecedentcomponent other than NCR was effective inthree applications (75%), with a mean percentreduction of 82.5%; protective equipment waseffective in a single application (33.3%), with amean percent reduction of 75.2%; and block-ing did not quite meet the criteria for effective-ness during the single application it wasattempted; there was a 73% percent reduction.Regarding treatment effectiveness, results of

the current analysis indicate that a successfultreatment of automatically reinforced SIB wasdeveloped for most individuals (59 out of 69).Of the 123 treatments of automatically rein-forced SIB evaluated, 69.1% of treatmentsattempted reduced SIB by 80% from baseline(a very similar finding to Iwata, Pace, Dorsey,et al., 1994), and the mean percent reductionacross all treatments was 69.3%; however,diverse outcomes were observed across specifictreatments. For example, given an equal num-ber of applications, DRA alone never producedan 80% reduction in SIB, with a mean percent

reduction of -4.4%; whereas, NCR with pro-tective equipment produced an 80% reductionin SIB in 100% of applications with a meanpercent reduction of 97.8%. These resultsshould be tempered by the potential for publi-cation bias, which has been noted with somebehavior analytic procedures (Sham & Smith,2014). However, the behavior analytic researchalso suggests that some behavioral treatments(e.g., NCR) may have similar effectiveness inthe published record and when used clinically(Phillips, Iannaccone, Rooker, & Hago-pian, 2017).Historical trends in the treatment of automati-

cally reinforced SIB. Figure 2 shows the numberof procedures attempted and successful whenreinforcement alone was implemented (toppanel), reinforcement plus other components(middle panel), and treatments without rein-forcement (bottom panel) presented cumula-tively across the 34-year period analyzed in thisstudy. A dramatic increase in the number ofreinforcement and reinforcement plus othercomponents was observed starting in 1996 andcontinuing to 2009, after which, reinforcementalone treatments were much more likely to bepublished than reinforcement plus other com-ponent treatments. Comparing the past decade(2006 – 2015) and the decade before (1996 –2005), both reinforcement alone and reinforce-ment with other components have becomemore effective over time. For reinforcementalone, treatment was effective in 15 of 25 appli-cations (60%) during the 1996 – 2005 decade;whereas, this treatment was effective in 15 of21 applications (71.4%) during the nextdecade. For reinforcement with additional com-ponents, treatment was effective in 24 of31 applications (77.4%) during the 1996 –2005 period; whereas, these treatments wereeffective in 5 of 6 applications (83.3%) during2006 – 2015 period. Similar analysis was notconducted for treatments without reinforce-ment, because so few applications were con-ducted during the same period.

GRIFFIN W. ROOKER et al.986

Figure 3 shows the data from table 2 acrossthe 34-year period in 5-year bins. Between1996 and 2005, a large increase in the numberof treatments using punishment and antecedent(other than NCR) components was observed.The most common form of punishment wasresponse cost (33.3%), followed closely by ahands-down procedure (21.2%) and overcor-rection (12.1%). No other punishment proce-dure was conducted more than twice. Since2005, there has been a slight increase in theuse of less restrictive components (antecedents,blocking, and equipment) and a large decreasein the use of more restrictive procedures

(punishment, restraint, and multiple compo-nents). Comparing data from the past decade(2006 – 2015) and the decade before (1996 –2005) indicates a substantial decrease in the useof punishment. In the past decade (2006 –2015), of the 15 articles reviewed only oneapplication of punishment was attempted (pro-portion of studies = 0.07); in the decadebefore, of the 23 articles reviewed 16 applica-tions of punishment were attempted (propor-tion of studies = 0.7). This translates to a 90%reduction in the use of punishment across thetwo decades.Because NCR was commonly evaluated

(making up more than half of the sample), addi-tional analyses regarding this treatment wereconducted. As previously mentioned, each NCRarticle was coded on the procedure used(e.g., CSA) to determine the item included inNCR. Across all 75 NCR treatments, we couldnot determine how an item came to be includedin NCR in 10 implementations. For the remain-ing 65 implementations, the item was selectedbased on CSAs in 33 implementations (44% ofNCR treatments), based on a preference assess-ment in 25 implementations (33.3% of NCRtreatments), and based on anecdotal observationor report in 7 implementations (9.3% of NCRtreatments).Figure 4 shows the historical use of preference

and competition procedures to identify the item

Figure 3. Cumulative use of nonreinforcement com-ponents since 1982 based on the studies used for thisreview.

Figure 2. Cumulative use and effectiveness of rein-forcement (top panel), reinforcement plus additional com-ponents (middle panel), and no reinforcement (bottompanel) treatments since 1982.

987TREATMENT OF AUTOMATICALLY REINFORCED SIB

included in NCR-alone (top panel) and in NCRplus additional components (bottom panel) over5-year periods since 1990. Neither preferencenor competition procedures were identified asbeing used for NCR prior to 1990. For NCR-alone treatments, trends indicate that preferenceassessments and CSAs were used approximatelythe same amount for the first decade (1990 –2000). However, since 2000, CSAs were muchmore likely when used for NCR-alone (toppanel). The use of CSAs for NCR with an addi-tional component also increased, but onlyslightly between 2010 and 2015 (bottom panel).Figure 5 shows the number and effectiveness

of NCR-alone and NCR plus other compo-nents treatments based on the type of procedure used to select the item included in

NCR. NCR alone was attempted 10 timesbased on a preference assessment, and waseffective 6 times (60%); whereas, this treatmentwas attempted 24 times based on a CSA proce-dure, and was effective 19 times (79.2%).NCR with an additional component wasattempted 15 times based on a preferenceassessment, and was effective 11 times(73.3%); whereas, this treatment was attempted9 times based on a CSA procedure and waseffective 7 times (77.8%). Across all NCRtreatments, when a preassessment was used toidentify the item(s) included in NCR, thistreatment was attempted 25 times based onpreference, and effective 17 times (68%);whereas, this treatment was attempted 33 timesbased on competition and found to be effective26 times (78.8%).

