Assessment and treatment of hand mouthing

MISSING IMAGE

Material Information

Title:
Assessment and treatment of hand mouthing
Physical Description:
Unknown
Creator:
Roscoe, Eileen Michelle
Publication Date:

Record Information

Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 27752311
oclc - 48541207
System ID:
AA00020463:00001

Full Text









ASSESSMENT AND TREATMENT OF HAND MOUTHING:
A LARGE-SCALE ANALYSIS


By

EILEEN MICHELLE ROSCOE


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA


2001














ACKNOWLEDGMENTS

I would like to thank those individuals who helped make this dissertation

possible. First, I thank Dr. Brian Iwata, my committee chair and advisor, for his patience

and support throughout my graduate education. His professional assistance and advice

throughout this project and others have been invaluable. I also thank the other committee

members, Dr. Jennifer Asmus, Dr. Timothy Hackenberg, Dr. Mark Lewis, and Dr.

Timothy Vollmer, for their assistance. Additional thanks are extended to Alonzo Cole,

Gerald Goff, and Liming Zhou, who assisted in the development and implementation of

this project. I also thank my family for their unconditional love and support.














TABLE OF CONTENTS

page

ACKN OW LEDGM ENTS................................................................................................. ii

ABSTRACT .................................................................................................................. iv

INTRODUCTION .......................................................................................................... 1

Prevalence................................................................................................................... 2
Risk .............................................................................................................................3
Function...................................................................................................................... 5
Interventions for Behavior Maintained by Automatic Reinforcement......................... 10

EXPERIMENT 1: PREVALENCE AND RISK OF HAND MOUTHING ................... 16

M ethod.................................................................. ................................................. 16
Results and Discussion.............................................................................................. 20

EXPERIMENT 2: FUNCTIONAL ANALYSIS OF HAND MOUTHING................... 23

M ethod......................................................................................................................23
Results and Discussion.............................................................................................. 24

EXPERIM ENT 3: TREATM ENT ................................................................................ 29

Participants and Setting............................................................................................. 30
Data Collection and Reliability.................................................................................. 30
Phase I: Functional Analysis..................................................................................... 30
Phase II: Preference Assessm ents............................................................................. 33
Results and Discussion for Phase II........................................................................... 34
Phase III: Treatm ent Evaluation ............................................................................... 36
Results and Discussion for Phase III.......................................................................... 39

DISCU SSION ............................................................................................................... 50

REFERENCES ............................................................................................................. 54

BIOGRAPHICAL SKETCH ......................................................................................... 59














Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

ASSESSMENT AND TREATMENT OF HAND MOUTHING:

A LARGE-SCALE ANALYSIS

By

Eileen Michelle Roscoe

August, 2001
Chairman: Brian A. Iwata
Major Department: Psychology

Hand mouthing (HM) is a chronic problem in individuals with developmental

disabilities. Some forms of HM are classified as self-injurious behavior when lesions,

tissue breakdown, and infection occur. Although the prevalence of mouthing has been

reported, data on the frequency, severity, or functions of the behavior were not included.

The purpose of the present study was to conduct such an analysis of HM in a large

population. The study was conducted at two public residential facilities for persons with

developmental disabilities, with a total population of 830 individuals.

In Experiment I, the prevalence and risk of HM were assessed using indirect and

direct observation methods. Results obtained from interviews showed that the prevalence

of HM in two institutional samples was comparable to that reported in earlier studies,

whereas direct observation yielded a lower estimate of prevalence. In addition, results

indicated that HM, when it did occur, was likely to produce tissue damage if left

untreated.







In Experiment II, modified functional analyses (FAs) were conducted, during

which alone, attention, and demand sessions were alternated in multielement designs with

a 2:1 ratio of alone to attention and demand sessions. Results indicated that, in 98.4% of

the 64 cases for which an FA was conducted, HM was most likely maintained by

automatic reinforcement.

In Experiment III, 14 individuals whose HM either occurred at very high-rates

during the FA or produced severe tissue damage to the hands participated. In Phase I, a

modified FA was conducted that determined participants' HM was maintained by

automatic reinforcement. In Phase II, preference assessments were conducted that

identified preferred leisure and/or edible items that were used during treatment. In Phase

III, the following interventions were implemented in sequential order: (a) noncontingent

reinforcement (NCR), (b) either NCR + differential reinforcement of alternative behavior

(DRA) + response blocking or NCR + response blocking only, and (c) NCR + brief

manual restraint. NCR, NCR + DRA + response blocking, NCR + response blocking,

and NCR + brief restraint resulted in increases in item contact and large reductions in

HM for six, five, two, and one participantss, respectively.













INTRODUCTION

The purpose of the current investigation was to conduct a large-scale assessment

and treatment of hand mouthing, a form of behavior exhibited by individuals with

developmental disabilities that may or may not be injurious. In Experiment 1, the

prevalence, frequency, and risk of HM were assessed using indirect and/or direct

observational methods. In Experiment 2, a modified functional analysis, based on that

described by Iwata et al. (1982/1994a) was conducted, to identify the extent to which

participants' HM was sensitive to (maintained by) social contingencies. In Experiment 3,

a progressive series of interventions was evaluated to determine which treatment, or

combination of treatments, most effectively reduced HM.

Hand mouthing (HM) has been defined as the placing of one or more digits of the

hand past the plane of the lips (Lockwood & Williams, 1994), insertion of the hand into

the mouth beyond the first knuckle (Dorsey, Iwata, Ong, & McSween, 1980), and/or any

contact between the hand and the mouth (Lerman & Iwata, 1996). Rast and Jack (1992)

made a distinction between hand mouthing and the mouthing of objects, noting that the

former involves contact of only the hands and the mouth, whereas the latter may include

several behaviors involving contact of the mouth with other objects. HM is

topographically very similar to hand biting, often making these responses

indistinguishable to observers. For example, Paisey, Whitney, and Wainczak (1993)

conducted a case study of a participant's self-injurious HM. Although the authors noted

that the participant frequently engaged in hand biting, they only recorded instances of








HM (defined as insertion of fingers or hand into mouth). However, based on their

definition, occurrences of HM may have included instances of biting. Assuming that HM

and hand biting may be functionally equivalent (i.e., they are both maintained by the

same environmental variable), lumping these responses together may not prove

problematic. However, if the responses are maintained by different consequences, clear

assessment and/or treatment outcomes may be greatly impeded. Given this, care should

be taken to use definitions that allow for accurate discrimination between hand mouthing

and hand biting for participants who exhibit both response forms.

HM that occurs at a high frequency and/or duration may result in injury.

Researchers have reported salivary dermatitis and/or associated infections requiring

medical treatment as a result of participants' HM (Ball, Campbell, & Barkemeyer, 1980;

Lockwood & Williams, 1994; McClure, Moss, McPeters, & Kirkpatrick, 1986). HM has

also been classified as a form of "stereotypy" because it is repetitive, has no apparent

function, and may interfere with adaptive behavior (LaGrow & Repp, 1984). Regardless

of whether a participant's HM is injurious, it may greatly impede her/his quality of life.

Most notably, the offensive odor associated with HM may serve as a barrier to social

interactions (McClure et al., 1986). In addition, mouthing may interfere with the

acquisition of appropriate behaviors (Ball et al., 1980).

Prevalence

Because HM has been classified in several different ways, estimating its

prevalence has been difficult. That is, some authors have described HM as a self-

injurious response and have included it in studies on the prevalence of self injury (Griffin

et al., 1986; Maurice & Trudel, 1982), whereas others have considered it a form of

stereotypy (Bodfish et al., 1995; LaGrow & Repp, 1984). Given that studies on the








prevalence of self-injurious behavior (SIB) typically exclude noninjurious response

topographies, and studies on the prevalence of noninjurious stereotypy exclude self-

injurious topographies, reports on the prevalence of HM are usually incomplete.

The only study to date that has assessed the prevalence of HM specifically was

conducted by Troster, Bambring, and Beelmann (1991), who surveyed parents of children

(10 mo.-to-6 yr.) who were blind. Parents were asked to indicate which of 11 different

stereotyped behaviors (e.g., sucking thumbs or fingers, body rocking, repetitive hand and

finger movements, etc.) they had observed in their child; they were also asked to indicate

the frequency of response topographies reported to occur. Results indicated that the

prevalence of HM that occurred (a) at least once per week was 49.4%, (b) once per day

was 36.5%, and (c) hourly was 4.7%. Although this study was the first to include HM as

a specific response topography, it was conducted only with normally developing blind

children; thus, the generality of these findings to adults or to individuals with

developmental disabilities is unclear. In addition, the authors used only indirect (survey)

methods for assessing the prevalence of SIB, stereotypy, and/or HM, which may have

been unreliable or inaccurate. In the present study, we assessed the prevalence of HM by

conducting a series of assessments, beginning with indirect assessments (i.e., interviews)

and ending with direct observation to verify the results obtained during the indirect

assessment. By conducting both indirect and direct assessments of HM, more accurate

measures may be obtained with respect to prevalence.


Risk
As noted previously, HM has been reported to result in severe self-injury (Ball et

al., 1980; Lockwood & Williams, 1994; McClure et al., 1986). Rast and Jack (1992)








noted that, for HM to be considered injurious, mouthing should result in "red or white

hands, chapped lips or face, or lesions on the hands or face" (p. 1). In addition, they

indicated that HM, when injurious, occurs both frequently and for long periods of time.

Based on their definition of injury, an adequate assessment of risk should include both

the frequency and severity of HM.

Few studies have assessed the frequency and/or duration of stereotypic behaviors.

Dura, Mulick, and Rasnake (1987) assessed the frequency and duration of various forms

of stereotypy, including rhythmic movements, bizarre posturing, self-restraint, and object

manipulation. Participants' direct care staff were interviewed to determine whether or

not an individual (a) engaged in each type of stereotypy and (b) engaged in the

stereotypic response for 1 min, 1-to-10 min, or greater than 10 min. Although assessing

the duration of HM may prove useful, it was unclear whether the duration reported was

per episode, per day, or per week. If the duration was per episode, it would have been

helpful to know how many episodes occurred per day. Rojahn (1986) determined the risk

of participants' stereotypy by conducting a survey, consisting of a 3-point scale for

severity and a 4-point scale for the frequency of occurrence. Schroeder, Schroeder,

Smith, and Dalldorf (1978) used an indirect assessment to determine the severity of

participants' SIB. After social workers referred individuals who exhibited SIB, a

pediatrician or psychologist interviewed the social worker to assess whether the SIB was

severe (it occurred at least once a day and had caused bleeding, bruises, broken bones, or

other tissue damage that required intervention by a medical staff) or mild (those referrals

that were not severe). The studies noted above are noteworthy in that they assessed the

frequency and/or severity of several different topographies of SIB and/or stereotypy.








