Research in Developmental Disabilities 30 (2009) 409–425

Contents lists available at ScienceDirect

Research in Developmental Disabilities

Review

A review of empirical support for differential reinforcement of alternative behavior

Erin S. Petscher a,*, Catalina Rey b, Jon S. Bailey b

a University of Florida, United States b Florida State University, United States

A R T I C L E I N F O A B S T R A C T

Article history: Received 16 July 2008 Received in revised form 12 August 2008 Accepted 30 August 2008

Keywords: Differential reinforcement of alternative behavior Functional communication training Empirically supported treatments Destructive behavior Problem behavior Aberrant behavior Behavior analysis

Differential reinforcement of alternative behavior (DRA) is one of the most common behavior analytic interventions used to decrease unwanted behavior. We reviewed the DRA literature from the past 30 years to identify the aspects that are thoroughly researched and those that would benefit from further emphasis. We found and coded 116 empirical studies that used DRA, later grouping them into categories that met APA Division 12 Task Force criteria. We found that DRA has been successful at reducing behaviors on a continuum from relatively minor problems like prelinguistic communication to life-threatening failure to thrive. DRA with and without extinction is well established for treating destructive behavior of those with developmental disabilities, and to combat food refusal.

� 2008 Elsevier Ltd. All rights reserved.

Contents

1. Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 411 1.1. Inclusion and exclusion criteria. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 411 1.2. Experimental design. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 411 1.3. Participant and setting characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 411 1.4. Behavior topography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 411 1.5. Assessments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 412 1.6. Procedural manipulations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 412 1.7. Generalization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 412

* Corresponding author. Tel.: +1 850 322 1797. E-mail address: [email protected] (E.S. Petscher).

0891-4222/$ – see front matter � 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.ridd.2008.08.008

410 E.S. Petscher et al. / Research in Developmental Disabilities 30 (2009) 409–425

1.8. Task Force criteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 412 1.9. Inter-rater reliability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 412

2. Results and discussion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 412 2.1. Experimental design. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 413 2.2. Participants and settings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 414 2.3. Behavior topography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 416 2.4. Assessments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 417 2.5. Procedural variations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 417

2.5.1. Schedule thinning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 417 2.6. Generalization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 418 2.7. Empirically supported treatments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 418

2.7.1. DRA with extinction is a well-established treatment for destructive behavior . . . 419 2.7.2. DRA without extinction is a well-established treatment for destructive behavior . . . 419 2.7.3. DRA with extinction is well established for the treatment of food refusal. . . . . . . 419 2.7.4. DRA plus NCR is an experimental treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 419

3. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 420 Acknowledgements. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 420 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 420

Differential reinforcement is the withholding of reinforcers for one behavior and delivering them for another. Through the years many specific variations of differential reinforcement have emerged to demonstrate impressive behavior reductions with difficult clients (see Vollmer & Iwata, 1992). In fact, Lennox, Miltenberger, Spengler, and Efanian (1988) found that differential reinforcement (DR) procedures are among the most frequently used to suppress unwanted behaviors.

In recent years the field of behavior analysis has moved toward the use of function-based treatments to reduce unwanted behaviors. Function-based treatments are considered to be among the most efficient and effective behavioral interventions (Beare, Severtson, & Brandt, 2004; Iwata, Dorsey, Slifer, Bauman, & Richman, 1982/1994). Differential reinforcement is particularly appropriate today as the extinction component requires the assessment of maintaining variables, helping it meet the definition a function-based treatment.

Extinction alone may produce more unwanted side effects than when it is combined with another intervention (Lerman, Iwata, & Wallace, 1999; Petscher & Bailey, 2008). Differential reinforcement of alternative behavior (DRA) withholds the reinforcer for unwanted behavior while simultaneously reinforcing a specific, alternative response (Cooper, Heron, & Heward, 2007; Vollmer, Roane, Ringdahl, & Marcus, 1999). DRA may be the ideal intervention in many cases because it reduces behavior without the concern of extinction-induced side effects and provides an appropriate option for the clients to earn valuable reinforcers once they are no longer provided for unwanted behaviors (Rolider & Van Houten, 1990).