DISCUSSION

Hagopian et al. (2017) reviewed the treat-ment of automatically reinforced SIB for20 individuals with 38 treatments, with thefocus on determining how subtypes of auto-matically reinforced SIB responded to similartreatments. Expanding the inclusion and exclu-sion criteria from the Hagopian et al. study, weidentified an additional 49 individuals who

Figure 5. Frequency of attempted and successfulNCR treatments using preference and competitionprocedures.

Figure 4. Cumulative competition and preferenceprocedures used to identify items included in NCR-alone(top panel) and NCR with additonal components (bottompanel) treatments.

GRIFFIN W. ROOKER et al.988

engaged in automatically reinforced SIB and anadditional 85 treatments for this behavior.Results of our analysis indicated that the

majority of articles published on automaticallyreinforced SIB were of a high quality. Severalstudies identified as “low quality” (e.g., havinga poor design) did in fact report highly success-ful treatment outcomes (e.g., Stokes & Luiselli,2008). Thus, we felt this justified the inclusionof these studies. Using our scoring procedure,approximately 73% of articles scored above an80% regarding the quality of their description,as well as their design. However, when analyz-ing these scores, the lowest proportion wasawarded for the FA description and data, aswell as the treatment data subscales.Eighty-eight percent of articles reported con-

ducting an FA. Of those articles that conductedan FA, the description of the FA was detailedenough to determine an appropriate test(e.g., alone) and control (e.g., play) condition.However, far fewer details were providedregarding the length of FAs (i.e., the number ofseries of the FA conducted to determine anautomatic function). Indeed, across all 49 stud-ies, the majority of studies did not provide evi-dence that an FA of sufficient length (threeseries) was used to determine behavior main-tained by automatic reinforcement; six studiesdid not conduct an FA, three studies providedevidence that less than three series had beenconducted (based on an FA graph), and22 studies provided no detail (did not have anFA graph). Given that it may have taken severalyears for FA methodology to become dissemi-nated to the research and clinical communities,it is not surprising that some researchers maynot have conducted an FA. However, the prac-tice in publication of simply reporting that anFA was conducted and that behavior was deter-mined to be automatic (rather than presentingdescription and data depiction of the FA to thereader) has reduced the quality of literaturewhen comparing FA results to treatmentresults, particularly for automatically reinforced

SIB (Hagopian et al., 2017). These results sug-gest that it may be unclear at present the typesof data that would be relevant for future retro-spective or meta-analysis, thus it would be inthe best interest of the community to providemore detailed information in publication whenprocedures are conducted, rather than referringto a common procedure and a brief statementof its outcomes.For treatment data, only the minority of

studies used a design that would not allow for ademonstration of experimental control. Muchmore commonly, a failure in level change wasobserved while using an appropriate single sub-ject design, thus not allowing for the demon-stration of experimental control. This findingshould be tempered by the known resistance ofautomatically reinforced SIB to treatment(e.g., Berg et al., 2016; Hagopian et al., 2017).However, given the heterogeneity of treatmentsattempted (21 treatment types reported in thisstudy), additional efforts should be made to usestrong experimental designs in future studies.Although the number of applications varied

across different treatment types, a number offindings are relevant for the clinical care of indi-viduals who engage in automatically reinforcedSIB. With the exception of DRA with DROand NCR, which was attempted only once,NCR was the most effective reinforcement-alonetreatment. When NCR was combined with anadditional component, the effectiveness of thisintervention improved for five out of six of thesecomponents. Although NCR was the most effec-tive reinforcement-alone treatment, both DRAand DRO were more significantly improvedthan NCR when other components were added.In fact, DRA and DRO with additional compo-nents were both more effective than NCR alone,and when all three reinforcement proceduresincluded additional components, the mean per-cent reduction was greater than 80% for alltreatments. Although this finding is likely relatedto a ceiling effect (i.e., NCR alone was alreadymostly effective, so there was little room for

989TREATMENT OF AUTOMATICALLY REINFORCED SIB

improvement), these results suggest that similareffectiveness may be obtained across reinforce-ment procedures when additional componentsare included, and that particular componentsmay complement each other in treatmentpackages.These results also suggested some differences

in the effectiveness of components used to treatautomatically reinforced SIB. As might beexpected, the most restrictive treatment(restraint) was highly effective regardless of thetype of reinforcement treatment used. Unex-pectedly, punishment (another component usu-ally considered very restrictive) when added toreinforcement alone was not as effective as oneof the less restrictive components (antecedentintervention), and was equally effective toanother of the less restrictive components(blocking). Research on the use of an antecedentintervention (other than NCR) in the treatmentof automatically reinforced SIB may be a goodarea for future investigation. These interventionsare less intrusive but appear highly effective;however, evaluations of antecedent componentsother than NCR are rare in the literature.Our analysis of the types of assessments used

to identify items for NCR and the subsequenteffectiveness of NCR is similar to findings byGroskreutz, Groskreutz, and Higbee (2011). Inthis study, the authors prospectively examinedthe effectiveness of NCR as a treatment for oneindividual’s vocal stereotypy using itemsderived from either a preference assessment orCSA. The authors found that using the itemsidentified from the CSA produced a moreeffective treatment. Although our study ana-lyzed the effectiveness of NCR across individ-uals based on how items were identified in aretrospective manner, and Groskreutz et al.examined stereotypy in a prospective manner,the results are similar and suggest that CSAsshould be used to identify items to use in NCRtreatments. Interestingly, data from Figure 4suggests that clinicians and researchers whopublished studies on the use of NCR for

automatically reinforced SIB have alreadyadopted the use of CSAs as the preferred meansto identify items for NCR.The method and results of the current analy-