However, because only indirect assessments were used, the reliability of the findings is

uncertain. Second, none of the studies reported data specifically on HM, making it

unclear if HM often results in injury. Although authors have noted that HM may result in

severe tissue damage, no studies to date have conducted a systematic assessment

indicating the degree of injury associated with HM. In the present study, we sought to

determine the risk of HM exclusively by conducting direct observations of the frequency

and severity of participants' HM.

Function

The development of functional analysis methodology is based on over 30 years of

research indicating that behavior problems (e.g., aggression, noncompliance, and SIB)

are often learned performances. That is, they are maintained by contingencies of positive

or negative reinforcement. Iwata, Dorsey, Slifer, Bauman, and Richman (1982/1994a)

proposed a general model for assessing behavioral sensitivity to a variety of

contingencies. In this study, three test conditions were used, each containing a

discriminative stimulus (SD), an establishing operation (EO), and a source of

reinforcement. A fourth condition, the "play" condition, was designed to be a control. In

the social disapproval or attention condition, toys were continuously available, and the

therapist sat in a chair across the room reading a book. Brief attention was delivered

contingent upon the target behavior, and all other responses were ignored. This condition

served as a test for behavior maintained by social positive reinforcement. In the demand

condition, a therapist presented instructional tasks, and problem behavior resulted in

escape from these tasks. This condition tested for SIB maintained by social-negative

reinforcement. In alone condition, the participant was alone in a room, and no materials

were present. Persistence of responding in this condition suggested that problem








behavior was maintained by automatic reinforcement (Vaughn & Michael, 1982).

During the play or control condition, leisure items were freely available, attention was

delivered on a fixed-time (FT) schedule, and no demands were presented. Because both

positive and negative reinforcers were noncontingently available during this condition,

the probability of behavior occurring should be greatly reduced. Results showed that 4 of

the 9 participants exhibited the highest levels of behavior in the alone condition,

suggesting their self-injury was maintained by automatic reinforcement, and 3

participants exhibited undifferentiated responding or responding that was high across all

conditions, suggesting their behavior may also have been maintained by automatic

reinforcement. Two participants exhibited the highest levels of SIB during demand

sessions, suggesting their behavior was maintained by social-negative reinforcement, and

only 1 participant exhibited differentially highest levels of self-injury during attention

sessions, suggesting his behavior was maintained by access to attention.

Although functional analysis provides information for effective treatment

development, few researchers have conducted functional analyses prior to treatment

selection for HM (e.g., Mazaleski, Iwata, Rodgers, & Zarcone, 1994; Vollmer, Marcus, &

LeBlanc, 1994; Lerman & Iwata, 1996). In addition, out of the studies that have included

a functional analysis prior to treatment, small sample sizes were used, limiting the

generality of their findings. The study with the largest sample size to date was conducted

by Goh et al. (1995), who conducted functional analyses of 12 individuals who exhibited

HM. Results indicated that in 10 of the 12 participants' (83%) HM was either (a) higher

in the alone condition relative to the other test conditions or (b) not differentially high

during either the attention or demand conditions. Thus, in a large majority of the cases,








HM persisted in the alone condition and was insensitive to social contingencies,

indicating the behavior was maintained by automatic reinforcement. However, whether

these data reflect the proportion of HM that is maintained by automatic reinforcement in

the larger population of individuals who engage in the behavior remains unclear.

Although functional analysis methodology has been used by many researchers to

identify the maintaining variables of a wide range of behavior disorders, several authors

have noted that this assessment may be too time consuming (Durand & Crimmins, 1988;

Repp, Singh, Olinger, & Olson, 1990). For example, in a large-scale study based on 152

functional analyses, Iwata et al. (1994b) reported that the duration of assessment ranged

from 8 to 66 sessions (2 hr to 16.5 hr). In the study noted above by Goh et al. (1995),

functional analyses averaged 25 sessions (6 hours).

A number of procedural variations have been designed to increase the efficiency

with which functional analyses can be conducted. One such variation is the "brief

functional analysis" (Cooper et al. 1992; Derby et al., 1992; Northup et al., 1991; Wacker

et al., 1990). Derby et al. assessed the effectiveness of the brief functional analysis for 79

participants. Each participant was evaluated during a 90-min period, and results showed

that the target behavior was observed and clear differentiation in the data patterns

emerged for 46.6% of the participants. In a more direct evaluation of the brief functional

analysis, Kahng and Iwata (1999) compared the outcomes from 50 functional analyses

consisting of brief (single) and extended exposures to test and control conditions and

found that interpretations from the brief and extended analyses matched for 66% of the

individuals. The development of the brief functional analysis has been a significant

contribution to the field of behavior analysis by offering an efficient and objective








assessment method for use under time constraints. However, some limitations of the

brief functional analysis are that single exposures to test conditions may be insufficient to

occasion behavior or to establish discrimination among contingencies associated with the

different conditions.

Another method for increasing the efficiency of functional analyses is to shorten

session length. This procedural variation would decrease the overall length of assessment

while permitting repeated exposures to each test condition. Wallace and Iwata (1999)

compared the results obtained from functional analyses based on varying session

durations (i.e., 15 min, 10 min, and 5 min). Functional analyses involving attention,

demand, alone, tangible, and play conditions were conducted in multielement designs.

Forty-six functional analyses based on 15-min sessions were used to generate shorter data

sets by subtracting the last 5 min (for the 10-min duration) and the last 10 min (for the 5-

min duration) from each session, allowing for comparisons among 15-, 10-, and 5-min

session durations. Evaluators who were blind to participant identity and session duration

visually inspected the graphs to determine whether the results indicated attention,

tangible, escape, and/or automatic functions. The authors then determined the extent to

which the interpretations based on each 15-min data set matched those based on 10- or 5-

min data sets. Results indicated perfect agreement for comparisons between the 15- and

10-min session durations and only three disagreements for comparisons between 15- and

5-min session durations. The mean durations of the 15-, 10-, and 5-min functional

analyses would have been 6.1 hr, 4.1 hr, and 2.1 hr, respectively. Based on these

findings, 5-min sessions often may be sufficient to reveal functional relationships and








may greatly reduce the amount of time required to conduct assessments involving

repeated exposure to test conditions.

Another method for increasing the efficiency of functional analyses involves

progressing from a brief functional analysis to a more extended analysis if needed.

Vollmer, Marcus, Ringdahl, and Roane (1995) developed a model for conducting

functional analyses using conditions that were based on those described by Iwata et al.

(1982/1994a). In this model, a brief functional analysis was first conducted (Phase I); if

differentiated results were obtained, the assessment was complete. If differentiated

results were not obtained, a more extended multielement format was used (Phase II). If

differentiated results were observed in the more extended analysis, the assessment was

complete. If undifferentiated results occurred, an extended no-interaction condition was

conducted to determine whether behavior persisted in the absence of social contingencies

(Phase III). If behavior occurred at high levels during this condition, the assessment was

complete; if behavior extinguished during the extended no-interaction condition,

indicating that behavior may be sensitive to social reinforcement, the functional analysis

conditions were presented in a reversal design (Phase IV). The function of participants'

problem behavior was identified for 17 of the 20 participants as a result of this

methodology.

The development and continued research on methods for increasing the efficiency

of functional analysis methods is important because it provides clinicians working under

time constraints with a more objective alternative to indirect methods. In the present

study, we assessed the utility of an abbreviated functional analysis methodology.

Because previous research indicates that HM is often maintained by automatic








reinforcement, we conducted a series of alone sessions that were interspersed with

attention and demand probes. If behavior persisted during the alone condition, with

lower levels in the attention and demand conditions, this would verify our hypothesis that

the participant's HIM was maintained by automatic reinforcement. If responding did not

persist during the alone condition, then a more extended functional analysis would be

conducted until a clear function could be identified.

Interventions for Stereotypy and/or Behavior

Maintained by Automatic Reinforcement

If results of a functional analysis verify that HM is maintained by automatic

reinforcement, procedures that have been found effective in reducing problem behavior

maintained by automatic reinforcement comprise the treatments of choice. One such

procedure is noncontingent reinforcement (NCR). For example, Goh et al. (1995)

presented leisure items to participants who exhibited HM maintained by automatic

reinforcement. Results indicated that NCR resulted in reduced levels of hand mouthing

and high levels of item manipulation for 3 of the 4 participants. These findings suggested

that stimulation from the leisure items might have substituted for the stimulation from

participants' HM. Shore et al. (1997) demonstrated the effectiveness of NCR in three

individuals who exhibited SIB in the form of arm rubbing (1 participant) or hand

mouthing (2 participants). Functional analyses indicated that SIB occurred most often in

the absence of social contingencies (i.e., the alone condition), suggesting participants'

SIB was maintained by automatic reinforcement. During baseline, when no interaction or

materials were delivered, SIB occurred at moderate-to-high levels; during NCR, when

preferred stimuli (based on a stimulus preference assessment) were continuously

available, SIB was greatly reduced, and item contact occurred at high levels. These








findings also suggested that the stimulation from the leisure items may have competed

with that produced by participants' SIB. A noteworthy feature of both the Goh et al. and

Shore et al. studies is that the authors conducted a functional analysis prior to evaluating

treatment. It is important to empirically demonstrate whether problem behavior is

maintained by automatic reinforcement. Without appropriately ruling out potential social

reinforcers that may be maintaining responding, treatment procedures may have limited

effectiveness.

In other studies, supplementary procedures, such as differential reinforcement of

alternative behavior (DRA) and/or prompting, have been used to decrease stereotypy and

increase object manipulation. Favell, McGimsey, and Schell (1982) found that DRA

effectively reduced participants' problem behaviors (HM, eye poking, and/or pica) and

increased appropriate item manipulation. After NCR was found ineffective in

suppressing HM to low levels and increasing adaptive behaviors (e.g., appropriate social

interaction and item interaction) to high levels, Homer (1980) assessed the effectiveness

of DRA and observed further decreases in SIB and increases in adaptive behaviors.

