Citation
Assessment and treatment of hand mouthing

Material Information

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

Subjects

Subjects / Keywords:
Demand analysis ( jstor )
Design analysis ( jstor )
Functional analysis ( jstor )
Manuals ( jstor )
Participant observation ( jstor )
Philosophical psychology ( jstor )
Recreation ( jstor )
Self injurious behavior ( jstor )
Stereotypies ( jstor )
Topography ( jstor )

Record Information

Rights Management:
Copyright [name of dissertation author]. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Resource Identifier:
27752311 ( ALEPH )
48541207 ( OCLC )

Downloads

This item has the following downloads:


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




Full Text

PAGE 1

$66(660(17 $1' 75($70(17 2) +$1' 0287+,1* $ /$5*(6&$/( $1$/<6,6 %\ (,/((1 0,&+(//( 526&2( $ ',66(57$7,21 35(6(17(' 72 7+( *5$'8$7( 6&+22/ 2) 7+( 81,9(56,7< 2) )/25,'$ ,1 3$57,$/ )8/),//0(17 2) 7+( 5(48,5(0(176 )25 7+( '(*5(( 2) '2&725 2) 3+,/2623+< 81,9(56,7< 2) )/25,'$

PAGE 2

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

PAGE 3

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

PAGE 4

$EVWUDFW RI 'LVVHUWDWLRQ 3UHVHQWHG WR WKH *UDGXDWH 6FKRRO RI WKH 8QLYHUVLW\ RI )ORULGD LQ 3DUWLDO )XOILOOPHQW RI WKH 5HTXLUHPHQWV IRU WKH 'HJUHH RI 'RFWRU RI 3KLORVRSK\ $66(660(17 $1' 75($70(17 2) +$1' 0287+,1* $ /$5*(6&$/( $1$/<6,6 %\ (LOHHQ 0LFKHOOH 5RVFRH $XJXVW &KDLUPDQ %ULDQ $ ,ZDWD 0DMRU 'HSDUWPHQW 3V\FKRORJ\ +DQG PRXWKLQJ +0f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

PAGE 5

,Q ([SHULPHQW ,, PRGLILHG IXQFWLRQDO DQDO\VHV )$Vf ZHUH FRQGXFWHG GXULQJ ZKLFK DORQH DWWHQWLRQ DQG GHPDQG VHVVLRQV ZHUH DOWHUQDWHG LQ PXOWLHOHPHQW GHVLJQV ZLWK D UDWLR RI DORQH WR DWWHQWLRQ DQG GHPDQG VHVVLRQV 5HVXOWV LQGLFDWHG WKDW LQ b RI WKH FDVHV IRU ZKLFK DQ )$ ZDV FRQGXFWHG +0 ZDV PRVW OLNHO\ PDLQWDLQHG E\ DXWRPDWLF UHLQIRUFHPHQW ,Q ([SHULPHQW ,,, LQGLYLGXDOV ZKRVH +0 HLWKHU RFFXUUHG DW YHU\ KLJKUDWHV GXULQJ WKH )$ RU SURGXFHG VHYHUH WLVVXH GDPDJH WR WKH KDQGV SDUWLFLSDWHG ,Q 3KDVH D PRGLILHG )$ ZDV FRQGXFWHG WKDW GHWHUPLQHG SDUWLFLSDQWVf +0 ZDV PDLQWDLQHG E\ DXWRPDWLF UHLQIRUFHPHQW ,Q 3KDVH ,, SUHIHUHQFH DVVHVVPHQWV ZHUH FRQGXFWHG WKDW LGHQWLILHG SUHIHUUHG OHLVXUH DQGRU HGLEOH LWHPV WKDW ZHUH XVHG GXULQJ WUHDWPHQW ,Q 3KDVH ,,, WKH IROORZLQJ LQWHUYHQWLRQV ZHUH LPSOHPHQWHG LQ VHTXHQWLDO RUGHU Df QRQFRQWLQJHQW UHLQIRUFHPHQW 1&5f Ef HLWKHU 1&5 GLIIHUHQWLDO UHLQIRUFHPHQW RI DOWHUQDWLYH EHKDYLRU '5$f UHVSRQVH EORFNLQJ RU 1&5 UHVSRQVH EORFNLQJ RQO\ DQG Ff 1&5 EULHI PDQXDO UHVWUDLQW 1&5 1&5 '5$ UHVSRQVH EORFNLQJ 1&5 UHVSRQVH EORFNLQJ DQG 1&5 EULHI UHVWUDLQW UHVXOWHG LQ LQFUHDVHV LQ LWHP FRQWDFW DQG ODUJH UHGXFWLRQV LQ +0 IRU VL[ ILYH WZR DQG RQH SDUWLFLSDQWVf UHVSHFWLYHO\ Y