sis were most comparable to the method andfinding in Kahng et al. (2002a), who reviewedpublished articles on behavioral treatments ofSIB (maintained by any function) over a35-year period. Our study overlapped withKahng et al. (2002a) in that both includedstudies with automatically reinforced SIBbetween 1982-2000. Similar to the Kahnget al. study, we found the use of punishmenthad decreased over time (90% reduction in thepast decade). In the place of punishment, wefound that reinforcement alone had becomemore effective in the past decade (an 11%increase in effectiveness over the decade prior)and that there was a minor increase in the useof less restrictive additional components. Theimprovement of treatments using reinforcementalone may be related to the increased use ofcompetition procedures to identify items forNCR over the same time period. However, incomparison to Kahng et al., across the entire34-year period, we found that reinforcementwas, on the whole, less effective. Kahnget al. found that the use of reinforcement aloneproduced a mean percent reduction in SIB of73.2% across all years included their study. Inthe current study, reinforcement alone was lesseffective, producing only a 52.5% mean per-cent reduction in SIB. It is unlikely that rein-forcement procedures have become lesseffective since the publication of Kahng et al.,thus the difference is likely related to the factthat we only included automatically reinforcedSIB in the review.There are several potential reasons why SIB

may be more resistant to reinforcement aloneas a treatment in our review. First, evidencefrom the current study suggests thatconsequence-based reinforcement interventions(i.e., DRA and DRO) are much less effectivefor automatically reinforced SIB (effective in

GRIFFIN W. ROOKER et al.990

only 5 of 16 consequence-only-based interven-tions) than NCR (effective in 28 of 37 anteced-ent-only-based interventions). It may be thatfor consequence-based interventions to be effec-tive, the individual must either engage in otherbehavior and/or refrain from engaging in SIBto obtain reinforcement. Thus, individuals whoengage in a high rate of automatically rein-forced SIB may never contact the consequence-based reinforcement contingency, due toresponse competition, decreasing the effective-ness of these interventions. Second, as discussedabove, using an item from a preference assess-ment in NCR may make NCR more likely tofail in the treatment of automatically reinforcedSIB, thus it is possible that our sample hadmore preference assessments than Kahnget al. (2002b). However, Kahng et al. likelyhad a similar distribution of preference assess-ments and CSAs (if not more preference assess-ments) used to identify items for NCR as thecurrent study (given the time period in whichthe studies Kahng et al. reviewed were con-ducted) and still found reinforcement moreeffective. In addition, in our study, there wereseven cases where a CSA was conducted andNCR was still found to be ineffective. Finally,there is evidence that some forms of automati-cally reinforced SIB are more resistant to treat-ment. For example, Hagopian et al. (2015) andHagopian et al. (2017) demonstrated that sub-types of automatically reinforced SIB mayrespond differently to reinforcement alone as atreatment. Furthermore, a subset of the indi-viduals in the current study was demonstratedto have subtype 2 SIB (see Hagopian et al.,2017). Unfortunately, in the current study, themajority of individuals could not be subtypeddue to the lack of published FAs. However, itis likely that on average reinforcement alonewas less effective in the current study becauseoverall treatment effectiveness was reduced bythe presence of subtypes of automatically rein-forced SIB associated with treatment resistance.

Several weaknesses in this study should benoted. First, because this was a retrospectivestudy of published research, the data are lim-ited. For example, it is highly likely that thedegree to which the individual interacts withitems in NCR determines its effectiveness(through response competition); however, thesedata could not be obtained from the studies wereviewed. Second, the inclusion criteria wereintentionally broad, and studies were includedwithout a conclusive demonstration that SIBwas maintained by automatic reinforcement(aside from the authors’ statement). Althoughthis may introduce some doubt into the currentfindings, in the majority of cases the authorseither presented an FA or noted that an FA wasconducted. Third, although we evaluated thecompleteness of articles, the degree to whichprocedures were implemented in exactly thesame way across studies was impossible todetermine, as these publications often did notpresent integrity data. Despite this, in mostcases, coders reported they could replicate thebasic components of the treatment (e.g., howoften reinforcement was programmed). Fourth,there was no uniform progression of treatmentsacross studies. Thus, the degree to which agiven treatment may or may not have beeneffective for an individual is unknowable andlimits the type of analyses that can be con-ducted. Finally, not all treatments were imple-mented as part of clinical care. That is, somestudies were not focused necessarily on themost effective treatment for automatically rein-forced SIB through an application of the leastto most intrusive treatments. Thus, more intru-sive treatments may have been attempted withindividuals who could have been successfullytreated with less intrusive treatments(e.g., Lerman, Iwata, Shore, & DeLeon, 1997).Alternatively, unpublished clinical treatmentsmay have been attempted prior to the pub-lished data we evaluated. In which case, infor-mation on the effectiveness or ineffectiveness oftreatments may exist and have informed the

991TREATMENT OF AUTOMATICALLY REINFORCED SIB

experimenters prior to their study. Both ofthese possibilities may artificially inflate ordeflate the degree to which additional treat-ment components were required to successfullytreat automatically reinforced SIB.Kahng et al. (2002a) indicated a decrease in

the use of punishment in the treatment of SIBin published studies across time, suggesting thatreinforcement alone as a treatment had becomemore effective for SIB. Results of the currentstudy demonstrated this was true for automati-cally reinforced SIB, but to a lesser extent.Thus, future research should examine means tomake reinforcement alone a more effectivetreatment for automatically reinforced SIB, andparticularly for treatment-resistant automati-cally reinforced SIB (e.g., subtype 2). For thesubset of individuals who engage in thistreatment-resistant SIB, research on makingalternative activities more attractive, trainingleisure skills, and making NCR more durablefor extended use is needed. In addition,research might examine how additional compo-nents influence item interaction in assessmentand treatment. Similarly, research is needed onthe least intrusive and most effective forms ofadditional components to complement rein-forcement treatments (e.g., response effortmanipulations). Finally, further research isneeded to examine individual, behavioral, andprocedural reasons why reinforcement alonefails as a treatment. The current analysis identi-fied NCR based on preference as one reasonNCR might be ineffective; however, in the cur-rent study, this is only a subset of the ineffec-tive reinforcement-alone treatments.For the clinical treatment of automatically