Singh and Millichamp (1987) evaluated the effectiveness of prompting to teach

individuals with profound mental retardation to engage in appropriate play and social

behaviors. Results indicated that stereotypy and inappropriate play behaviors decreased

to low levels when appropriate behaviors increased.

When reinforcement-based interventions, such as NCR and DRA, do not result in

concomitant decreases in problem behaviors, interventions aimed at directly reducing

problem behaviors, such as response blocking and manual restraint, have been used.

Lerman and Iwata (1996) assessed the effects of response blocking on one participant's








HM that was maintained by automatic reinforcement. Baseline and response blocking

sessions were conducted in a reversal design. Baseline sessions were identical to the

alone condition of the functional analysis in that no interaction was delivered. During

response blocking sessions, the therapist manually blocked some or all attempts at HM.

Results indicated that when HM was blocked, HM attempts decreased to near-zero levels.

Lindberg, Iwata, and Kahng (1999) assessed the effectiveness of response blocking with

two individuals who exhibited several different topographies of SIB (head or body

hitting, rubbing, finger pressing against face). Results of a functional analysis indicated

that their behavior was not maintained by social reinforcement. The authors

implemented a progressive series of treatment components (NCR, NCR with prompting,

NCR and DRA with prompting, NCR and DRA with prompting and response blocking)

in a multiple baseline design. During NCR, when preferred leisure items were

continuously available, moderate-to-high levels of SIB occurred, and little or no item

contact occurred. Similar findings were obtained during the NCR with prompting

condition. During NCR and DRA with prompting, when food items were delivered

contingent upon specified durations of independent item contact, SIB continued to occur

at high levels, and item contact did not increase. When the response-blocking component

was added to the other treatments, item contact did not increase; however, HM did

decrease for one of the two participants. The studies noted above lend some support for

the efficacy of response blocking in reducing HM and other forms of stereotypy.

Turner, Realon, Irvin, and Robinson (1996) assessed the effectiveness of brief

manual restraint in three individuals with profound mental retardation. Results of a

functional analysis indicated that participants' HM was not maintained by social








reinforcement. NCR and NCR with restraint sessions were conducted in a reversal

design. During NCR, when leisure items were continuously available, HM occurred at

high levels, and item contact occurred at low levels; during NCR with restraint, when the

therapist placed the participant's hand in her/his lap for 3 s following each HM attempt,

HM decreased to low levels, and item contact increased to high levels. Thompson, Iwata,

Conners, and Roscoe (1999) demonstrated the importance of including preferred items

when using manual restraint procedures. After conducting a functional analysis of

participants' SIB (hand mouthing, head and body hitting) to rule out behavioral

maintenance by social contingencies, an assessment was conducted to identify

participants' preferred leisure items. Next, the effects of brief manual restraint were

evaluated in reversal or multiple baseline designs. During baseline and manual restraint

conditions, participants were also exposed to reinforcement and no-reinforcement

conditions, which were alternated in a multielement design. Baseline sessions were

identical to the no-interaction condition (the therapist did not use manual restraint during

this condition) of the functional analysis. During manual restraint sessions, the therapist

held a participant's hands in her/his lap or across her/his chest for 15 s contingent upon

SIB attempts. During no-reinforcement sessions, the participant did not have access to

leisure items; during reinforcement sessions, participants had continuous access to their

top preferred leisure item identified in the preference assessment. Results indicated that

although manual restraint with no reinforcement resulted in decreases of participants'

problem behavior, manual restraint was more effective when combined with

reinforcement. These findings lend support for the use of manual restraint when

reinforcement-based interventions prove ineffective. In addition, the findings of








Thompson et al. demonstrate the importance of including preferred leisure items when

using more intrusive interventions, such as manual restraint.

In summary, a number of treatment procedures, ranging from benign interventions

such as NCR to more intensive ones such as restraint, have been found effective in

suppressing problem behavior maintained by automatic reinforcement. However, there

are some limitations of research on the treatment of HM. First, few of these studies have

focused on HM specifically, making it unclear whether procedures found effective with

other response topographies would also prove effective for HM. Second, treatment

procedures have often been implemented without first conducting a functional analysis.

Although research indicates that stereotypy often persists in the absence of social

contingencies (Lindberg et al., 1999), it is important to conduct functional analyses to

rule out a potential social function. Third, researchers often did not use systematic

preference assessments prior to implementing NCR and DRA treatment components.

Because the reinforcer for behavior maintained by automatic reinforcement is unknown,

it is important to identify potent reinforcers that may effectively compete with the

sensory stimulation the problem behavior produces. Finally, researchers have often used

intrusive interventions, such as restraint and protective equipment, without first assessing

the effectiveness of reinforcement-based interventions. The present study attempted to

extend current research on the treatment of HM and behavior maintained by automatic

reinforcement in the following ways: (a) A functional analysis was conducted to rule out

behavioral sensitivity to social contingencies, (b) systematic preference assessments were

conducted to identify preferred leisure materials that may compete with HM and to

identify edible items that may reinforce appropriate play behaviors, (c) a progressive





15


series of interventions was used, beginning with NCR and gradually introducing more

intrusive components such as response blocking and brief manual restrain on an as

needed basis.














STUDY 1: PREVALENCE AND RISK OF HAND MOUTHING

Method

Participants

All individuals living in two state residential facilities for persons with

developmental disabilities, one in Florida (N=501) and one in Tennessee (N=329),

participated in the initial screening (see below). Participants' ages ranged from 16 to 55

years, their levels of mental retardation ranged from mild to profound, and their verbal

skills were extremely variable but generally very limited. All 830 or some subset of these

individuals participated in the procedures noted below.

Response Definition

During all assessment procedures, HM was defined as placement of the hand past

the plane of the lips or as repetitive contact between the hand and the mouth. In addition,

staff members were told that participants' HM should be repetitive and chronic to be

included (to exclude socially acceptable forms of HM, such as when eating).

Indirect Assessment

During the initial screening, Behavior analysts and/or direct care staff from each

home at both residential facilities were contacted through email or telephone and were

asked to identify individuals in their home who exhibited HM. At the institution in

Florida, staff members were given a list of names and asked to indicate whether or not

each participant exhibited HM. At the institution in Florida, Behavior analysts who

worked closely with the clients were asked to list the names of individuals who exhibit








HM. As a result of this initial screening, 130 participants were reported to engage in

HM. These 130 individuals participated (see Table 1 for their demographic information)

in all subsequent assessments unless otherwise noted. The proportion of the participant

Table 1
Demographic Characteristics of Participant Sample (N=130)

Number Percentage
Sex
Male 75 57.7%
Female 55 42.3%


Age
1-10
11-20
21-30
31-40
41-50
51+

Degree of retardation
Mild/moderate
severe
profound

Impairments
visual
auditory
physical

Receptive verbal skills
none
minimal
follows directions

Expressive verbal skills
none
minimal
one-word utterances
Two-to-ten-word utterances
>10 word utterances


4
2
124


0.0%
2.3%
16.2%
28.5%
37.7%
15.4%


3.1%
1.5%
95.4%


33.1%
29.2%
36.2%


11.5%
53.1%
35.4%


79.2%
3.1%
6.9%
6.2%
S60o/.








sample classified as having profound mental retardation was very high (95.4%) and not

representative of the total population of 830.

Following the initial screening, more extensive interviews were conducted to

gather information about the 130 individuals identified. A graduate student met with a

staff member who worked with the participant in an office or a quiet room and asked the

staff member specifically whether or not a participant exhibited HM. If HM was reported

to occur, the interviewer also asked staff to estimate whether HM occurred at a low rate

(less than once per hour) or at a higher rate (more than once per hour). Interobserver

agreement was not assessed for these indirect assessments.

Direct Assessment

Direct observations were conducted for two purposes. First, we wanted to verify

results of the interviews by observing whether a participant reported to engage in HM

actually did so. Second, we wanted to determine whether participants engaged in high or

low rates of HM. Because HM that occurs at higher frequencies may be more likely to

result in injury, this measure provided information related to the risk of participants' HM.

One hundred and two of the 130 individuals reported to engage in HM during the

indirect assessment participated. We were unable to obtain consent for 28 of the

individuals. Sessions were conducted in therapy rooms located in a clinic for the

assessment and treatment of SIB, or were conducted in quiet rooms in participants'

homes. If participants wore protective equipment, the equipment was removed prior to

conducting observations.

Trained observers recorded instances of HM using either handheld computers

(Assistant model AST 102) or a data sheet segmented into 10-s bins. Sessions were

either 5 or 10 min, and the total observation was 60 min. Data were summarized as the








percentage of 10-s intervals during which HM occurred. Reliability was assessed by

having a second observer collect data independently during an average of 34.2% of the

sessions. Observers' records were compared on an interval-by-interval basis, and

percentage agreement was calculated by dividing the number of agreements by the

number of agreements plus disagreements and multiplying by 100%. The mean percent

agreement was 96% (range, 91.1% to 97.7%).

The majority of sessions resembled the alone condition of the functional analysis

(i.e., individuals had no materials, and no interaction was provided). Also, to rule out the

possibility that HM may not occur during an alone condition as a result of behavioral

maintenance by social contingencies, at least one attention and at least one demand

condition were conducted (see below, functional analysis). By directly observing

participants under these conditions, we increased the likelihood that low rates or an

absence of responding during no interaction conditions were not the result of extinction.

Information regarding whether or not HM occurred during the direct assessment was used

to determine the accuracy of the indirect assessment. Information regarding the amount

of HM that occurred during the assessment provided information concerning the risk of

HM. For example, individuals who exhibited HM during less than 20% of the intervals

were classified as having low-rate HM, whereas individuals who exhibited HM during

20% or more of the intervals were classified as having high-rate HM.

Severity

The 130 participants reported to engage in HM during the interview participated.

The settings were identical to those used for the direct assessment.

During the severity assessment, either a behavior analyst or a graduate student

examined a participant's hands and rated the extent of injury on a 4-point scale according








to the following criteria: (a) A "1" was scored if no cuts, bruises, or swelling were

observed; (b) a "2" was scored if reddened areas, but no cuts, bruises, or swelling were

present; (c) a "3" was scored if there was current injury, such as scabs, swelling, or

cracks; and (d) a "4" was scored if the participant wore protective equipment prescribed

to prevent HM. HM was classified as injurious if a score of a 3 or a 4 was recorded.