PAGE 6

,1752'8&7,21 7KH SXUSRVH RI WKH FXUUHQW LQYHVWLJDWLRQ ZDV WR FRQGXFW D ODUJHVFDOH DVVHVVPHQW DQG WUHDWPHQW RI KDQG PRXWKLQJ D IRUP RI EHKDYLRU H[KLELWHG E\ LQGLYLGXDOV ZLWK GHYHORSPHQWDO GLVDELOLWLHV WKDW PD\ RU PD\ QRW EH LQMXULRXV ,Q ([SHULPHQW WKH SUHYDOHQFH IUHTXHQF\ DQG ULVN RI +0 ZHUH DVVHVVHG XVLQJ LQGLUHFW DQGRU GLUHFW REVHUYDWLRQDO PHWKRGV ,Q ([SHULPHQW D PRGLILHG IXQFWLRQDO DQDO\VLV EDVHG RQ WKDW GHVFULEHG E\ ,ZDWD HW DO Df ZDV FRQGXFWHG WR LGHQWLI\ WKH H[WHQW WR ZKLFK SDUWLFLSDQWVf +0 ZDV VHQVLWLYH WR PDLQWDLQHG E\f VRFLDO FRQWLQJHQFLHV ,Q ([SHULPHQW D SURJUHVVLYH VHULHV RI LQWHUYHQWLRQV ZDV HYDOXDWHG WR GHWHUPLQH ZKLFK WUHDWPHQW RU FRPELQDWLRQ RI WUHDWPHQWV PRVW HIIHFWLYHO\ UHGXFHG +0 +DQG PRXWKLQJ +0f KDV EHHQ GHILQHG DV WKH SODFLQJ RI RQH RU PRUH GLJLWV RI WKH KDQG SDVW WKH SODQH RI WKH OLSV /RFNZRRG t :LOOLDPV f LQVHUWLRQ RI WKH KDQG LQWR WKH PRXWK EH\RQG WKH ILUVW NQXFNOH 'RUVH\ ,ZDWD 2QJ t 0F6ZHHQ f DQGRU DQ\ FRQWDFW EHWZHHQ WKH KDQG DQG WKH PRXWK /HUPDQ t ,ZDWD f 5DVW DQG -DFN f PDGH D GLVWLQFWLRQ EHWZHHQ KDQG PRXWKLQJ DQG WKH PRXWKLQJ RI REMHFWV QRWLQJ WKDW WKH IRUPHU LQYROYHV FRQWDFW RI RQO\ WKH KDQGV DQG WKH PRXWK ZKHUHDV WKH ODWWHU PD\ LQFOXGH VHYHUDO EHKDYLRUV LQYROYLQJ FRQWDFW RI WKH PRXWK ZLWK RWKHU REMHFWV +0 LV WRSRJUDSKLFDOO\ YHU\ VLPLODU WR KDQG ELWLQJ RIWHQ PDNLQJ WKHVH UHVSRQVHV LQGLVWLQJXLVKDEOH WR REVHUYHUV )RU H[DPSOH 3DLVH\ :KLWQH\ DQG :DLQF]DN f FRQGXFWHG D FDVH VWXG\ RI D SDUWLFLSDQWfV VHOILQMXULRXV +0 $OWKRXJK WKH DXWKRUV QRWHG WKDW WKH SDUWLFLSDQW IUHTXHQWO\ HQJDJHG LQ KDQG ELWLQJ WKH\ RQO\ UHFRUGHG LQVWDQFHV RI

