Preference for blocking or response redirection during stereotypy treatment

Abstract

Response redirection and response blocking reduce stereotypy maintained by automatic reinforcement. The current study evaluated the effects of redirection and response blocking on the stereotypic responding of three elementary-age children diagnosed with autism. During the treatment evaluation, redirection and response blocking were evaluated using an alternating treatment embedded in a reversal design. Both procedures resulted in comparably low levels of motor stereotypy. Following treatment evaluation, a concurrent chain was conducted to evaluate participant preference for redirection or response blocking. All three participants preferred redirection. Practitioners may wish to consider participant preference when developing and implementing treatments for stereotypy.

Introduction

Stereotypy involves repetitive motor and vocal responses and is part of the diagnostic criteria for autism (Bodfish, Symons, Parker, & Lewis, 2000). Although stereotypy is often part of typical development (Foster, 1998), it has a greater prevalence in individuals with autism (Bodfish et al., 2000). Stereotypy often warrants intervention because it is associated with undesirable social consequences (Jones, Wint, & Ellis, 1990) and can interfere with acquisition of new skills (Koegel & Covert, 1972).

Identification of effective interventions for stereotypy may challenge researchers and clinicians because stereotypy is frequently maintained by automatic reinforcement (Rapp & Vollmer, 2005). Automatic reinforcement refers to operant behavior that occurs independent of social consequences. During a functional analysis, automatic reinforcement is identified when differentially high levels of stereotypy occur during an alone condition (when no consequences are delivered contingent on stereotypy) or when high levels of stereotypy are observed during all conditions. When high and undifferentiated patterns of responding are observed across conditions, an extended series of alone sessions can provide further support for whether the response maintains in the absence of social consequences (Vollmer, Marcus, Ringdahl, & Roane, 1995).

When a behavior is maintained by automatic reinforcement, it may be difficult to identify an effective treatment because the response directly produces the reinforcer. Despite this potential difficulty, effective treatments have been developed for decreasing stereotypy. Reinforcement-based treatment options include non-contingent reinforcement (Lindberg, Iwata, Roscoe, Worsdell, & Hanley, 2003), differential reinforcement of other behavior (Cowdery, Iwata, & Pace, 1990), and differential reinforcement of alternative behavior (Wacker et al., 1990).

Reinforcement-based interventions are not always effective at decreasing automatically reinforced stereotypy (Lindberg, Iwata, & Kahng, 1999). In these cases, punishment-based interventions such as response blocking (Lerman & Iwata, 1996) and response redirection (Ahearn, Clark, MacDonald, & Chung, 2007) may be necessary to effectively reduce stereotypy. Response blocking involves either the prevention of a response (e.g., by placing one's hand between the participant's hand and eye when a participant attempts to eye poke) or physical redirection of the response (e.g., physically removing one's hand from their eye socket). Response blocking decreases several topographies of automatically reinforced problem behavior, including hand mouthing (e.g., Lerman and Iwata, 1996, Smith et al., 1999), eye poking (Lalli, Livezey, & Kates, 1996), and motor stereotypy (Lerman, Kelley, Vorndran, & Van Camp, 2003). When an intermittent schedule of response blocking is implemented, stereotypy increases, suggesting that the decreases in stereotypy observed following response blocking are a function of punishment processes and not extinction (Lerman & Iwata, 1996).

Response redirection involves a caregiver presenting prompts to engage in a response other than the problem behavior contingent on stereotypy. Response interruption and redirection decreases vocal stereotypy (Athens et al., 2008, Cassella et al., 2011, Lui-Gitz and Banda, 2010) and motor stereotypy (Ahrens, Lerman, Kodak, Worsdell, & Keegan, 2011). In an early evaluation of this procedure, Ahearn et al. (2007) evaluated response interruption and redirection (RIRD) for treating vocal stereotypy. Contingent on vocal stereotypy, the participants were asked to respond to simple questions (e.g., “What's your name?”) or perform vocal imitations (e.g., “Say ‘ball.”’) until they responded to three directives in the absence of vocal stereotypy. Response interruption and redirection decreased vocal stereotypy and was associated with increased appropriate vocalizations. Similar to response blocking, stereotypy increases when an intermittent schedule of response redirection is implemented, suggesting that the decreases in stereotypy observed following redirection are due to punishment (Ahrens et al., 2011).

Because there are several treatment options available for decreasing automatically reinforced stereotypy, including reinforcement- and punishment-based interventions, it may not always be clear to clinicians which interventions are most appropriate for their clients. A growing area of research involves investigating participant preference for different treatments (Hanley et al., 2005, Layer et al., 2008, Luczynski and Hanley, 2009). Accounting for a client's preference for treatments may lead to better long-term outcomes, perhaps because the client may be less likely to resist treatment. In addition, providing a choice between treatments may establish a particular treatment as being more reinforcing (Hanley, 2010).

One difficulty associated with identifying preferred treatments for individuals with autism is that those individuals may be unable to vocally communicate their preferences. An alternative to vocally communicating preference is to use a concurrent-chain procedure. When concurrent-chain procedures are used in this way, an individual is provided with a choice between several treatment options, often through the presentation of different colored cards or microswitches. An initial response (such as touching the card or microswitch) results in brief exposure to the treatment contingencies associated with that choice. Preference is determined by evaluating how frequently the participant selects each of the treatment options. Concurrent-chain procedures have been use to identify participant preference for functional communication training (FCT) over noncontingent reinforcement (NCR; Hanley, Piazza, Fisher, Contrucci, & Maglieri, 1997), for contingent over noncontingent social reinforcement (Luczynski & Hanley, 2009), and for FCT with punishment over FCT with extinction (Hanley et al., 2005).

Although the concurrent-chain procedure has identified that children may prefer treatments involving punishment over those without a punishment component, research has not evaluated this procedure for determining preference between two behavior-reduction procedures, such as response blocking and RIRD. The purpose of this experiment was twofold. First, we compared the efficacy of response blocking and RIRD for treating stereotypy. Second, we used a concurrent-chain procedure to determine which treatment participants preferred.