Some DRA literature has been reviewed for specific client types. For example, Mirenda reviewed the use of functional communication training through augmentative alternative communication devices (1997). Matson, Dixon and Matson (2005) reviewed treatment of aggressive behaviors among those with developmental disabilities, finding that a variation of DRA, functional communication training (FCT), is one of the most common interventions. However, a comprehensive review of DRA has not yet been performed. The authors of the current study feel that demonstrating the empirical support of DRA to those outside of applied behavior analysis is important for the dissemination of the field. Therefore, in addition to summarizing the status of DRA literature, the current study also utilizes the model offered by Severtson, Carr, and Lepper (2008) by coding articles according to the Task Force criteria.

Since 1995, The Division 12 Task Force on the Promotion and Dissemination of Psychological Procedures (Task Force) has been identifying treatments that meet the criteria they set to demonstrate empirical validation. Members of the Task Force publish lists of interventions found to meet their standards as well established, probably efficacious, or experimental (Chambless et al., 1996; Task Force, 1995). Unfortunately, while many behavioral interventions may be effective and efficacious, few have been included in the Task Force lists (see Chambless & Hollon, 1998).

411 E.S. Petscher et al. / Research in Developmental Disabilities 30 (2009) 409–425

The purposes of the current study are to review the published studies on DRA in the past 30 years, report the characteristics that meet criteria for being well established or probably efficacious, and to identify the variants that need further research.

1. Methods

The articles selected were each evaluated by the first or second author, and some were also reviewed by a second, independent observer who was trained to search for the relevant elements. Observers coded information about the participants, variables, interventions, results and follow-up.

1.1. Inclusion and exclusion criteria

Peer-reviewed journal articles that reported data on an intervention with DRA since 1977 were reviewed for this paper. The search terms, ‘‘differential reinforcement of alternative behavior’’, and ‘‘functional communication training’’, ‘‘DRA’’, and ‘‘FCT’’, were used within the search engines PsychInfo and Educational Resources Information Center (ERIC), Journal of Applied Behavior Analysis abstracts and Journal of the Experimental Analysis of Behavior abstracts. Reference lists of the articles obtained were also scanned for interventions that matched the definition but did not actually report it as DRA or FCT (e.g., Coe et al., 1997; Riordan, Iwata, Finney, Wohl, & Stanley, 1984). Search criteria permitted all dates through the end of 2007 and only the English language. Eliminations were made for those who either did not present data or used non-human subjects. Review articles and group designs that did not present individual participants’ data were excluded.

1.2. Experimental design

Experimental design was tracked by participants rather than studies. The designs were matched by definitions given by Bailey and Burch (2002). AB designs were listed if intervention just followed baseline. A multielement design involved the rapid alternation between at least two conditions. A reversal design required that at least two conditions were introduced twice. Variations of reversal designs, such as an ABCBA, where additional conditions were added to the design, were included. A multiple baseline was marked if two or more baselines of different durations were followed by the intervention. Those that added other conditions without replication, such as an ABCDE were categorized as additive designs. Changing criterion designs involved the stepwise progression of behavior requirements. Finally, if the participants were introduced to designs with multiple characteristics, this was tracked as a combination.

The experimenter who implemented the intervention was recorded as staff, parents or family members, teachers, therapist, or a behavior analyst. When not specified or the data collectors were trained observers or interns, these were considered behavior analysts.

1.3. Participant and setting characteristics

The number and age of participants was recorded, along with the first two diagnoses reported for each. Participants were grouped as children (up to 18 years old), or adults (19 and over). If a participant was diagnosed with an intellectual or physical disability that could be considered a developmental disability other than autism this was also recorded. Autism was tracked separately to show any distinction, although it was combined with developmental disabilities when studies were evaluated for Task Force criteria. The settings were inpatient facilities, clinics, schools, home, and vocational programs.