PAGE 7

+0 GHILQHG DV LQVHUWLRQ RI ILQJHUV RU KDQG LQWR PRXWKf +RZHYHU EDVHG RQ WKHLU GHILQLWLRQ RFFXUUHQFHV RI +0 PD\ KDYH LQFOXGHG LQVWDQFHV RI ELWLQJ $VVXPLQJ WKDW +0 DQG KDQG ELWLQJ PD\ EH IXQFWLRQDOO\ HTXLYDOHQW LH WKH\ DUH ERWK PDLQWDLQHG E\ WKH VDPH HQYLURQPHQWDO YDULDEOHf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f +0 %DOO &DPSEHOO t %DUNHPH\HU /RFNZRRG t :LOOLDPV 0F&OXUH 0RVV 0F3HWHUV t .LUNSDWULFN f +0 KDV DOVR EHHQ FODVVLILHG DV D IRUP RI fVWHUHRW\S\f EHFDXVH LW LV UHSHWLWLYH KDV QR DSSDUHQW IXQFWLRQ DQG PD\ LQWHUIHUH ZLWK DGDSWLYH EHKDYLRU /D*URZ t 5HSS f 5HJDUGOHVV RI ZKHWKHU D SDUWLFLSDQWfV +0 LV LQMXULRXV LW PD\ JUHDWO\ LPSHGH KHUKLV TXDOLW\ RI OLIH 0RVW QRWDEO\ WKH RIIHQVLYH RGRU DVVRFLDWHG ZLWK +0 PD\ VHUYH DV D EDUULHU WR VRFLDO LQWHUDFWLRQV 0F&OXUH HW DO f ,Q DGGLWLRQ PRXWKLQJ PD\ LQWHUIHUH ZLWK WKH DFTXLVLWLRQ RI DSSURSULDWH EHKDYLRUV %DOO HW DO f 3UHYDOHQFH %HFDXVH +0 KDV EHHQ FODVVLILHG LQ VHYHUDO GLIIHUHQW ZD\V HVWLPDWLQJ LWV SUHYDOHQFH KDV EHHQ GLIILFXOW 7KDW LV VRPH DXWKRUV KDYH GHVFULEHG +0 DV D VHOI LQMXULRXV UHVSRQVH DQG KDYH LQFOXGHG LW LQ VWXGLHV RQ WKH SUHYDOHQFH RI VHOI LQMXU\ *ULIILQ HW DO 0DXULFH t 7UXGHO f ZKHUHDV RWKHUV KDYH FRQVLGHUHG LW D IRUP RI VWHUHRW\S\ %RGILVK HW DO /D*URZ t 5HSS f *LYHQ WKDW VWXGLHV RQ WKH