A second observer collected data independently on an average of 54.6% of these

assessments. In comparing observers' records, an agreement was defined as both

observers having recorded the same score for a given participant. Interobserver

agreement was calculated by dividing agreements by agreements plus disagreements and

multiplying by 100%. The mean percentage agreement across participants was 91.5%

(range, 0% to 100%).

Results and Discussion

Table 2 shows the results from the indirect assessment. Out of the entire

population (830 individuals), 130 (15.7%) were reported by staff to engage in HM, and

59 (7.1%) were reported to engage in high-rate HM (greater than once per hour).

Although the reported prevalence of HM was much less than that reported by Troster et

al. (1991), instances of high rate HM were comparable to those reported by Troster et al.

Out of the 130 individuals reported to engage in HM, direct observations were

conducted for 102. To estimate the number of individuals who exhibited HM out of the

larger sample size (n = 130), we first determined the proportion of individuals who were

observed to exhibit HM out of the smaller sample (n = 102). Sixty-four out of the 102 or

62.7% were observed to engage in HM. Thus, the estimated number of individuals who

exhibited HM out of the larger sample of 130 was 62.7% or 82. Out of the total

population of 830 individuals, the estimated number that engaged in HM was 82 or 9.9%.








We used the same method to estimate the number of individuals who engaged in high-

rate HM (50 or 6% of the total population of 830). The percentage of high rate HM,

given that HM was observed, was 61%. Thus, results obtained from direct observation

yielded lower figures than those obtained from the indirect assessment for overall

prevalence as well as for prevalence of high-rate HM. These findings also suggest that

the majority of individuals who exhibit HM engage in this behavior at a high rate.

Table 2
Experiment I: Prevalence, Frequency, and Severity Results

Number Proportion
Indirect Assessment (N=830)
Overall prevalence 130/830 15.7%
High rate HM 59/830 7.1%

Direct Observation (N=130*)
Overall prevalence 82/830 9.9%
High rate HM 50/830 6.0%
High rate HM/AI1 observed HM 50/82 61.0%

Severity Ratings (N=130)
Overall prevalence of severe HM 29/830 3.5%
Severe HM/All observed HM 29/82 35.4%
Numbers reflect estimates for the 130 individuals reported to engage in HM based on
direct observation of 102 of these individuals.

All 130 individuals reported to engage in HM received a severity measure through

direct observation. Twenty-nine individuals or 3.5% of the total population (830) were

observed to have severe HM (i.e., a 3 or a 4 severity score). The proportion of

individuals who engaged in severe HM (HM that produced observable tissue damage or

required protective equipment to prevent such damage), given that HM was observed,

was 35.4%.

A strength of this investigation was that both interview and direct observation

methods were used to determine the prevalence of HM. Results of a recent study by








McGill, Hughes, Teer, and Rye (2001) indicate a high degree of variability obtained

through indirect assessment methods. After conducting interviews with staff members

regarding the occurrence and frequency of various topographies of problem behavior, the

authors found large differences in the frequency of behaviors reported by staff members.

It is interesting to note that larger differences were obtained for stereotypic behavior and

for higher-frequency problem behavior. Given that HM may be a form of stereotypy

and/or a high-frequency topography, direct observations seems particularly warranted.

A weakness of the present experiment was that we did not check false negatives.

That is, it is possible that a greater number of individuals than those actually reported

exhibited HM. Future research could assess false negatives by conducting direct

observations with some proportion of the individuals not reported to engage in HM to

verify the absence of HM.














EXPERIMENT 2: FUNCTIONAL ANALYSIS OF HAND MOUTHING

Method

Participants and Setting

All individuals who were directly observed to engage in HM (N=64) during

Experiment 1 participated in Experiment 2. Observations were conducted in a quiet room

in the individual's home or at a day program for the assessment and treatment of their

problem behaviors.

Data Collection and Reliability

HM was defined as either contact between the hand/fingers and the mouth or

insertion of the fingers past the plane of the lips. Trained observers recorded the presence

or absence of HM during continuous 10-s intervals, using a pencil and scoring sheet or a

hand-held computer (Assistant model AST 102). Data were summarized as the

percentage of intervals during which HM occurred. Interobserver agreement was

assessed during a mean of 32% of the sessions. Percentage agreement scores were

calculated based on an interval-by interval comparison of the observers' records by

dividing the number of agreements by the total number of intervals and multiplying by

100%. Mean agreement was 95.6% (range, 90.1% to 98.9%).

Functional Analysis

Procedures based on those described by Iwata et al. (1982/1994a) were conducted

to identify the functional properties of HM and were modified in several ways to increase

the efficiency of assessment. First, in light of data reported by Wallace and Iwata (1999)








indicating that results from 5- or 10-min sessions generally correspond to those obtained

from 15-min sessions, all functional analysis sessions were either 5- or 10-min in

duration. In addition, based on the Goh et al. (1995) findings, which suggested that HM

is unlikely to be maintained by social contingencies, the functional analysis was designed

primarily to determine whether HM persisted in the absence of social consequences,

while still providing some exposure to social contingencies. More specifically,

participants were initially exposed only to attention, demand, and alone sessions.

Conditions were alternated in a multielement design with a 2:1 ratio of alone to

attention/demand sessions. If results were clear after 14 sessions (10 alone, two attention,

and two demand sessions), assessment was terminated. However, if results were unclear,

assessment was extended until a clear function emerged. The specific conditions

extended depended upon the pattern of responding obtained during the initial 14-session

assessment. For example, if results were undifferentiated, with slightly lower levels of

HM in the alone condition relative to the demand condition, alone and demand conditions

were extended and alternated in a 1:1 ratio. However, if results were undifferentiated,

with slightly lower levels of HM in the alone condition relative to attention and demand

conditions, then an extended multielement functional analysis (including alone, attention,

play, and demand conditions) was conducted.

Results and Discussion

Figure 1 shows results of the functional analysis for 3 participants. Each

participant represents a pattern of responding typically observed. In the first pattern, HM

occurred at high levels only in the alone condition, with much lower levels occurring

during the attention and demand conditions (top panel). Out of 64 individuals, 17 or

26.6% exhibited responding characteristic of this pattern. In the second pattern, a high










Pattern I 26.6%)

S" / ^


A-


Pattern 2 (45.3%)

80-1 A-.
60-
40-I


Il-I


--- Alone

---- Attention

P- lay

-----Demand


-5-*- -*


-A


100 Pattern 3(2%)

80
60-
40-
20
n--^ _.i-____________________0 ----


100-

80-
60-


u-I


A


*~'

A-


SESSIONS


Figure 1: Percentage of intervals hand mouthing during functional analysis sessions.
The top graph depicts pattern 1, the second from top graph depicts pattern 2,
the second from the bottom graph depicts pattern 3, and the bottom graph
depicts pattern 4. See text for details








level of responding occurred across conditions, with slightly lower levels in the attention

and demand conditions (second panel). Out of 64 participants, 29 or 45.3% exhibited

responding characteristic of this pattern. In the third pattern, HM occurred during all

conditions but was not differentially high during any condition (third panel). Out of 64

participants, 17 or 26.6% exhibited responding characteristic of this pattern. As noted

previously, when this pattern was observed, either a series of extended alone sessions or

1:1 alternating sessions continued until a clear and differentiated pattern emerged.

Results for all individuals who received an extended analysis indicated that their HM was

either higher in the alone condition or not differentially high during attention and demand

conditions. A fourth pattern of responding was observed in one participant, in which HM

occurred at high levels during the attention condition, with much lower levels occurring

during the alone and demand conditions (see bottom panel of Figure 1), indicating that

the behavior was maintained by social positive reinforcement in the form of attention. In

summary, 63 of 64 or 98.4% of the functional analyses revealed patterns indicating

participants' HM was maintained by automatic reinforcement.

In Experiment 2, we used an abbreviated functional analysis design whose

modifications were based on previous research. We hypothesized that the participants'

HM was most likely maintained by automatic reinforcement, and, as a result, conducted

alone sessions primarily, which were interspersed with attention and demand sessions.

This design proved useful in identifying a behavioral function for all participants in a

relatively brief amount of time.

The functional analysis method used in this study extends previous research on

brief functional analyses. Vollmer, Marcus, Ringdahl, and Roane (1995) assessed the








effectiveness of a brief functional analysis in which 2-to-3 sessions were conducted for

each condition, and within-session response patterns were analyzed for each of the

sessions to determine function. Assessments took 1 to 2 hr, and 30% of the assessments

yielded differentiated results during this brief analysis. The present study differs from

that used by Vollmer et al. in that we used repeated measures of responding in all

conditions with proportionally more alone sessions relative to attention and demand

sessions. However, the total duration of the 14-session assessment used in the present

study was comparable to that reported by Vollmer et al. (1 hr 10 min when sessions were

5-min; 2 hr 20 min when sessions were 10-min). In addition, the 14-session assessment

yielded differentiated results for 47 of the 64 (or 73.4%).

Hanley, Iwata, and Thompson (2001) used another procedural variation when

conducting functional analyses. In their analysis, two conditions were compared: A test

condition (either attention or tangible), in which a suspected reinforcer was delivered

contingent upon problem behavior, and a control condition, in which the same reinforcer

was available noncontingently. There are several noteworthy features of the authors'

procedural variation of the typical functional analysis method. First, the method was

very efficient (taking under 2 hours to implement). Second, the assessment allowed for

repeated measures of the test condition assessed. Third, the specific test condition used

was based on reports from staff and results from structured observations. Fourth, all

participants were successfully treated by interventions based on the results of their

functional analyses. Our method extended the Hanley et al. method by (a) testing for

automatic reinforcement primarily and (b) by conducting probes of all test conditions to

rule out behavioral maintenance by social reinforcement.





28


A potential limitation of our method was our selective use of test conditions.

Including a relatively small number of attention and demand sessions may have limited

the sensitivity of the analysis in detecting social reinforcement functions. Future research

could compare results obtained through the modified design used in this study with

results obtained by the more extended design used by Iwata et al. (1982/1994a).














EXPERIMENT 3: TREATMENT

Data from Experiment 2 indicated that the HM is most often maintained by

automatic reinforcement. Based on this information, treatments that have been found

effective for behavior maintained by automatic reinforcement are perhaps most

appropriate for HM.