1.4. Behavior topography

The topographies of alternative behavior and problem behaviors were reported. Aggression, self- injurious behavior, property destruction, and disruptions were later grouped to encompass destructive behavior (Fisher, Thompson, Hagopian, Bowman, & Krug, 2000). The article reviewers

412 E.S. Petscher et al. / Research in Developmental Disabilities 30 (2009) 409–425

also tracked whether the individual data were presented for problem behavior, replacement behavior or both.

1.5. Assessments

Articles were scanned for reports and results of preference, reinforcer, and functional assessments. Functional assessments were tracked as descriptive, indirect, antecedent-behavior-consequence, motivation assessment scale (Durand, 1990), functional analysis or as a modified experimental analysis.

1.6. Procedural manipulations

While all studies included in this paper involved the use of DRA, some could be further distinguished as differential negative reinforcement of alternative behavior (DNRA) and FCT. DNRA specifies that the alternative behavior is negatively reinforced by the removal of an aversive stimulus. FCT requires that the presumed function of the unwanted behavior is made contingent upon the alternative response, rather than utilizing any or several arbitrary reinforcers. The specific style of DRA used was coded by reviewers.

Many studies manipulated antecedents and consequences in addition to the DRA intervention. Some were combined with DRA and these modifications were tracked as function-based or arbitrary. In addition, recorders reported the schedule thinning techniques that were utilized.

1.7. Generalization

When experimenters reported a generalization or follow-up condition these and the data trend were monitored. In addition, if side effects were reported, the perceived value of them was tracked as desired or undesired.

1.8. Task Force criteria

The criteria set by the Task Force for single subject designs require that the intervention is compared to a control or other acceptable treatment, has a treatment manual or other specific instructions, appropriate design and practically significant results, participants with details reported and similar characteristics, and at least two different investigators performed the studies (Chambless et al., 1996; Task Force, 1995). It is well established if at least 10 single subject cases meet criteria.

The Task Force criteria were used to group studies along identical interventions and similar participant characteristics and target behaviors. Reversal, multielement, some combinations and concurrent multiple baselines were all considered appropriate experimental designs if the first author found visually significant differences in trend, level, and variability (Bailey & Burch, 2002). However, a multiple baseline differed from the previous definition as a minimum of three panels were required to have the possibility of demonstrating control (Chambless & Hollon, 1998).

1.9. Inter-rater reliability

A trained, second reader independently analyzed 35% of the articles to determine inter-rater agreement on the evaluation of the studies. Agreement was calculated by an item-by-item comparison of data sheets. The number of item agreements was divided by the total number of agreements plus disagreements and multipled by 100 (Hanley, Iwata, & McCord, 2003). Average inter- rater agreement was 93% (range 82–100%).

2. Results and discussion

The initial search identified 538 articles and 116 remained after the exclusionary criteria were applied. These articles were each reviewed multiple times, and are all denoted in the reference list.

413 E.S. Petscher et al. / Research in Developmental Disabilities 30 (2009) 409–425

Table 1 Experimental design and experimenter type for all reviewed studies

Number Percent

Design AB 5 4.3 Additive 8 6.9 Combination 32 27.6 Multielement 7 6.0 Multiple baseline 29 25.0 Other 10 8.7 Reversal 25 21.6

Experimenter Behavior analyst 70 60.3 Combination 3 2.6 Parent/family 11 9.5 Staff 13 11.2 Teacher 8 6.9 Therapist 11 9.5

Although the findings and designs of all 116 did not meet Task Force criteria, many still offered substantial contributions to the DRA body of literature, so were included in the sample of articles used for Tables 1–4.