PAGE 8

SUHYDOHQFH RI VHOILQMXULRXV EHKDYLRU 6,%f W\SLFDOO\ H[FOXGH QRQLQMXULRXV UHVSRQVH WRSRJUDSKLHV DQG VWXGLHV RQ WKH SUHYDOHQFH RI QRQLQMXULRXV VWHUHRW\S\ H[FOXGH VHOI LQMXULRXV WRSRJUDSKLHV UHSRUWV RQ WKH SUHYDOHQFH RI +0 DUH XVXDOO\ LQFRPSOHWH 7KH RQO\ VWXG\ WR GDWH WKDW KDV DVVHVVHG WKH SUHYDOHQFH RI +0 VSHFLILFDOO\ ZDV FRQGXFWHG E\ 7URVWHU %DPEULQJ DQG %HHOPDQQ f ZKR VXUYH\HG SDUHQWV RI FKLOGUHQ PRWR \Uf ZKR ZHUH EOLQG 3DUHQWV ZHUH DVNHG WR LQGLFDWH ZKLFK RI GLIIHUHQW VWHUHRW\SHG EHKDYLRUV HJ VXFNLQJ WKXPEV RU ILQJHUV ERG\ URFNLQJ UHSHWLWLYH KDQG DQG ILQJHU PRYHPHQWV HWFf WKH\ KDG REVHUYHG LQ WKHLU FKLOG WKH\ ZHUH DOVR DVNHG WR LQGLFDWH WKH IUHTXHQF\ RI UHVSRQVH WRSRJUDSKLHV UHSRUWHG WR RFFXU 5HVXOWV LQGLFDWHG WKDW WKH SUHYDOHQFH RI +0 WKDW RFFXUUHG Df DW OHDVW RQFH SHU ZHHN ZDV b Ef RQFH SHU GD\ ZDV b DQG Ff KRXUO\ ZDV b $OWKRXJK WKLV VWXG\ ZDV WKH ILUVW WR LQFOXGH +0 DV D VSHFLILF UHVSRQVH WRSRJUDSK\ LW ZDV FRQGXFWHG RQO\ ZLWK QRUPDOO\ GHYHORSLQJ EOLQG FKLOGUHQ WKXV WKH JHQHUDOLW\ RI WKHVH ILQGLQJV WR DGXOWV RU WR LQGLYLGXDOV ZLWK GHYHORSPHQWDO GLVDELOLWLHV LV XQFOHDU ,Q DGGLWLRQ WKH DXWKRUV XVHG RQO\ LQGLUHFW VXUYH\f PHWKRGV IRU DVVHVVLQJ WKH SUHYDOHQFH RI 6,% VWHUHRW\S\ DQGRU +0 ZKLFK PD\ KDYH EHHQ XQUHOLDEOH RU LQDFFXUDWH ,Q WKH SUHVHQW VWXG\ ZH DVVHVVHG WKH SUHYDOHQFH RI +0 E\ FRQGXFWLQJ D VHULHV RI DVVHVVPHQWV EHJLQQLQJ ZLWK LQGLUHFW DVVHVVPHQWV LH LQWHUYLHZVf DQG HQGLQJ ZLWK GLUHFW REVHUYDWLRQ WR YHULI\ WKH UHVXOWV REWDLQHG GXULQJ WKH LQGLUHFW DVVHVVPHQW %\ FRQGXFWLQJ ERWK LQGLUHFW DQG GLUHFW DVVHVVPHQWV RI +0 PRUH DFFXUDWH PHDVXUHV PD\ EH REWDLQHG ZLWK UHVSHFW WR SUHYDOHQFH 5LVN $V QRWHG SUHYLRXVO\ +0 KDV EHHQ UHSRUWHG WR UHVXOW LQ VHYHUH VHOILQMXU\ %DOO HW DO /RFNZRRG t :LOOLDPV 0F&OXUH HW DO f 5DVW DQG -DFN f