Various interventions for behavior maintained by automatic reinforcement have

been applied to HM. For example, NCR, DRA, response blocking, and manual restraint

have been shown to effectively suppress HM. However, few studies have examined

these procedures in the context of functional analyses and systematic preference

assessments. In addition, the literature on treatment for HM contains mostly small N

studies and few comparative analyses of various treatment components. Furthermore, no

study to date has conducted a progressive series of interventions on a large number of

participants exhibiting HM.

The purpose of this experiment was to evaluate and compare a series of

interventions for HM maintained by automatic reinforcement and to implement the

interventions in a manner progressing from least to most intrusive. To this end, three

different phases were conducted during Experiment 3. First, functional analyses were

conducted for all participants to rule out behavioral sensitivity to social reinforcement.

Second, systematic preference assessments were conducted to identify preferred leisure

items that may compete with HM and preferred food items that may be used as

reinforcers to increase item manipulation. Third, we sought to evaluate the utility of a








progressive series of interventions, focusing on the least restrictive procedure necessary

for effectively reducing HM and increasing item contact.

Participants and Setting

Out of the 29 individuals whose HM was observed to be either severe or severe

and high-rate during Experiment 1, 14 participated in the treatment study. These 14

participants were selected because they specifically had been referred for assessment and

treatment, making them high priority cases, or because they did not have scheduled

activities that conflicted with sessions.

Data Collection and Reliability

As in Experiments 1 and 2, HM was defined as either placement of the hand past

the plane of the lips or as repetitive contact between the hand and the mouth. Trained

observers recorded instances of HM using either handheld computers (Assistant model

AST 102) or a data sheet segmented into 10-s bins. Data was summarized as the

percentage of intervals during which HM occurred. Reliability was calculated by having

a second observer collect data independently on an average of 31.2% of the sessions.

Observers' records were compared on an interval-by-interval basis, and percentage

agreement was calculated by dividing the number of agreements by the number of

agreements plus disagreements and multiplying by 100%. The mean percent agreement

for HM was 95.3% (range, 87.4% to 100%).

Phase I: Functional Analysis

Because all participants received a functional analysis in the previous study, we

did not conduct another functional analysis. Functional analyses from experiment 2 for

individuals participating in experiment 3 are shown in Figure 2. Deanna, Gina, Dan, Ted,

Andy, and Beth (top 2 rows) exhibited high levels of HM during the alone condition and












Deanna


5 10 15




10- \
0-


0*
5 10 15 20


10-

0




5 10 15
Deborah

0-

0-
0-

0-


5 10 15 20 25
0-
Mike
0-
0-

0.

D
10 20 30 40


Gina







5 10 15










5 10 Is
And














5 10 15










5 10 15 20







Lim







5 10 15 20

SESSIONS


5 10 15






Beth 10 ----


5 10 15


5 10 15


5 10 15 20


-- Alone
-- Attention
Play
-*--- Demand


Figure 2: Percentage of intervals HM during the functional analysis from Experiment 2
for all Experiment 3 participants. See text for details.








Table 3
FA patterns for all participants during the functional analysis from Experiment 2

Participant FA Pattern

Deanna High in alone, lower levels in attention and demand
Gina High in alone, lower levels in attention and demand
Dan High in alone, lower levels in attention and demand
Ted High in alone, lower levels in attention and demand
Andy High in alone, lower levels in attention and demand
Beth High in alone, lower levels in attention and demand
Kate High and undifferentiated across all conditions
Lynn High and undifferentiated across all conditions
Avery High and undifferentiated across all conditions
Deborah Extended alone and demand sessions; not differentially high in demand
James Extended alone and demand sessions; not differentially high in demand
Keri Extended alone and demand sessions; not differentially high in demand
Mike Extended alone, attention, play, and demand sessions; undifferentiated
across conditions
Lisa Extended alone sessions; high in alone, lower levels in attention and
demand conditions


lower levels in the attention and demand conditions. Kate and Lynn exhibited high and

undifferentiated levels of responding across alone, attention, and demand conditions,

whereas Avery exhibited high and undifferentiated levels of responding across alone and

lower levels in the attention and demand conditions. Kate and Lynn exhibited high and

undifferentiated levels of responding across alone, attention, and demand conditions,

whereas Avery exhibited high and undifferentiated levels of responding across alone and

attention conditions, with much lower levels in the demand condition (middle row).

Deborah, James, Keri, Mike, and Lisa exhibited undifferentiated responding that was not

differentially high in the alone condition, requiring a more extended analysis (bottom two

rows). Deborah, James, Keri, and Lisa exhibited high levels of responding during both








alone and demand condition; thus, we extended and alternated alone and demand

conditions for them. Mike exhibited high levels of HM across alone, attention, and

demand conditions, and responding was not differentially high during the alone

condition; thus, we conducted alone, attention, play, and demand condition in a 1:1 ratio

until a clear function could be determined. All patterns indicated HM was maintained by

automatic reinforcement. Table 3 shows a summary of the functional analysis patterns

observed during Experiment 2 for the 14 individuals who participated in Experiment 3.

Phase II: Preference Assessments

Leisure Item Assessment

The purpose of this assessment was to identify preferred leisure items that might

effectively compete with participants' HM during NCR sessions. Because all participants

were exposed to the NCR treatment, all participants received a leisure item assessment.

Leisure items presumed to provide different types of sensory stimulation (tactile,

auditory, olfactory) were selected for use during the leisure item assessment. Item

contact was defined as any contact between the participant's hand and the leisure item.

Observers recorded the duration of item contact using a stopwatch, paper, and pencil.

Procedures were based on those described by DeLeon, Iwata, Conners, and Wallace

(1999). Ten-to-15 items were singly presented. Each item was presented for a 2-min

trial, which was repeated 3 times, for a total duration of 6-min per item. Data were

summarized as the total duration of item contact. Interobserver agreement was assessed

during a mean of 35.7% of the trials. Mean interobserver agreement for duration of item

contact was 92.2% (range, 70.7% to 100%).








Food Item Assessment

The purpose of the food item assessment was to identify a preferred edible that

might function as a reinforcer for shaping item contact during DRA sessions (see below).

Five individuals who continued to engage in high levels of HM and low levels of item

contact during NCR sessions (see below) received a food preference assessment before

participating in the DRA intervention. The assessment procedure was similar to that

described by Pace, Ivancic, Edwards, Iwata, and Page (1985). Ten edibles were

presented singly three times each. Observers recorded whether or not an approach

response occurred for each trial of food presentation. Approach responses were defined

as the participant's placement of the edible into her/his mouth. We altered this definition

slightly for individuals who could not eat independently (i.e., approach was defined as

swallowing the item after it was placed in the participant's mouth). Data were

summarized as percentage of approach responding by dividing the number of trials

during which an approach response occurred for an item by the total number of trials

during which the item was presented and multiplying by 100%. Interobserver agreement

was assessed during a mean of 33.3% of the trials. Mean interobserver agreement for

approach responding was 100% (range, 100% to 100%).

Results and Discussion

Results of the leisure item assessment indicated a hierarchy of preferred leisure

items for 12 of the 14 participants (the remaining 2 participants interacted with all items

100% of the time). Figure 2 (top panel) shows data from Lynn's preference assessment,

which are representative data of the 12 participants who exhibited differential responding

during the assessment. The three items associated with the highest duration of item









~ 0
360-



240-l


0
S120-
< * *LYNN


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
U LEISURE ITEMS

ii
z


,_, 100-
80

i 0







O .
1 111 11







1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
FOOD ITEMS
Figure 2: Duration of contact with each of the items presented during Lynn's leisure item
assessment (top panel), and percentage of trials approach for each item
presented during Mike's food assessment (bottom panel)








contact for Lynn were a string of beads, a ribbon, and a plastic ring. At least one of these

three items was used during Lynn's NCR sessions.

Results of the food item assessment suggested several preferred food items for all

participants. Data for three of the five participants indicated a hierarchy of preferred

leisure items; data for the remaining two participants indicated 100% approach

responding across all items. Figure 2 (bottom panel) shows Mike's data from this

assessment, which are representative of those obtained for participants who exhibited

differential responding during this assessment. The three edibles associated with the

highest percentage of approach responding for Mike were Fig Newtons, graham crackers,

and cookies. At least one of these edibles was delivered during Mike's DRA sessions.

Phase III: Treatment Evaluation

Participants and Settings

The same 14 individuals who received a functional analysis in Phase I

participated in the treatment evaluation phase. Settings were identical to those of Phases

1 and2.

Data Collection and Reliability

Data on the percentage of intervals containing HM were collected during baseline,

and data on the percentage of intervals containing HM and item contact were recorded

during all subsequent treatment conditions. Item contact was defined as any contact

between a participant's hand and a leisure item. During sessions in which response

blocking or restraint was implemented, the percentage of intervals in which HM attempts

occurred was recorded. Interobserver agreement was assessed during a mean of 29.1% of

the sessions. Mean interobserver agreement was 94% (range, 85% to 100%) and 95.5%

(range, 85.5% to 100%), for hand mouthing and item contact, respectively.








Experimental Sequence and Design

All participants were exposed to baseline and one or more treatment conditions.

Everyone was exposed to NCR (see below). Based on results observed during the NCR

condition, treatment was complete or additional components were evaluated. The effects

of interventions were evaluated in multiple baseline designs, except as noted below. All

sessions were 10 min in length.

Baseline and Treatment Components

Baseline. All participants were exposed to this condition. Baseline sessions were

similar to the alone condition of the functional analysis in that a therapist was present;

however, no interaction or materials were delivered.

Noncontingent reinforcement (NCR). All participants were exposed to this

condition. During NCR sessions, participants had continuous access to one or more of

their top three preferred leisure items identified during the leisure item preference

assessment. The item(s) that were used were held constant across all NCR sessions.

That is, if the top preferred item was the only item used, it was the only item during all

NCR sessions, and if the second most preferred item was used, this item was used during

all NCR sessions. A therapist was present; however, no interaction was delivered.