2.1. Experimental design

Table 1 displays the experimental design and experimenter type for 116 articles that utilized DRA. Many research designs replicated variables in order to properly demonstrate experimental control (Ringdahl et al., 2002; Roberts, Mace, & Daggett, 1995; Vollmer, Iwata, Smith, & Rodgers, 1992). Some authors selected experimental designs that involved no or limited opportunity to demonstrate experimental control (Beare et al., 2004; Dura, 1991; Earles & Myles, 1994; Walsh, 1991). As the control is demonstrated by the replication of the findings in single subject designs, such findings

Table 2 Age of participants, different settings, and first two diagnoses for participants

Characteristic Number Percent

Age category Child 277 82.4 Adult 59 17.6

Total participants 336 100

Diagnosis Autism 116 27.4 Developmental disability 271 64.1 Gastroesophageal reflux 6 1.4 None 3 0.7 Other 27 6.4

Total diagnoses 423 100

Setting Clinic 17 14.7 Home 21 18.1 Hospital 36 31.0 Multiple 4 3.4 Other 3 2.6 School 32 27.6 Vocational 3 2.6

Total studies 116 100

414 E.S. Petscher et al. / Research in Developmental Disabilities 30 (2009) 409–425

Table 3 Topography of alternative and unwanted behaviors

Number Percent

Problem behavior Aggression 31 9.2 Destruction/combination 160 47.6 Disruption 9 2.7 Food rejection/packing 15 4.5 Not specified/unclear 11 3.3 Other 26 7.7 Self-Injury 74 22.0 Vocalizations 10 3.0

Alternative behavior Communication 287 85.4 Compliance 8 2.4 Exchange cigarette 4 1.1 Food acceptance 18 5.4 Play/social 4 1.1 Task engagement 13 3.9 Transitions 2 0.6

should be replicated either across participants, settings, behaviors, or intervention phases. The AB and most additive designs alone, therefore, added little to indicate that the interventions were responsible for any behavior changes.

Most studies used a combination of at least two designs, which often were needed to evaluate multiple components of interventions. For example, in one study a reversal design was used first with full implementation of DRA compared to baseline. Next, modified versions of DRA were alternated to demonstrate the effects of partial DRA implementation on problem and alternative behavior (Vollmer et al., 1999). Other combination designs did not provide the opportunity to display experimental control, but still suggested interventions for further inquiry. For example, Fisher et al. (2005) combined a reversal and multielement design for Kim. However, the baseline conditions were too short to indicate a trend. These data suggested a positive response to the intervention and indicated that further study could be beneficial.

Behavior analysts typically performed the interventions. Such studies are likely to produce consistent results and treatment fidelity (Vollmer et al., 1999). However, those studies in which parents or staff implemented the intervention were programmed for generalization from the beginning. For example, Bird, Dores, Moniz, and Robinson (1989) performed their study in a classroom and the teacher enforced the intervention. This should increase consumer’s confidence that the intervention would continue successfully after the study ended. Unfortunately, it can be especially difficult to train teachers or parents to perform interventions while they are also held responsible for their typical duties. It can also confound the intervention and cause confusion if it does not sufficiently produce behavior change. Therefore, an approach commonly employed is to implement the intervention first with a trained experimenter, then transfer the training to someone who naturally interacts with the participant in the absence of a study. For example, after Goh, Iwata, and Kahng (1999) successfully reduced cigarette pica, they reportedly transferred the intervention to multiple settings and therapists.

2.2. Participants and settings

The number, first two diagnoses and age of participants are reported in Table 2. Over 80% of the participants were children, which corresponded with a similarly high percent of studies located in schools. Wright-Gallo et al. (2006) performed one such study, in which they achieved their goal of performing classroom-based functional analyses and DRA treatments for students diagnosed with emotional/behavioral disorders. Furthermore, their findings support the use of function-based interventions in the actual classroom, rather than an empty treatment room at a school or other location.