reinforced SIB, this review provides several dis-tinct recommendations. Importantly, NCR isthe most thoroughly studied treatment optionand when using a reinforcement procedurealone, NCR has been used successfully mostoften. Further, the effectiveness of NCR can beimproved by using items in the treatment thatare determined to compete with SIB in a

pretreatment assessment (i.e., CSA). However,NCR alone cannot be relied on to be effectivein every case of automatically reinforced SIB.Thus, in some cases additional treatment com-ponents will be necessary. When this is thecase, punishment has been used most often.However, our results suggest that a number ofless intrusive procedures (e.g., antecedents otherthan NCR, blocking, and protective equip-ment) have been equally effective and may infact produce a greater percent reduction inautomatically reinforced SIB. Thus, the pru-dent course when faced with automaticallyreinforced SIB would be to first use empiricalmeans to determine competing stimuli, evalu-ate those stimuli through NCR, and then addadditional, less restrictive components, relyingon punishment and restraint for only the mosttreatment-resistant cases.

REFERENCES

Baghdadli, A., Pascal, C., Grisi, S., & Aussilloux, C.(2003). Risk factors for self-injurious behavioursamong 222 young children with autistic disorders.Journal of Intellectual Disability Research, 47, 622-627. https://doi.org/10.1046/j.1365-2788.2003.00507.x

*Banda, D. R., McAfee, J. K., & Hart, S. L. (2012).Decreasing self-injurious behavior and fading self-restraint in a student with autism and Tourette syn-drome. Behavioral Interventions, 27, 164-174. https://doi.org/10.1002/bin.1344

Beavers, G. A., Iwata, B. A., & Lerman, D. C. (2013).Thirty years of research on the functional analysis ofproblem behavior. Journal of Applied BehaviorAnalysis, 46, 1-21. https://doi.org/10.1002/jaba.30

Berg, W. K., Wacker, D. P., Ringdahl, J. E., Stricker, J.,Vinquist, K., Salil Kumar Dutt, A., ... & Mews, J.(2016). An integrated model for guiding the selectionof treatment components for problem behavior main-tained by automatic reinforcement. Journal of AppliedBehavior Analysis, 49, 617-638. https://doi.org/10.1002/jaba.303

Bodfish, J. W., Symons, F. J., Parker, D. E., &Lewis, M. H. (2000). Varieties of repetitivebehavior in autism: Comparisons to mental retarda-tion. Journal of Autism and Developmental Disorders,30, 237-243. https://doi.org/10.1023/A:1005596502855

GRIFFIN W. ROOKER et al.992

Bowman-Perrott, L., Burke, M. D., Zaini, S.,Zhang, N., & Vannest, K. (2015). Promoting posi-tive behavior using the good behavior game a meta-analysis of single-case research. Journal of PositiveBehavior Interventions. 18, 180-190. https://doi.org/10.1177/1098300715592355

*Cavalari, R. N., DuBard, M., & Luiselli, J. K. (2013).Simplified habit reversal and treatment fading forchronic skin picking in an adolescent with autism.Clinical Case Studies, 13, 190-198 https://doi.org/10.1177/1534650113510348

Courtemanche, A., Schroeder, S., Sheldon, J.,Sherman, J., & Fowler, A. (2012). Observing signs ofpain in relation to self-injurious behaviour amongindividuals with intellectual and developmental dis-abilities. Journal of Intellectual Disability Research, 56,501-515. Accessed at: https://onlinelibrary.wiley.com/doi/10.1111/j.1365-2788.2011.01492.x/full

*Cowdery, G. E., Iwata, B. A., & Pace, G. M. (1990).Effects and side effects of DRO as treatment for self-injurious behavior. Journal of Applied Behavior Analy-sis, 23, 497-506. https://doi.org/10.1901/jaba.1990.23-497

*Davis, T. N., Dacus, S., Strickland, E.,Machalicek, W., & Coviello, L. (2013). Reduction ofautomatically maintained self-injurious behavior uti-lizing noncontingent matched stimuli. DevelopmentalNeurorehabilitation, 16, 166-171. https://doi.org/10.3109/17518423.2013.766819

*DeLeon, I. G., Anders, B. M., Rodriguez-Catter, V., &Neidert, P. L. (2000). The effects of noncontingentaccess to single- versus multiple-stimulus sets on self-injurious behavior. Journal of Applied Behavior Analy-sis, 33, 623-626. https://doi.org/10.1901/jaba.2000.33-623

*Demanche, J., & Chok, J. T. (2013). The use of wristweights and vibratory stimulation to treat self-injurious behavior. Journal of Developmental and Phys-ical Disabilities, 25, 79-90. https://doi.org/10.1007/s10882-012-9304-2

Didden, R., Korzilius, H., & Curfs, L. M. (2007). Skin-picking in individuals with Prader-Willi syndrome:Prevalence, functional assessment, and its comorbid-ity with compulsive and self-injurious behaviours.Journal of Applied Research in Intellectual Disabilities,20, 409-419. https://doi.org/10.1111/j.1468-3148.2007.00388.x

*Dozier, C. L., Vollmer, T. R., Borrero, J. C.,Borrero, C. S., Rapp, J. T., Bourret, J., &Gutierrez, A. (2007). Assessment of preference forbehavioral treatment versus baseline conditions.Behavioral Interventions, 22, 245-261. https://doi.org/10.1002/bin.241

Elsea, S. H., & Girirajan, S. (2008). Smith–Magenis syn-drome. European Journal of Human Genetics, 16,412-421. https://doi.org/10.1038/sj.ejhg.5202009

*Favell, J. E., McGimsey, J. F., & Schell, R. M. (1982).Treatment of self-injury by providing alternate

sensory activities. Analysis and Intervention in Devel-opmental Disabilities, 2, 83-104. https://doi.org/10.1016/0270-4684(82)90007-6

Fisher, W., Piazza, C. C., Bowman, L. G.,Hagopian, L. P., Owens, J. C., & Slevin, I. (1992).A comparison of two approaches for identifying rein-forcers for persons with severe and profound disabil-ities. Journal of Applied Behavior Analysis, 25, 491-498. https://doi.org/10.1901/jaba.1992.25-491