PAGE 9

QRWHG WKDW IRU +0 WR EH FRQVLGHUHG LQMXULRXV PRXWKLQJ VKRXOG UHVXOW LQ fUHG RU ZKLWH KDQGV FKDSSHG OLSV RU IDFH RU OHVLRQV RQ WKH KDQGV RU IDFHf S f ,Q DGGLWLRQ WKH\ LQGLFDWHG WKDW +0 ZKHQ LQMXULRXV RFFXUV ERWK IUHTXHQWO\ DQG IRU ORQJ SHULRGV RI WLPH %DVHG RQ WKHLU GHILQLWLRQ RI LQMXU\ DQ DGHTXDWH DVVHVVPHQW RI ULVN VKRXOG LQFOXGH ERWK WKH IUHTXHQF\ DQG VHYHULW\ RI +0 )HZ VWXGLHV KDYH DVVHVVHG WKH IUHTXHQF\ DQGRU GXUDWLRQ RI VWHUHRW\SLF EHKDYLRUV 'XUD 0XOLFN DQG 5DVQDNH f DVVHVVHG WKH IUHTXHQF\ DQG GXUDWLRQ RI YDULRXV IRUPV RI VWHUHRW\S\ LQFOXGLQJ UK\WKPLF PRYHPHQWV EL]DUUH SRVWXULQJ VHOIUHVWUDLQW DQG REMHFW PDQLSXODWLRQ 3DUWLFLSDQWVf GLUHFW FDUH VWDII ZHUH LQWHUYLHZHG WR GHWHUPLQH ZKHWKHU RU QRW DQ LQGLYLGXDO Df HQJDJHG LQ HDFK W\SH RI VWHUHRW\S\ DQG Ef HQJDJHG LQ WKH VWHUHRW\SLF UHVSRQVH IRU PLQ OWR PLQ RU JUHDWHU WKDQ PLQ $OWKRXJK DVVHVVLQJ WKH GXUDWLRQ RI +0 PD\ SURYH XVHIXO LW ZDV XQFOHDU ZKHWKHU WKH GXUDWLRQ UHSRUWHG ZDV SHU HSLVRGH SHU GD\ RU SHU ZHHN ,I WKH GXUDWLRQ ZDV SHU HSLVRGH LW ZRXOG KDYH EHHQ KHOSIXO WR NQRZ KRZ PDQ\ HSLVRGHV RFFXUUHG SHU GD\ 5RMDKQ f GHWHUPLQHG WKH ULVN RI SDUWLFLSDQWVf VWHUHRW\S\ E\ FRQGXFWLQJ D VXUYH\ FRQVLVWLQJ RI D SRLQW VFDOH IRU VHYHULW\ DQG D SRLQW VFDOH IRU WKH IUHTXHQF\ RI RFFXUUHQFH 6FKURHGHU 6FKURHGHU 6PLWK DQG 'DOOGRUI f XVHG DQ LQGLUHFW DVVHVVPHQW WR GHWHUPLQH WKH VHYHULW\ RI SDUWLFLSDQWVf 6,% $IWHU VRFLDO ZRUNHUV UHIHUUHG LQGLYLGXDOV ZKR H[KLELWHG 6,% D SHGLDWULFLDQ RU SV\FKRORJLVW LQWHUYLHZHG WKH VRFLDO ZRUNHU WR DVVHVV ZKHWKHU WKH 6,% ZDV VHYHUH LW RFFXUUHG DW OHDVW RQFH D GD\ DQG KDG FDXVHG EOHHGLQJ EUXLVHV EURNHQ ERQHV RU RWKHU WLVVXH GDPDJH WKDW UHTXLUHG LQWHUYHQWLRQ E\ D PHGLFDO VWDIIf RU PLOG WKRVH UHIHUUDOV WKDW ZHUH QRW VHYHUHf 7KH VWXGLHV QRWHG DERYH DUH QRWHZRUWK\ LQ WKDW WKH\ DVVHVVHG WKH IUHTXHQF\ DQGRU VHYHULW\ RI VHYHUDO GLIIHUHQW WRSRJUDSKLHV RI 6,% DQGRU VWHUHRW\S\