415 E.S. Petscher et al. / Research in Developmental Disabilities 30 (2009) 409–425

Table 4 Studies that demonstrated DRA with extinction as a well-established treatment of destructive behavior for children and adults with developmental disabilities

Study Participant Problem behaviora Alternative behaviorb Function

Children Bowman et al. (1997) Ben PD, SIB, AGG COM Compliance from

others Carr and Durand (1985) Jim AGG, TAN, OP Academic tasks Attention Carr and Durand (1985) Eve AGG, SIB, OP Academic tasks Attention Carr and Durand (1985) Tom AGG, TAN, OS Academic tasks Attention Carr and Durand (1985) Sue AGG, TAN, SIB Academic tasks Attention DeLeon et al. (2000) Jake AGG COM Tangible Derby et al. (1998) Lana SIB COM Attention Durand (1999) Allison Crying, screaming COM Tangible Durand (1999) Mike SIB, PD COM Tangible Durand (1999) Ron AGG COM Attention Durand (1999) David SIB COM Escape Durand (1999) Matt SIB, screaming COM Escape Durand and Carr (1991) Hal SIB Correct responses Escape Durand and Carr (1991) Ben AGG Correct responses Attention, escape Durand and Carr (1991) Tim SIB, AGG Correct responses Escape Durand and Carr (1992) Sam TAN, PD Correct responses Attention Durand and Carr (1992) Ted TAN, PD, OP Correct responses Attention Durand and Carr (1992) Ian TAN, PD, OP Correct responses Attention Durand and Carr (1992) Ray AGG, TAN, PD Correct responses Attention Durand and Carr (1992) Mike TAN Correct responses Attention Durand and Carr (1992) Jaynie OP, TAN Correct responses Attention Fisher, Adelinis, Thompson, Tina AGG COM Escape

Worsdell, and Zarcone (1998) Fisher, Adelinis, et al. (1998) Ike AGG COM Escape Fisher, Kuhn, et al. (1998) Ned AGG, PD COM Attention Fisher, Kuhn, et al. (1998) Amy SIB, AGG, PD COM Tangible, attention Fisher et al. (2000) Ken AGG, SIB, PD COM Attention Fyffe, Kahng, Fittro, and Russell Matt ISB COM Attention

(2004) Hagopian et al. (1998) Case 17 SIB, AGG, DIS COM Tangible Hagopian et al. (1998) Case 19 SIB, AGG, DIS COM Attention, escape Hagopian, Kuhn, Long, and Rush Stephen SIB, AGG, DIS COM Attention

(2005) Hagopian et al. (2005) James SIB, AGG, DIS COM Attention Hagopian et al. (2005) Matt AGG COM Tangible Hanley et al. (1997) Tony AGG, DIS COM Attention Hanley et al. (1997) Carla DIS, AGG COM Attention Kahng, Hendrickson, and Vu (2000) Ashby SIB, AGG, DIS COM Tangible Kelley, Lerman, and Van Camp Gary AGG COM Escape

(2002) Lalli et al. (1995) Joe SIB COM Escape Lalli et al. (1995) Jen SIB COM Escape Lalli et al. (1995) Kim AGG COM Escape Mancil et al. (2006) Scott TAN COM Tangible Marcus and Vollmer (1996) CJ SIB, AGG, DIS COM Tangible Piazza et al. (1999) Ike TAN COM Attention Roane et al. (2004) Carl AGG COM Tangible Roberts et al. (1995) Mary SIB Tasks Escape Sigafoos and Meikle (1996) Pete AGG, SIB, DIS COM Attention, tangible Sigafoos and Meikle (1996) Dale AGG, SIB, DIS COM Attention, tangible Thompson, Fisher, Piazza, Ernie AGG COM Attention

and Kuhn (1998) Vollmer et al. (1999) Rachel SIB, AGG Compliance, COM Escape Vollmer et al. (1999) Todd SIB Compliance, COM Tangible Vollmer et al. (1999) Kyle AGG Compliance, COM Escape

Adults Bird et al. (1989) Jim AGG, SIB COM Multiple Fisher et al. (2000) Glen SIB, AGG, PD COM Tangible