Groskreutz, M. P., Groskreutz, N. C., & Higbee, T. S.(2011). Response competition and stimulus prefer-ence in the treatment of automatically reinforcedbehavior: A comparison. Journal of Applied BehaviorAnalysis, 44, 211-215. https://doi.org/10.1901/jaba.2011.44-211

*Hagopian, L. P., Paclawskyj, T. R., & Kuhn, S. C.(2005). The use of conditional probability analysis toidentify a response chain leading to the occurrence ofeye poking. Research in Developmental Disabilities, 26,393-397. https://doi.org/10.1016/j.ridd.2003.09.002

Hagopian, L. P., Rooker, G. W., & Zarcone, J. R.(2015). Delineating subtypes of self-injurious behav-ior maintained by automatic reinforcement. Journalof Applied Behavior Analysis, 48, 523-543. https://doi.org/10.1002/jaba.236

Hagopian, L. P., Rooker, W. G., Zarcone, J. R.,Bonner, A. C., & Arevalo, A. R., (2017). Furtheranalysis of subtypes of automatically reinforced SIB:A replication and quantitative analysis of publisheddatasets. Journal of Applied Behavior Analysis, 50, 48-66. https://doi.org/10.1002/jaba.368

Hanley, G. P., Iwata, B. A., & McCord, B. E. (2003).Functional analysis of problem behavior: A review.Journal of Applied Behavior Analysis, 36, 147-185.https://doi.org/10.1901/jaba.2003.36-147

*Hanley, G. P., Piazza, C. C., Keeney, K. M., Blakeley-Smith, A. B., & Worsdell, A. S. (1998). Effects ofwrist weights on self-injurious and adaptive behaviors.Journal of Applied Behavior Analysis, 31, 307-310.https://doi.org/10.1901/jaba.1998.31-307

*Healey, J. J., Ahearn, W. H., Graff, R. B., &Libby, M. E. (2001). Extended analysis and treat-ment of self-injurious behavior. BehavioralInterventions, 16, 181-195. https://doi.org/10.1002/bin.91

*Higgins, W. J., Falcomata, T. S., Roane, H. S., &Thomas, K. M. (2012). Effects of body positioningon automatically reinforced self-injurious behaviorexhibited by a child with Duane syndrome and Gold-enhar syndrome. Journal of Developmental and Physi-cal Disabilities, 24, 135-144. https://doi.org/10.1007/s10882-011-9262-0

Hustyi, K. M., Hammond, J. L., Rezvani, A. B., &Hall, S. S. (2013). An analysis of the topography,severity, potential sources of reinforcement, andtreatments utilized for skin picking in Prader–Willisyndrome. Research in Developmental Disabilities, 34,

993TREATMENT OF AUTOMATICALLY REINFORCED SIB

2890-2899. https://doi.org/10.1016/j.ridd.2013.06.014

Iwata, B. A., Dorsey, M. F., Slifer, K. J.,Bauman, K. E., & Richman, G. S. (1994). Toward afunctional analysis of self-injury. Journal of AppliedBehavior Analysis, 27, 197-209. (Reprinted fromAnalysis and Intervention in DevelopmentalDisabilities, 2, 3-20, 1982) https://doi.org/10.1901/jaba.1994.27-197

*Iwata, B. A., Pace, G. M., Cowdery, G. E., &Miltenberger, R. G. (1994). What makes extinctionwork: An analysis of procedural form and function.Journal of Applied Behavior Analysis, 27, 131-144.https://doi.org/10.1901/jaba.1994.27-131

Iwata, B. A., Pace, G. M., Dorsey, M. F., Zarcone, J. R.,Vollmer, T. R., Smith, R. G., ... Willis, K. D.(1994). The functions of self-injurious behavior: Anexperimental-epidemiological analysis. Journal ofApplied Behavior Analysis, 27, 215-240. https://doi.org/10.1901/jaba.1994.27-215

*Jennett, H., Jann, K., & Hagopian, L. P. (2011). Evalu-ation of response blocking and re-presentation in acompeting stimulus assessment. Journal of AppliedBehavior Analysis, 44, 925-929. https://doi.org/10.1901/jaba.2011.44-925

*Kahng, S., Abt, K. A., & Wilder, D. A. (2001). Treat-ment of self-injury correlated with mechanicalrestraints. Behavioral Interventions, 16, 105-110.https://doi.org/10.1002/bin.86

Kahng, S., Iwata, B. A., & Lewin, A. B. (2002a). Behav-ioral treatment of self-injury, 1964 to 2000. AmericanJournal on Mental Retardation, 107, 212-221. https://doi.org/10.1352/0895-8017(2002)107%3C0212:BTOSIT%3E2.0.CO;2

Kahng, S., Iwata, B. A., & Lewin, A. B. (2002b). Theimpact of functional assessment on the treatment ofself-injurious behavior. In S. Schroeder, M. L. Oster-Granite, & T. Thompson (Eds.), Self-injurious behav-ior: Gene-brain-behavior relationships (pp. 119-131).Washington, DC: American PsychologicalAssociation.

Kennedy, C. H. (1994). Automatic reinforcement: Oxy-moron or hypothetical construct? Journal of Behav-ioral Education, 4, 387-395. https://doi.org/10.1007/BF01539540

*Kennedy, C. H., & Souza, G. (1995). Functional analy-sis and treatment of eye poking. Journal of AppliedBehavior Analysis, 28, 27-37. https://doi.org/10.1901/jaba.1995.28-27

Kennedy, C. H., & Thompson, T. (2000). Health condi-tions contributing to problem behavior among peoplewith mental retardation and developmental disabil-ities. In M. Wehmeyer & J. Patten (Eds.), Mentalretardation in the 21st century (pp. 211-231). Austin,TX: Pro-Ed.