Differential reinforcement of alternative behavior (DRA). Individuals who

exhibited low levels of item contact during NCR sessions were exposed to this

component in an attempt to increase their engagement with leisure items. In addition,

one participant who exhibited high levels of item contact (and high levels of HM) was

also exposed to this treatment component. During these sessions, participants had

continuous access to their top preferred leisure items and were physically prompted to

engage in item contact. A three-step prompting sequence (vocal, model, and physical








guidance) was initiated every 30 s, unless the participant was already engaging in item

contact at the beginning of the interval. The prompting sequence began with a vocal

prompt (e.g., "Play with the item"). If the participant did not begin manipulating the item

within 5 s, then the therapist modeled the appropriate behavior while repeating the vocal

prompt. If the participant did not engage with the item within 5 s of the model prompt,

the therapist physically guided her/him to do so. Participants' most preferred food item,

as identified from the food preference assessment, was delivered contingent upon item

contact. The duration of item contact required for reinforcer delivery varied for each

participant depending upon the mean duration exhibited during one of the NCR sessions.

The specific durations required for reinforcer delivery ranged from 2-to-5 s.

Response blocking. Eight individuals whose HM occurred at high levels during

the NCR condition were exposed to this treatment component in an attempt to suppress

their HM. During this intervention, a therapist stood behind the participant and blocked

participants' HM attempts on a fixed-ratio 1 schedule. The therapist blocked

participants' hands from entering their mouths by placing the palm of the therapist's hand

about 2 cm in front of the participants' mouths. HM attempts were scored each time a

participant's hand contacted the back of the therapist's hand.

Brief manual restraint. One individual, Avery, whose HM was not responsive to

other interventions, was exposed to brief manual restraint. During these sessions, Avery

had continuous access to her top preferred leisure items. Contingent upon HM attempts,

the therapist held Avery's hands in her lap for 10-s. Session time did not include time

during which Avery was in brief restraint.








Results and Discussion

NCR

Figure 3 shows treatment results for Debra, Deanna, Gina, Ted, Dan, and James. During

baseline, participants exhibited moderate-to-high levels of HM. During NCR, HM

decreased to very low levels, and item contact occurred at very high levels for all

participants. Decreases in HM and increases in item contact were immediate for all

participants except Dan. Dan exhibited low-to-moderate levels of item contact during the

first 2 NCR sessions; however, item contact increased to high levels during subsequent

sessions. Ted and James never exhibited HM and engaged in item contact during 100%

of the intervals across all NCR sessions. Six of the 14 individuals who received the NCR

component were successfully treated by this intervention. The 8 individuals who were

not successfully treated during NCR received one or more of the subsequent treatment

components.

NCR + DRA + Response Blocking

Figure 4 shows treatment results for Andy, Keri, Beth, Mike, and Kate. During

baseline, participants exhibited moderate-to-high levels of HM. During NCR, Kate's HM

did not decrease and continued to occur at high levels; Beth's, Mike's, and Andy's HM

decreased initially and then increased and maintained at high levels, and Keri's HM

decreased and maintained at low levels. All participants except for Mike exhibited low

levels of item contact during NCR sessions. In an attempt to reduce HM (for Beth, Mike,

Cate, and Andy) and to increase item contact (Beth, Keri, Kate, and Andy), participants

were exposed to the NCR + DRA + Response blocking intervention. During this






40


BL NCR BL NCR
100 -

80 HM.D
Item
60- Contact

40

20- DEBRA

0- __ ^______ --------------

100- y1






100O- o ; c- --/
80-

60 -

1 40- -

20- DEANNA a



100-

80-

60-

40-\

20- ONA JAN

0152
5 10 15 20 5 10 15
SESSIONS

Figure 3: Percentage of intervals HM or HM and item contact during BL and NCR
sessions for Debra (left top panel), Deanna (left middle panel), Gina (left
bottom panel), Ted (right top panel), Dan (right middle panel), and James
(right bottom panel).


ED


]








IES






41


BL NCR NCR + DRA + Blocking


80- Item
60- Contact
40- I I D
20- HETH

100-
80-
60-
40-
2O- __ ~- KE R1
20-



MIKEL

80-
610-
S80-
r 60-
0
40-
20-
U 0- ---=v- -
~100 S
80-
60-
40
20-
KAME
0~
S100- 0
80-
60-
40-
20 -l ANDY?*^ r~--


20-

5 10 SESSIONS 15 20




Figure 4: Percentage of intervals HM or HM and item contact during BL, NCR, and
NCR + DRA + response blocking sessions for Beth (top panel), Keri (second
panel), Mike (third panel), Kate (fourth panel), and Andy (bottom panel)








condition, HM decreased to much lower levels (Beth, Mike, Kate, and Andy) or

continued to occur at low levels (Keri). Item contact increased for Beth, Keri, Kate, and

Andy and maintained at high levels for Mike. In summary, 4 out of 7 individuals who

exhibited high levels of HM during NCR received the NCR + DRA + blocking

component, and all 4 were successfully treated by this intervention. Four out of 4

individuals who exhibited low levels of item contact during NCR also received the NCR

+ DRA + blocking component, and item contact successfully increased during this

intervention. After conducting the NCR + DRA + blocking component with one

participant (Mike) who exhibited high levels of item contact during NCR, we modified

the treatment package by just conducting NCR + blocking for all subsequent participants

exhibiting high levels of item contact during NCR (see below).

NCR + response blocking

Figure 5 shows treatment results for Lynn and Lisa. During baseline, they

exhibited moderate-to-high levels of HM. During NCR, Lynn initially exhibited

moderate levels of item contact; however, item contact increased to high levels across

subsequent sessions. In addition, HM continued to occur at moderate-to-high levels.

During NCR, Lisa immediately exhibited high levels of item contact that maintained

throughout the NCR phase. Lisa's HM occurred at variable levels, often occurring at

very low levels for several consecutive NCR sessions and then reemerging at very high

levels. Because Lynn's and Lisa's item contact occurred at high levels during NCR,

there was no need for the DRA component. Therefore, Lynn and Lisa were exposed to

NCR + response blocking only. During NCR + response blocking, Lynn's HM



















25- I I
> LYNN
o v, _^ \_,
s _____ __ rh
1 0 0 U .^

S75- i

50.

25-1\ r\


10 20 30 40 50 60
SESSIONS

Figure 5: Percentage of intervals HM or HM and item contact during baseline, NCR, and
NCR + response blocking (without DRA) for Lynn (top panel) and Lisa
(bottom panel)








immediately decreased, but remained variable on several sessions until eventually

reaching low levels. In addition, after the first 4 sessions, Lynn's item contact decreased

to very low levels; however, during the 12th session, item contact recovered and remained

high during all subsequent sessions. During NCR + response blocking, Lisa's HM

decreased to lower levels with less variability than that observed during the NCR

component. In addition, Lisa's item contact remained at very high levels.

NCR + brief manual restraint

Avery's results are shown in Figure 6. During baseline, Avery exhibited high levels of

HM. During NCR, HM decreased but continued to occur at moderate levels, and item

contact occurred at high levels. During the NCR + response blocking intervention, Avery

continued to exhibit moderate levels of HM and high levels of item contact. Because

NCR + response blocking did not successfully reduce Avery's HM, we

assessed the effectiveness of NCR + brief manual restraint. During a return to baseline,

Avery's HM returned to high levels similar to that observed during the previous BL

condition. During NCR + manual restraint, HM immediately decreased to very low

levels. During the 10th session of the NCR + manual restraint condition, HM increased

slightly for 7 sessions and then decreased again to low levels. Item contact occurred at

high levels throughout the NCR + manual restraint condition. During a second return to

baseline, Avery exhibited high levels of HM. During the final NCR + restraint condition,

HM immediately decreased to near-zero levels and item contact occurred at high levels.

These levels of HM and item contact maintained throughout the final NCR + brief

manual restraint phase.

Table 3 shows a summary of the number of participants who received each

treatment component and the number who were successfully treated by that component.

























RESTRAINT BL


Iiem 4







25- V I



10 20 30 40 50 60 70 80
SESSIONS



Figure 6: Percentage of intervals HM or HM and item contact during baseline, NCR,
NCR + response blocking, and NCR + brief restraint conditions for Avery








NCR was implemented with all 14 participants, and 6 were successfully treated by this

intervention. NCR + DRA + response blocking was implemented with 5 individuals, and

all 5 were successfully treated by this intervention. NCR + response blocking (without

DRA) was implemented with 3 individuals, and 2 of the 3 participants were successfully

treated by this intervention. Finally, 1 individual participated in the NCR + brief restraint

condition, and she was successfully treated by this intervention. In summary, the

progressive series of interventions effectively reduced HM and increased item contact for

all 14 participants.


Table 3: Treatment Summary


Intervention N participated N successful
NCR 14 6
NCR + DRA + R Block 5 5
NCR + R Block 3 2
NCR + Brief Restraint 1 1
TOTAL 14

The present findings are noteworthy for a number of reasons. First, because all

participants were successfully treated with interventions consistent with a hypothesis that

behavior was maintained by automatic reinforcement (i.e., none of the interventions

involved manipulation of social consequences), the validity of the modified functional

analysis used in this investigation was supported. Second, participants were successfully

treated with various combinations of NCR, DRA, response blocking, and brief manual

restraint, replicating previous research demonstrating the utility of these interventions.

Third, progressively adding on different treatment components and basing treatment

selection on response patterns from previous components was effective in identifying the

least intrusive intervention for effectively reducing HM and/or increasing item contact for

all participants. For example, response blocking + DRA was implemented with









individuals who exhibited high levels of HM and/or low levels of item contact during

NCR, whereas response blocking (without the DRA component) was implemented with

individuals who exhibited high levels of HM and high levels of item contact during NCR.

In addition, the most intrusive intervention, brief manual restraint, was required for only

one participant.

The additive treatment design used in this investigation extends that used by

Lindberg et al. (1999), who progressively introduced treatment components in an additive

fashion for problem behaviors maintained by automatic reinforcement. For example,

they sequentially introduced NCR, prompting, DRA, response blocking, and protective

equipment with two participants who exhibited stereotypy. Results indicated that

response blocking or protective equipment was necessary for reducing participants'

stereotypy. The present study extends that used by Lindberg et al. in two ways: (a) we

implemented the treatment design with a greater number of participants and (b) we

obtained successful results with each of the different treatment components used. The

treatment design used in the present study also extends that used by Cooper et al. (1995),

who assessed the effectiveness of various treatment components for reducing pediatric

feeding problems. Cooper et al. used a similar design to that used in the present study in

that they progressively altered treatment components. However, their design differed

from that used in the present study in that the entire treatment package was implemented

first, and, if it was found effective, selected components were progressively removed to

evaluate their independent contributions. For example, during the first treatment phase,

several different interventions noncontingentt access to toys and social interaction,

escape extinction, praise and a sip of liquid contingent on acceptance of food) were in








effect. Next, one component (e.g., noncontingent access to toys and social interaction)

was removed while the other components remained. Our design differs from that used by

Cooper et al. in that we evaluated the effectiveness of a specific component initially (i.e.,

NCR), and, if that component was found ineffective, we progressively added more

components. Both methods allow for an evaluation of specific treatment components.