PAGE 10

+RZHYHU EHFDXVH RQO\ LQGLUHFW DVVHVVPHQWV ZHUH XVHG WKH UHOLDELOLW\ RI WKH ILQGLQJV LV XQFHUWDLQ 6HFRQG QRQH RI WKH VWXGLHV UHSRUWHG GDWD VSHFLILFDOO\ RQ +0 PDNLQJ LW XQFOHDU LI +0 RIWHQ UHVXOWV LQ LQMXU\ $OWKRXJK DXWKRUV KDYH QRWHG WKDW +0 PD\ UHVXOW LQ VHYHUH WLVVXH GDPDJH QR VWXGLHV WR GDWH KDYH FRQGXFWHG D V\VWHPDWLF DVVHVVPHQW LQGLFDWLQJ WKH GHJUHH RI LQMXU\ DVVRFLDWHG ZLWK +0 ,Q WKH SUHVHQW VWXG\ ZH VRXJKW WR GHWHUPLQH WKH ULVN RI +0 H[FOXVLYHO\ E\ FRQGXFWLQJ GLUHFW REVHUYDWLRQV RI WKH IUHTXHQF\ DQG VHYHULW\ RI SDUWLFLSDQWVf +0 )XQFWLRQ 7KH GHYHORSPHQW RI IXQFWLRQDO DQDO\VLV PHWKRGRORJ\ LV EDVHG RQ RYHU \HDUV RI UHVHDUFK LQGLFDWLQJ WKDW EHKDYLRU SUREOHPV HJ DJJUHVVLRQ QRQFRPSOLDQFH DQG 6,%f DUH RIWHQ OHDUQHG SHUIRUPDQFHV 7KDW LV WKH\ DUH PDLQWDLQHG E\ FRQWLQJHQFLHV RI SRVLWLYH RU QHJDWLYH UHLQIRUFHPHQW ,ZDWD 'RUVH\ 6OLIHU %DXPDQ DQG 5LFKPDQ Df SURSRVHG D JHQHUDO PRGHO IRU DVVHVVLQJ EHKDYLRUDO VHQVLWLYLW\ WR D YDULHW\ RI FRQWLQJHQFLHV ,Q WKLV VWXG\ WKUHH WHVW FRQGLWLRQV ZHUH XVHG HDFK FRQWDLQLQJ D GLVFULPLQDWLYH VWLPXOXV 6'f DQ HVWDEOLVKLQJ RSHUDWLRQ (2f DQG D VRXUFH RI UHLQIRUFHPHQW $ IRXUWK FRQGLWLRQ WKH fSOD\f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