*Kern, L., Bailin, D., & Mauk, J. E. (2003). Effects of atopical anesthetic on non-socially maintained self-injurious behavior. Developmental Medicine & Child

Neurology, 45, 769-771. https://doi.org/10.1017/S0012162203001427

*Kuhn, D. E., DeLeon, I. G., Fisher, W. W., &Wilke, A. E. (1999). Clarifying an ambiguous func-tional analysis with matched and mismatched extinc-tion procedures. Journal of Applied Behavior Analysis,32, 99-102. https://doi.org/10.1901/jaba.1999.32-99

*Ladd, M. V., Luiselli, J. K., & Baker, L. (2009). Contin-uous access to competing stimulation as interventionfor self-injurious skin picking in a child with autism.Child & Family Behavior Therapy, 31, 54-60. https://doi.org/10.1080/07317100802701400

*Lane, K. L., Thompson, A., Reske, C. L.,Gable, L. M., & Barton-Arwood, S. (2006). Reduc-ing skin picking via competing activities. Journal ofApplied Behavior Analysis, 39, 459-462. https://doi.org/10.1901/jaba.2006.62-05

*Lang, R., Didden, R., Sigafoos, J., Rispoli, M.,Regester, A., & Lancioni, G. E. (2009). Treatmentof chronic skin-picking in an adolescent with Asper-ger syndrome and borderline intellectual disability.Clinical Case Studies, 8, 317-325. https://doi.org/10.1177/1534650109341841

Lanovaz, M. J., & Sladeczek, I. E. (2012). Vocal stereo-typy in individuals with autism spectrum disorders areview of behavioral interventions. BehaviorModification, 36, 146-164. https://doi.org/10.1177/0145445511427192

LeBlanc, L. A., Patel, M. R., & Carr, J. E. (2000). Recentadvances in the assessment of aberrant behavior main-tained by automatic reinforcement in individuals withdevelopmental disabilities. Journal of Behavior Ther-apy and Experimental Psychiatry, 31, 137-154.https://doi.org/10.1016/S0005-7916(00)00017-3

Lee, J. H., Berkowitz, R. J., & Choi, B. J. (2002). Oralself-mutilation in the Lesch-Nyhan syndrome. Jour-nal of Dentistry for Children, 69, 66-69. Accessedat: https://www.ingentaconnect.com/content/aapd/jodc/2002/00000069/00000001/art00014#expand/collapse

*Lerman, D. C., Iwata, B. A., Shore, B. A., &DeLeon, I. G. (1997). Effects of intermittent punish-ment on self-injurious behavior: An evaluation ofschedule thinning. Journal of Applied Behavior Analy-sis, 30, 187-201. https://doi.org/10.1901/jaba.1997.30-187

Levy, A., DeLeon, I. G., Martinez, C. K., Fernandez, N.,Gage, N. A., Sigurdsson, S. O., & Frank-Crawford, M. (2017). A quantitative review of over-justification effects in persons with intellectual anddevelopmental disabilities. Journal of Applied BehaviorAnalysis. 50, 206-221. https://doi.org/10.1002/jaba.359

*Lindberg, J. S., Iwata, B. A., & Kahng, S. (1999). Onthe relation between object manipulation and stereo-typic self-injurious behavior. Journal of Applied Behav-ior Analysis, 32, 51-62. https://doi.org/10.1901/jaba.1999.32-51

GRIFFIN W. ROOKER et al.994

*Lindberg, J. S., Iwata, B. A., Roscoe, E. M.,Worsdell, A. S., & Hanley, G. P. (2003). Treatmentefficacy of noncontingent reinforcement during briefand extended application. Journal of Applied BehaviorAnalysis, 36, 1-19. https://doi.org/10.1901/jaba.2003.36-1

*Linscheid, T. R., Iwata, B. A., Ricketts, R. W.,Williams, D. E., & Griffin, J. C. (1990). Clinicalevaluation of the self-injurious behavior inhibitingsystem (SIBIS). Journal of Applied Behavior Analysis,23, 53-78. https://doi.org/10.1901/jaba.1990.23-53

*Long, E. S., Hagopian, L. P., DeLeon, I. G.,Marhefka, J. M., & Resau, D. (2005). Competingstimuli in the treatment of multiply controlled prob-lem behavior during hygiene routines. Research inDevelopmental Disabilities, 26, 57-69. https://doi.org/10.1016/j.ridd.2003.01.001

Ma, G., Wang, Q., Lv, C., Qiang, F., Hua, Q., Chu, H.,... & Gong, W. (2015). The prognostic significanceof HOTAIR for predicting clinical outcome inpatients with digestive system tumors. Journal of can-cer research and clinical oncology, 141, 2139-2145.https://doi.org/10.1007/s00432-015-1980-8

*Magnusson, A. F., & Gould, D. D. (2007). Reductionof automatically-maintained self-injury using contin-gent equipment removal. Behavioral Interventions, 22,57-68. https://doi.org/10.1002/bin.231

*Mason, S. A., & Iwata, B. A. (1990). Artifactual effectsof sensory-integrative therapy on self-injurious behav-ior. Journal of Applied Behavior Analysis, 23, 361.https://doi.org/10.1901/jaba.1990.23-361

Matson, J. L., & LoVullo, S. V. (2008). A review ofbehavioral treatments for self-injurious behaviors ofpersons with autism spectrum disorders. BehaviorModification, 32, 61-76. https://doi.org/10.1177/0145445507304581

*Morrison, H., Roscoe, E. M., & Atwell, A. (2011). Anevaluation of antecedent exercise on behavior main-tained by automatic reinforcement using a three-component multiple schedule. Journal of AppliedBehavior Analysis, 44, 523-541. https://doi.org/10.1901/jaba.2011.44-523

Nyhan, W. L. (1976). Behavior in the Lesch-Nyhansyndrome. Journal of Autism and ChildhoodSchizophrenia, 6, 235-252. https://doi.org/10.1007/BF01543464

*O’Reilly, M. F. (1996). Assessment and treatment ofepisodic self-injury: A case study. Research in Develop-mental Disabilities, 17, 349-361. https://doi.org/10.1016/0891-4222(96)00018-2