However, an advantage of progressively adding on treatment components rather than

removing them is that if an initial component is demonstrated effective, additional

interventions are unnecessary. Thus, the participant does not experience more intrusive

components unless they are needed to obtain a successful outcome. The limitation of

progressively adding on components is that, in the event that the initial components are

ineffective, it may take longer to identify an effective treatment package.

A surprising finding was the limited effectiveness of NCR, suppressing HM in

only 8 of the 14 participants (57.1%). One possible explanation for the limited

effectiveness of NCR may have been the fact that several of the participants did not have

item manipulation in their current repertoires. In addition, for individuals who did have

item contact in their repertoires, the leisure items may not have effectively competed with

the sensory stimulation produced by HM. For example, although Mike engaged in item

contact during NCR, his HM did not decrease, suggesting that the stimulation produced

by the leisure items did not effectively compete with Mike's HM. Future research could

identify what factors were responsible for the ineffectiveness of NCR (lack of item

contact in repertoire or lack of competition) by conducting NCR alone both before and

after DRA. If NCR is ineffective because participants do not have item contact in their

repertoire, NCR alone might maintain high levels of item contact and low levels of








problem behavior once item contact is shaped during DRA. In addition, future research

is needed to determine factors that may increase the likelihood of identifying leisure

items that may compete with HM. One way to increase the probability that leisure items

used during NCR will compete with item contact is to measure HM during the preference

assessment and only use items correlated with both high levels of item contact and low

levels of HM (Roscoe, Iwata, & Goh, 1998). Another method that may used to increase

the likelihood that items used during NCR will compete with HM is to identify items that

produce stimulation similar to that produced by the behavior (Piazza, Adelinis, Hanely,

Goh, & Delia, 2000).

The only treatment component that successfully reduced HM and/or increased

item contact for all participants (N=5) was the DRA intervention. There are three

possible reasons for the high degree of effectiveness of DRA in the present investigation.

First, the reinforcers used during DRA were identified through a formal preference

assessment. Second, item contact, when shaped, effectively competed with participants'

HM. Third, DRA was always combined with response blocking, which may have

contributed to the effects obtained. Future research could determine under what

conditions DRA is most effective (i.e., when implemented alone or when combined with

response blocking) to assess the extent to which DRA depends on being combined with

response blocking.













DISCUSSION

Hand mouthing has been referred to as a form of both stereotypic and self-

injurious behavior, and it has been described as behavior not maintained by social

contingencies. However, the classification of hand mouthing into any one of these

categories seems questionable in the absence of appropriate assessment methods. Direct

observations of participants' injury associated with HM indicated that HM, when it does

occur, is injurious about a third of the time. Because HM may or may not be injurious, it

may be best not to refer to HM as exclusively a stereotypic or a self-injurious behavior.

However, in the present study, results of functional analyses indicated that HM was

overwhelmingly maintained by automatic reinforcement. Nevertheless, given that one

participant's HM was maintained by social reinforcement (attention), it is important to

conduct a functional analysis to definitely rule out behavioral maintenance by social

contingencies. If results of the functional analysis indicate that the behavior is

maintained by social contingencies, interventions found effective for behavior maintained

by social reinforcement should be used.

It is unclear why such a high proportion of participants' HM was maintained by

automatic reinforcement. One explanation may be that, unlike other forms of SIB that

produce more immediate damage (e.g., head banging), HM may emerge initially as a

fairly benign behavior. As a result, it may not produce immediate social consequences,

and, as a result, may be less likely than more severe forms of SIB to acquire a social








function. Thus, when HM eventually produces injuries (and results in social

consequences), these consequences do not play a role in maintenance.

In this study, a relatively efficient functional analysis, consisting primarily of

alone sessions interspersed with attention and demand sessions, was developed. Only 17

of the 64 functional analyses needed to be extended before a function could be

determined. Although the results from the present study provided support for the

modified functional analysis design used in this investigation, future research is needed to

determine the reliability and validity of these findings. For example, researchers could

assess whether this functional analysis method is effective when applied to response

topographies often maintained by automatic reinforcement (e.g., noninjurious stereotypic

behavior).

A strength of this study was that the utility of the functional analysis design was

assessed by conducting treatment procedures based on the findings obtained during the

functional analysis. Given that all participants were successfully treated by interventions

specifically designed for behavior maintained by automatic reinforcement, the validity of

the modified functional analysis design was supported.

Because most participants' HM was maintained by automatic reinforcement, the

utility of the specific functional analysis sequence that we used may not be appropriate

for behavior suspected to be maintained by social contingencies. Because previous

research indicated that hand mouthing is often maintained by nonsocial variables, we

conducted alone sessions in a 2:1 ratio with attention and demand sessions. By contrast,

for response topographies presumed to be maintained by social variables (e.g.,








aggression), a modified functional analysis consisting of mostly attention and demand

sessions could be conducted with very few play and/or alone sessions.

The treatment model used during Experiment 3 demonstrated the efficacy of

starting with a minimally intrusive and efficient treatment component, such as NCR, and

then progressively introducing more effortful and/or intrusive interventions as needed.

This model proved highly effective in that all 14 participants were successfully treated by

one or more of the treatment components. In addition, brief manual restraint was

required for only one participant. Future research could evaluate the effectiveness of the

specific treatment sequence used in this study or other sequences that progressively

introduce more restrictive interventions across other response topographies maintained by

automatic reinforcement. In addition, this method of program implementation (starting

with reinforcement-based interventions or interventions requiring minimal effort and

gradually introducing more restrictive treatment components as needed) could also be

evaluated for behavior maintained by social reinforcement.

The treatment study also extended previous research by conducting several

comparative analyses. For example, the effects of NCR alone were compared with NCR

+ response blocking, and the effects of NCR + response blocking were compared with

NCR + brief manual restraint. Because the purpose of our treatment design was to

progressively introduce more intrusive interventions as needed, each component was not

evaluated separately (e.g., response blocking was always combined with NCR, and DRA

was always combined with response blocking). Future research is needed to determine

which of the various treatment components is most effective when compared in isolation.








Future research is also needed to assess methods for promoting maintenance and

generalization of interventions for HM and other topographies maintained by automatic

reinforcement. Although effective interventions were identified for all 14 participants in

the present study, it remains unclear if treatment effects would maintain over long periods

of time. In addition, it is unclear to what extent these effects would generalize to other

settings. Many of the participants in this study required the DRA + response blocking

component, which was very labor intensive and may not have been practical for staff

implementation for long. Although this procedure was often highly successful in

producing clinically acceptable outcomes, steps should be taken to fade the amount of

effort required for program implementation to a more practical level. For example, the

duration of item contact required for reinforcer delivery could be gradually increased

during DRA, and/or the schedule of blocking and restraint could be gradually reduced.

Future research is needed to explore these and other methods for promoting

generalization and maintenance of the treatment components used in this study.















REFERENCES


Ball, T. S., Campbell, R., & Barkemeyer, R. (1980). Air splints applied to control
self-injurious finger sucking in profoundly retarded individuals. Journal of Behavior
Therapy and Experimental Psychiatry. 11, 267-271.

Bodfish, J. W., Crawford, T. W., Powell, S. B., Parker, D. E., Golden, R. N., &
Lewis, M. H. (1995). Compulsion in adults with mental retardation: Prevalence,
phenomenology, and comorbidity with stereotypy and self-injury. American Journal on
Mental Retardation. 100, 183-192.

Cooper, L. J., Wacker, D. P., Thursby, D., Plagmann, L. A., Harding, J., Millard,
T., & Derby, M. (1992). Analysis of the effects of task preferences, task demands, and
adult attention on child behavior in outpatient and classroom settings. Journal of Applied
Behavior Analysis. 25, 823-840.

Cooper, L. J., Wacker, D. P., McComas, J. J., Brown, K., Peck, S. M., Richman,
D., Drew, J., Frischmeyer, P., & Millard, T. (1995). Use of component analyses to
identify active variables in treatment packages for children with feeding disorders. Journal
of Applied Behavior Analysis. 28, 139-153.

DeLeon, I. G., Iwata, B. A., Conners, J., & Wallace, M. D. (1999). Examination of
ambiguous stimulus preferences with duration-based measures. Journal of Applied
Behavior Analysis. 32, 111-114.

Derby, K. M., Wacker, D. P., Sasso, G., Steege, M., Northup, J., Cigrand, K., &
Asmus, J. (1992). Brief functional assessment techniques to evaluate aberrant behavior in
an outpatient setting: A summary of 79 cases. Journal of Applied Behavior Analysis. 25,
713-721.

Dorsey, M. F., Iwata, B. A., Ong, P., & McSween, T. E. (1980). Treatment of
self-injurious behavior using a water mist: Initial response suppression and generalization.
Journal of Applied Behavior Analysis. 13, 343-353.

Dura, J.R., Mulick, J.A., & Rasnake, L.K. (1987). Prevalence of stereotypy
among institionalized nonambulatory profoundly mentally retarded people. American
Journal of Mental Deficiency. 91, 548-549.








Durand, V. M., & Crimmins, D. B. (1988). Identifying the variables maintaining
self-injurious behavior. Journal of Autism and Developmental Disorders. 18,99-117.

Favell, J. E., McGimsey, J. F., & Schell, R. M. (1982). Treatment of self-injury
by providing alternate sensory activities. Analysis and Intervention in Developmental
Disabilities. 2, 83-104.

Goh, H., & Iwata, B. A. (1994). Behavioral persistence and variability during
extinction of self-injury maintained by escape. Journal of Applied Behavior Analysis. 27,
173-174.

Goh, H., Iwata, B. A., Shore, B. A., DeLeon, I. G., Lerman, D. C., Ulrich, S. M.,
& Smith, R. G. (1995). An analysis of the reinforcing properties of hand mouthing.
Journal of Applied Behavior Analysis. 28, 269-283.