416 E.S. Petscher et al. / Research in Developmental Disabilities 30 (2009) 409–425

Table 4 (Continued )

Study Participant Problem behaviora Alternative behaviorb Function

Hanley, Iwata, and Thompson Karen SIB COM Attention (2001)

Hanley et al. (2001) Jake SIB COM Tangible Hanley et al. (2001) Julie SIB, AGG COM Attention Kahng, Iwata, DeLeon, and Todd SIB COM Escape

Worsdell (1997) Kahng et al. (1997) Lynn SIB COM Attention Kahng et al. (1997) Bob SIB COM Escape Lindauer et al. (2002) Sam SIB COM Attention Vollmer et al. (1992) Bob SIB Compliance Escape, attention

a AGG = aggression; SIB = self-injurious behavior; DIS = disruption; ISB = inappropriate sexual behavior; TAN = tantrum; PD = property destruction; OP = oppositional behavior; OS = out of seat. b COM = communication.

Adults exhibiting behavior problems may pose additional challenges as their longer learning histories and physical size is difficult when aggression or escape-maintained behaviors are present. In one study, a 24-year old with a developmental disability was treated in a multielement design comparing baseline and two forms of FCT. Although one form of FCT improved communication, both still resulted in variable rates of aggression (Bailey, McComas, Benavides, & Lovascz, 2002).

Most studies were performed in inpatient hospitals, where clients typically have severe and urgent needs for interventions that work quickly. The choice to utilize DRA in these cases indicates that experimenters are confident of its efficacy. Schools housed many studies as well, supporting the use of FCT as an instructional technique necessary for successful school behavior.

The vast majority of participants in DRA studies were diagnosed with autism or another developmental disability. Considering the impressive overall findings of this literature, the preference for clients with developmental disabilities over other diagnoses may suggest that they are the best fit for DRA interventions. However, some authors also utilized DRA to improve problem behavior of those with other diagnoses. Wilder and colleagues used DRA to decrease bizarre vocalizations for a 43-year old diagnosed with Schizophrenia (2001). The impressive behavior reduction indicated that DRA may be effective with similar clients and aberrant behavior in the future.

2.3. Behavior topography

Table 3 lists the unwanted and alternative behavior topographies that were targeted during DRA studies. Destruction included a combination of aggressive and disruptive behaviors, and was the most common presenting problem. Many authors published data on several of their clients’ destructive behaviors (e.g., LeBlanc, Hagopian, Marhefka, & Wilke, 2001; Piazza et al., 1999), while others opted to demonstrate how DRA improved specific topographies (e.g., McCord, Thomson, & Iwata, 2001). In one study, authors tracked changes in aggression and self-injury separately, demonstrating that the behavior frequencies varied based on the corresponding reinforcer schedule (Lindauer, Zarcone, Richman, & Schroeder, 2002).

In addition to reducing destructive behaviors, DRA has been effective in treating severe behaviors for children diagnosed with a failure to thrive (Kahng, Tarbox, & Wilke, 2001; Peck, Wacker, Berg, & Cooper, 1996). In these cases it may be vital for effective interventions to be provided immediately, and the data show that DRA satisfies the requirements.

Some behaviors targeted by authors are less severe but can lead to more restrictive placements if left untreated. For example, inappropriate vocalizations are often targeted in schools and many produce positive results (e.g., Dixon et al., 2004; Keen, Sigafoos, & Woodyatt, 2001; Lee, McComas, & Jawor, 2002). Alternative behaviors ranged from task completion to cigarette exchanges, with appropriate communication being the most frequent. It is an obvious target with DRA because conceptually the need is to improve client’s repertoires by adding appropriate ways for them to obtain reinforcers. When these reinforcers become available for other behaviors, those that require the least response effort will be utilized. In a fascinating display, the DRA intervention was provided as baseline

417 E.S. Petscher et al. / Research in De