*O’Reilly, M. F., Murray, N., Lancioni, G. E.,Sigafoos, J., & Lacey, C. (2003). Functional analysisand intervention to reduce self-injurious and agitatedbehavior when removing protective equipment forbrief time periods. Behavior Modification, 27, 538-559. https://doi.org/10.1177/0145445503255573

Pace, G. M., Ivancic, M. T., Edwards, G. L.,Iwata, B. A., & Page, T. J. (1985). Assessment of

stimulus preference and reinforcer value with pro-foundly retarded individuals. Journal of AppliedBehavior Analysis, 18, 249-255. https://doi.org/10.1901/jaba.1985.18-249

*Patel, M. R., Carr, J. E., Kim, C., Robles, A., &Eastridge, D. (2000). Functional analysis of aberrantbehavior maintained by automatic reinforcement:Assessments of specific sensory reinforcers. Research inDevelopmental Disabilities, 21, 393-407. https://doi.org/10.1016/S0891-4222(00)00051-2

Phillips, C. L., Iannaccone, J. A., Rooker, G. W., &Hagopian, L. P. (2017). Noncontingent reinforce-ment for the treatment of severe problem behavior:An analysis of 27 consecutive applications. Journal ofApplied Behavior Analysis, 2, 357-376. https://doi.org/10.1002/jaba.376

Piazza, C. C., Bowman, L. G., Contrucci, S. A.,Delia, M. D., Adelinis, J. D., & Goh, H. L. (1999).An evaluation of the properties of attention as rein-forcement for destructive and appropriate behavior.Journal of Applied Behavior Analysis, 32, 437-449.https://doi.org/10.1901/jaba.1999.32-437

*Piazza, C. C., Fisher, W. W., Hanley, G. P.,Hilker, K., & Derby, K. M. (1996). A preliminaryprocedure for predicting the positive and negativeeffects of reinforcement-based procedures. Journal ofApplied Behavior Analysis, 29, 137-152. https://doi.org/10.1901/jaba.1996.29-137

Poisson, A., Nicolas, A., Cochat, P., Sanlaville, D.,Rigard, C., de Leersnyder, H., ... & Demily, C.(2015). Behavioral disturbance and treatment strate-gies in Smith-Magenis syndrome. Orphanet Journalof Rare Diseases, 10, 111. https://doi.org/10.1186/s13023-015-0330-x

Prieto, M. A., Vázquez, J. A., & Murado, M. A. (2014).A new mathematical model to quantify and character-ize the response to pro-and anti-oxidants of thecopper-induced oxidation of LDL assay. A tool forexamination of potential preventive compounds andclinical risk prediction. Food Research International,66, 501-513. https://doi.org/10.1016/j.foodres.2014.07.046

Rapp, J. T., & Vollmer, T. R. (2005). Stereotypy I: Areview of behavioral assessment and treatment.Research in Developmental Disabilities, 26, 527-547.https://doi.org/10.1016/j.ridd.2004.11.005

Richman, D. M., & Lindauer, S. E. (2005). Longitudinalassessment of stereotypic, proto-injurious, and self-injurious behavior exhibited by young children withdevelopmental delays. American Journal on MentalRetardation, 110, 439-450. https://doi.org/10.1352/0895-8017(2005)110%5B439:LAOSPA%5D2.0.CO;2

*Richman, D. M., Wacker, D. P., Asmus, J. M., &Casey, S. D. (1998). Functional analysis and extinc-tion of different behavior problems exhibited by thesame individual. Journal of Applied Behavior Analysis,

995TREATMENT OF AUTOMATICALLY REINFORCED SIB

31, 475-478. https://doi.org/10.1901/jaba.1998.31-475

*Ringdahl, J. E., Vollmer, T. R., Marcus, B. A., &Roane, H. S. (1997). An analogue evaluation of envi-ronmental enrichment: The role of stimulus prefer-ence. Journal of Applied Behavior Analysis, 30, 203-216. https://doi.org/10.1901/jaba.1997.30-203

*Rosales, R., Worsdell, A., & Trahan, M. (2010). Com-parison of methods for varying item presentation dur-ing noncontingent reinforcement. Research in AutismSpectrum Disorders, 4, 367-376. https://doi.org/10.1016/j.rasd.2009.10.004

*Roscoe, E. M., Iwata, B. A., & Goh, H. L. (1998). Acomparison of noncontingent reinforcement and sen-sory extinction as treatments for self-injurious behav-ior. Journal of Applied Behavior Analysis, 31, 635-646.https://doi.org/10.1901/jaba.1998.31-635

Roscoe, E. M., Iwata, B. A., & Zhou, L. (2013). Assess-ment and treatment of chronic hand mouthing. Jour-nal of Applied Behavior Analysis, 46, 181-198. https://doi.org/10.1002/jaba.14

Sham, E., & Smith, T. (2014). Publication bias in studiesof an applied behavior-analytic intervention: An ini-tial analysis. Journal of Applied Behavior Analysis, 47,663-678. https://doi.org/10.1002/jaba.146

*Shore, B. A., Iwata, B. A., DeLeon, I. G., Kahng, S., &Smith, R. G. (1997). An analysis of reinforcer substi-tutability using object manipulation and self-injury ascompeting responses. Journal of Applied BehaviorAnalysis, 30, 21-41. https://doi.org/10.1901/jaba.1997.30-21

*Slifer, K. J., Iwata, B. A., & Dorsey, M. F. (1984).Reduction of eye gouging using a response interrup-tion procedure. Journal of Behavior Therapy andExperimental Psychiatry, 15, 369-375. https://doi.org/10.1016/0005-7916(84)90104-6

*Smith, R. G., Iwata, B. A., Vollmer, T. R., &Zarcone, J. R. (1993). Experimental analysis andtreatment of multiply controlled self-injury. Journal ofApplied Behavior Analysis, 26, 183-196. https://doi.org/10.1901/jaba.1993.26-183