Griffin, J. C., Williams, D. E., Stark, M. T., Altmeyer, B. K., & Mason, M.
(1986). Self-injurious behavior: A state-wide prevalence survey of the extent and
circumstances. Applied Research in Mental Retardation. 7, 105-116.

Homer, R. D. (1980). The effects of an environmental "enrichment" program on
the behavior of institutionalized profoundly retarded children. Journal of Applied
Behavior Analysis. 13,473-491.

Iwata, B. A., Dorsey, M. F., Slifer, K. J., Bauman, K. E., & Richman, G. S.
(1994a). Toward a functional analysis of self-injury. Journal of Applied Behavior
Analysis. 27. 197-209. (Reprinted from Analysis and Intervention in Developmental
Disabilities, 2,3-20,1982)

Iwata, B. A., Pace, G. M., Dorsey, M. F., Zarcone, J. R., Vollmer, T. R., Smith,
R. G., Rodgers, T. A., Lerman, D. C., Shore, B. A., Mazaleski, J. L., Goh, H., Cowdery,
G. E., Kalsher, M. J., & Willis, K. D. (1994b). The functions of self-injurious behavior:
An experimental-epidemiological analysis. Journal of Applied Behavior Analysis. 27,
215-240.

Kahng, S., & Iwata, B. A. (1999). Correspondence between outcomes of brief and
extended functional analyses. Journal of Applied Behavior Analysis. 32, 149-159.

LaGrow, S. J., & Repp, A. C. (1984). Stereotypic responding: A review of
intervention research. American Journal of Mental Deficiency. 88, 595-609.

Lerman, D. C., & Iwata, B. A. (1996). A methodology for distinguishing between
extinction and punishment effects associated with response blocking. Journal of Applied
Behavior Analysis. 29, 231-233.








Lindberg, J. S., Iwata, B. A., & Kahng, S. K. (1999). On the relation between
object manipulation and stereotypic self-injurious behavior. Journal of Applied Behavior
Analysis. 32, 51-62.

Lockwood, K., & Williams, D. E. (1994). Treatment and extended follow-up of
chronic hand mouthing. Journal of Behavior Therapy and Experimental Psychiatry. 25,
161-169.

Maurice, P., & Trudel, G. (1982). Self-injurious behavior: Prevalence and
relationships to environmental events. In J. H. Hollis & C. E. Meyers (Eds.), Life-
threatening behavior: Analysis and intervention (pp. 81-103). Washington: American
Association on Mental Deficiency.

Mazaleski, J. L., Iwata, B. A., Rodgers, T. A., Vollmer, T. R., & Zarcone, J. R.
(1994). Protective equipment as treatment for stereotypic hand mouthing: Sensory
extinction or punishment effects? Journal of Applied Behavior Analysis. 27, 345-355.

McClure, J. T., Moss, R. A., McPeters, J. W., & Kirkpatrick, M. A. (1986).
Reduction of hand mouthing by a boy with profound mental retardation. Mental
Retardation. 24, 219-222.

McGill, P., Hughes, D., Teer, K., & Rye, L. (2001). Variability in staff reports of
the frequency of challenging behavior. Research in Developmental Disabilities. 22, 221-
231.

Northup, J., Wacker, D., Sasso, G., Steege, M., Cigrand, K., Cook, J., & DeRaad,
A. (1991). A brief functional analysis of aggressive and alternative behavior in an outclinic
setting. Journal of Applied Behavior Analysis. 24, 509-522.

Pace, G. M., Ivancic, M. T., Edwards, G. L., Iwata, B. A., & Page, T. J. (1985).
Assessment of stimulus preference and reinforcer value with profoundly retarded
individuals. Journal of Applied Behavior Analysis. 18, 249-255.

Paisey, T. J. H., Whitney, R. B., & Wainczak, S. M. (1993). Case study:
Noninvasive behavioral treatment of self-injurious hand stereotypy in a child with Rett
syndrome. Behavioral Residential Treatment, 8, 133-145.

Piazza, C. C., Adelinis, J. D., Hanley, G. P., Goh, H., & Delia, M. D. (2000). An
evaluation of the effects of matched stimuli on behaviors maintained by automatic
reinforcement Journal of Applied Behavior Analysis. 33, 13-27.

Rast, J., & Jack, S. (1992). Mouthing. In E.A. Konarski, J.E. Favell (Eds.),
Manual for the assessment and treatment of the behavior disorders of people with mental
retardation (pp. 1-11). Morganton, NC: Western Carolina Center Foundation.








Repp, A. C., Singh, N. N., Olinger, E., & Olson, D. R. (1990). The use of
functional analyses to test causes of self-injurious behavior: Rationale, current status, and
future directions. Journal of Mental Deficiency Research. 34, 95-105.

Ringdahl, J. E., Vollmer, T. R., Marcus, B. A., & Roane, H. S. (1997). An
analogue evaluation of environmental enrichment: The role of stimulus preference.
Journal of Applied Behavior Analysis. 30, 203-216.

Rojahn, J. (1986). Self-injurious and stereotypic behavior of noninstitutionalized
mentally retarded people: Prevalence and classification. American Journal of Mental
Deficiency. 91, 268-276.

Roscoe, E. M., Iwata, B. A., & Goh, H. (1998). A comparison of noncontingent
reinforcement and sensory extinction as treatments for self-injurious behavior. Journal of
Applied Behavior Analysis. 31, 635-646.

Schroeder, S. R., Schroeder, C. S., Smith, B., & Dalldorf, J. (1978). Prevalence of
self-injurious behaviors in a large state facility for the retarded: A three year follow-up
study. Journal of Autism and Childhood Schizophrenia. 8, 261-269.

Shore, B. A., Iwata, B. A., DeLeon, I. G., Kahng, S., & Smith, R. G. (1997). An
analysis of reinforcer substitutability using object manipulation and self-injury as
competing responses. Journal of Applied Behavior Analysis. 30, 21-41.

Singh, N. N., & Millichamp, C. J. (1987). Independent and social play among
profoundly mentally retarded adults: Training, generalization, and long-term follow-up.
Journal of Applied Behavior Analysis. 20, 23-34.

Thompson, R. H., Iwata, B. A., Conners, J., & Roscoe, E. M. (1999). Effects of
reinforcement for alternative behavior during punishment of self-injury. Journal of
Applied Behavior Analysis. 32, 317-328.

Troster, H., Brambring, M., & Beelmann, A. (1991). Prevalence and situational
causes of stereotyped behaviors in blind infants and preschoolers. Journal of Abnormal
Child Psychology. 19, 569-590

Turner, W. D., Realon, R. E., Irvin, D., & Robinson, E. (1996). The effects of
implementing program consequences with a group of individuals who engaged in sensory
maintained hand mouthing. Research in Developmental Disabilities. 17, 311-330.

Vaughan, M. E. & Michael, J. L. (1982). Automatic reinforcement: An important
but ignored concept. Behaviorism. 10, 217-228.








Vollmer, T. R., Marcus, B. A., & LeBlanc, L. (1994). Treatment of self-injury and
hand mouthing following inconclusive functional analyses. Journal of Applied Behavior
Analysis. 27, 331-344.

Vollmer, T. R., Marcus, B. A., Ringdahl, J. E., & Roane, H. S. (1995). Progressing
from brief assessments to extended experimental analyses in the evaluation of aberrant
behavior. Journal of Applied Behavior Analysis. 28, 561-576.

Wacker, D., Steege, M., Northup, J., Reimers, T., Berg, W., & Sasso, G. (1990).
Use of functional analysis and acceptability measures to assess and treat severe behavior
problems: An outpatient clinic model. In A. C. Repp & N. N. Singh (Eds.), Perspectives
on the use of nonaversive and aversive interventions for persons with developmental
disabilities (pp. 349-359). Sycamore, IL: Sycamore Publishing Co.

Wallace, M. D., & Iwata, B. A. (1999). Effects of session duration on functional
analysis outcomes. Journal of Applied Behavior Analysis. 32, 175-183.













BIOGRAPHICAL SKETCH

Eileen Roscoe was born in Miami, FL. She entered the University of Florida in

1992 and graduated in 1996 with a degree in psychology. During her undergraduate

studies, she enrolled in the applied behavior analysis lab and gained her first exposure to

behavior analysis and individuals diagnosed with developmental disabilities. Following

this, she served as a volunteer at the Florida Center on Self-Injury, where she gained

experience conducting assessments and treatments of severe behavior disorders. Her

experience at the Florida Center on Self-Injury led her to seek graduate training in the

field of behavior analysis. In the Fall of 1996, she enrolled in the behavior analysis

program at the University of Florida to pursue a doctoral degree in psychology. She

worked as a graduate research assistant at the Florida Center on Self-Injury from 1996 to

2001 under the direction of Dr. Brian Iwata. Expecting graduation in August, 2001,

Eileen has accepted an appointment as a post-doctoral fellow at the Marcus Institute, a

hospital located in Atlanta, GA, and affiliated with Emory University. She plans to

further her career in applied behavior analysis by conducting clinical work and research

in the application of behavioral principles to the treatment of behavior disorders.









I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and quality,
as a dissertation for the degree of Doctor of Philosophy.




Bn6 A. Iwata, Gnair
Professor of Psychology



I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and quality,
as a dissertation for the degree of Doctor of Philosophy.




Jenner M. Asmus
Assistant Professor of Educational
Psychology


I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and quality,
as a dissertation for the degree of Doctor of Philosophy.



(Tim thy D.PHakenberg
ssciate Professor of Psychology(


I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and quality,
as a dissertation for the degree of Doctor of Philosophy.



Mark H. Lewis
Professor of Psychology










I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and quality,
as a dissertation for the degree of Doctor of Philosophy.



TimothyR. Vollmer
Assistant Professor of Psychology





This dissertation was submitted to the Graduate Faculty of the Department of
Psychology in the College of Liberal Arts and Sciences and to the Graduate School and
was accepted as partial fulfillment of the requirements for the degree of Doctor of
Philosophy.

August, 2001 __________
Dean, Graduate School


















LD

I; ~-.- .~'-

20
sr.s.-L~r


UNIVERSITY OF FLORIDA
I III 13IIIIIIIIiiiiiiiIIII li ll 2IIII 1 I 0
3 1262 08555 0308