Soke, G. N., Rosenberg, S. A., Hamman, R. F.,Fingerlin, T., Robinson, C., Carpenter, L., ... &DiGuiseppi, C. (2016). Brief report: Prevalence ofself-injurious behaviors among children with AutismSpectrum Disorder—A population-based study. Jour-nal of Autism and Developmental Disorders, 46, 3607-3614. https://doi.org/10.1007/s10803-016-2879-1

*Stokes, J. V., & Luiselli, J. K. (2008). Applied behavioranalysis assessment and intervention for health:Threatening self-injury (rectal picking) in an adultwith Prader-Willi syndrome. Clinical Case Studies, 8,38-47. https://doi.org/10.1177/1534650108327011

Symons, F. J., Butler, M. G., Sanders, M. D.,Feurer, I. D., & Thompson, T. (1999). Self-injuriousbehavior and Prader-Willi syndrome: behavioralforms and body locations. American Journal on Men-tal Retardation, 104, 260-269. https://doi.org/10.

1352/0895-8017(1999)104%3C0260:SBAPSB%3E2.0.CO;2

Symons, F. J., Byiers, B. J., Raspa, M., Bishop, E., &Bailey, D. B. (2010). Self-injurious behavior andfragile X syndrome: Findings from the national fragileX survey. American Journal of Intellectual andDevelopmental Disabilities, 115, 473-481. https://doi.org/10.1352/1944-7558-115.6.473

Tate, B. G., & Baroff, G. S. (1966). Aversive control ofself-injurious behavior in a psychotic boy. BehaviourResearch and Therapy, 4, 281-287. https://doi.org/10.1016/0005-7967(66)90024-6

*Thompson, R. H., Iwata, B. A., Conners, J., &Roscoe, E. M. (1999). Effects of reinforcement foralternative behavior during punishment of self-injury.Journal of Applied Behavior Analysis, 32, 317-328.https://doi.org/10.1901/jaba.1999.32-317

Thompson, T., & Caruso, M. (2002). Self-injury: Know-ing what we’re looking for. In S. R. Schroeder,M. L. Oster-Granite, & T. Thompson (Eds.), Self-injurious Behavior: Gene-brain-behavior relationships(pp. 3-21). Washington, DC: American PsychologicalAssociation.

Thompson, T., Symons, F., Delaney, D., & England, C.(1995). Self-injurious behavior as endogenousneurochemical self-administration. DevelopmentalDisabilities Research Reviews, 1, 137-148. Accessedat: https://onlinelibrary.wiley.com/doi/10.1002/mrdd.1410010210/full

*Tiger, J. H., Fisher, W. W., & Bouxsein, K. J. (2009).Therapist- and self-monitored DRO contingencies asa treatment for the self-injurious skin picking of ayoung man with Asperger syndrome. Journal ofApplied Behavior Analysis, 42, 315-319. https://doi.org/10.1901/jaba.2009.42-315

*Toole, L. M., DeLeon, I. G., Kahng, S., Ruffin, G. E.,Pletcher, C. A., & Bowman, L. G. (2004). Re-evaluation of constant versus varied punishers usingempirically derived consequences. Research in Devel-opmental Disabilities, 25, 577-586. https://doi.org/10.1016/j.ridd.2004.03.005

*Van Camp, C. M., Lerman, D. C., Kelley, M. E.,Roane, H. S., Contrucci, S. A., & Vorndran, C. M.(2000). Further analysis of idiosyncratic antecedentinfluences during the assessment and treatment of prob-lem behavior. Journal of Applied Behavior Analysis, 33,207-221. https://doi.org/10.1901/jaba.2000.33-207

*Van Camp, C. M., Vollmer, T. R., & Daniel, D.(2001). A systematic evaluation of stimulus prefer-ence, response effort, and stimulus control in thetreatment of automatically reinforced self-injury.Behavior Therapy, 32, 603-613. https://doi.org/10.1016/S0005-7894(01)80037-X

*Van Houten, R. (1993). The use of wrist weights toreduce self-injury maintained by sensory reinforce-ment. Journal of Applied Behavior Analysis, 26, 197-203. https://doi.org/10.1901/jaba.1993.26-197

GRIFFIN W. ROOKER et al.996

Vaughan, M. E., & Michael, J. L. (1982). Automaticreinforcement: An important but ignored concept.Behaviorism, 10, 217-227. Accessed at: http://www.jstor.org/tc/accept?origin=/stable/pdf/27759007.pdf

Vollmer, T. R. (1994). The concept of automatic rein-forcement: Implications for behavioral research indevelopmental disabilities. Research in DevelopmentalDisabilities, 15, 187-207. https://doi.org/10.1016/0891-4222(94)90011-6

*Vollmer, T. R., Marcus, B. A., & LeBlanc, L. (1994).Treatment of self-injury and hand mouthing follow-ing inconclusive functional analyses. Journal ofApplied Behavior Analysis, 27, 331-344. https://doi.org/10.1901/jaba.1994.27-331

*Wesolowski, M. D., & Zawlocki, R. J. (1982). The dif-ferential effects of procedures to eliminate an injuri-ous self-stimulatory behavior (digito-ocular sign) inblind retarded twins. Behavior Therapy, 13, 334-345.https://doi.org/10.1016/S0005-7894(82)80042-7

Zhou, Q., Chen, A., Song, H., Tao, J., Yang, H., &Zuo, M. (2015). Prognostic value of cancer stem cell

marker CD133 in ovarian cancer: a meta-analysis.International Journal of Clinical & ExperimentalMedicine, 8, 3080-3088. Accessed at: http://pubmedcentralcanada.ca/pmcc/articles/PMC4443030/

Studies marked with an asterisk include datasummarized in this review.

Received October 14, 2016Final acceptance October 19, 2017Action Editor, SungWoo Kahng.

SUPPORTING INFORMATION

Additional Supporting Information may befound in the online version of this article at thepublisher’s website.

997TREATMENT OF AUTOMATICALLY REINFORCED SIB