PAGE 11

EHKDYLRU ZDV PDLQWDLQHG E\ DXWRPDWLF UHLQIRUFHPHQW 9DXJKQ t 0LFKDHO f 'XULQJ WKH SOD\ RU FRQWURO FRQGLWLRQ OHLVXUH LWHPV ZHUH IUHHO\ DYDLODEOH DWWHQWLRQ ZDV GHOLYHUHG RQ D IL[HGWLPH )7f VFKHGXOH DQG QR GHPDQGV ZHUH SUHVHQWHG %HFDXVH ERWK SRVLWLYH DQG QHJDWLYH UHLQIRUFHUV ZHUH QRQFRQWLQJHQWO\ DYDLODEOH GXULQJ WKLV FRQGLWLRQ WKH SUREDELOLW\ RI EHKDYLRU RFFXUULQJ VKRXOG EH JUHDWO\ UHGXFHG 5HVXOWV VKRZHG WKDW RI WKH SDUWLFLSDQWV H[KLELWHG WKH KLJKHVW OHYHOV RI EHKDYLRU LQ WKH DORQH FRQGLWLRQ VXJJHVWLQJ WKHLU VHOILQMXU\ ZDV PDLQWDLQHG E\ DXWRPDWLF UHLQIRUFHPHQW DQG SDUWLFLSDQWV H[KLELWHG XQGLIIHUHQWLDWHG UHVSRQGLQJ RU UHVSRQGLQJ WKDW ZDV KLJK DFURVV DOO FRQGLWLRQV VXJJHVWLQJ WKHLU EHKDYLRU PD\ DOVR KDYH EHHQ PDLQWDLQHG E\ DXWRPDWLF UHLQIRUFHPHQW 7ZR SDUWLFLSDQWV H[KLELWHG WKH KLJKHVW OHYHOV RI 6,% GXULQJ GHPDQG VHVVLRQV VXJJHVWLQJ WKHLU EHKDYLRU ZDV PDLQWDLQHG E\ VRFLDOQHJDWLYH UHLQIRUFHPHQW DQG RQO\ SDUWLFLSDQW H[KLELWHG GLIIHUHQWLDOO\ KLJKHVW OHYHOV RI VHOILQMXU\ GXULQJ DWWHQWLRQ VHVVLRQV VXJJHVWLQJ KLV EHKDYLRU ZDV PDLQWDLQHG E\ DFFHVV WR DWWHQWLRQ $OWKRXJK IXQFWLRQDO DQDO\VLV SURYLGHV LQIRUPDWLRQ IRU HIIHFWLYH WUHDWPHQW GHYHORSPHQW IHZ UHVHDUFKHUV KDYH FRQGXFWHG IXQFWLRQDO DQDO\VHV SULRU WR WUHDWPHQW VHOHFWLRQ IRU +0 HJ 0D]DOHVNL ,ZDWD 5RGJHUV t =DUFRQH 9ROOPHU 0DUFXV t /H%ODQF /HUPDQ t ,ZDWD f ,Q DGGLWLRQ RXW RI WKH VWXGLHV WKDW KDYH LQFOXGHG D IXQFWLRQDO DQDO\VLV SULRU WR WUHDWPHQW VPDOO VDPSOH VL]HV ZHUH XVHG OLPLWLQJ WKH JHQHUDOLW\ RI WKHLU ILQGLQJV 7KH VWXG\ ZLWK WKH ODUJHVW VDPSOH VL]H WR GDWH ZDV FRQGXFWHG E\ *RK HW DO f ZKR FRQGXFWHG IXQFWLRQDO DQDO\VHV RI LQGLYLGXDOV ZKR H[KLELWHG +0 5HVXOWV LQGLFDWHG WKDW LQ RI WKH SDUWLFLSDQWVf bf +0 ZDV HLWKHU Df KLJKHU LQ WKH DORQH FRQGLWLRQ UHODWLYH WR WKH RWKHU WHVW FRQGLWLRQV RU Ef QRW GLIIHUHQWLDOO\ KLJK GXULQJ HLWKHU WKH DWWHQWLRQ RU GHPDQG FRQGLWLRQV 7KXV LQ D ODUJH PDMRULW\ RI WKH FDVHV

PAGE 12

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t &ULPPLQV 5HSS 6LQJK 2OLQJHU t 2OVRQ f )RU H[DPSOH LQ D ODUJHVFDOH VWXG\ EDVHG RQ IXQFWLRQDO DQDO\VHV ,ZDWD HW DO Ef UHSRUWHG WKDW WKH GXUDWLRQ RI DVVHVVPHQW UDQJHG IURP WR VHVVLRQV KU WR KUf ,Q WKH VWXG\ QRWHG DERYH E\ *RK HW DO f IXQFWLRQDO DQDO\VHV DYHUDJHG VHVVLRQV KRXUVf $ QXPEHU RI SURFHGXUDO YDULDWLRQV KDYH EHHQ GHVLJQHG WR LQFUHDVH WKH HIILFLHQF\ ZLWK ZKLFK IXQFWLRQDO DQDO\VHV FDQ EH FRQGXFWHG 2QH VXFK YDULDWLRQ LV WKH fEULHI IXQFWLRQDO DQDO\VLVf &RRSHU HW DO 'HUE\ HW DO 1RUWKXS HW DO :DFNHU HW DO f 'HUE\ HW DO DVVHVVHG WKH HIIHFWLYHQHVV RI WKH EULHI IXQFWLRQDO DQDO\VLV IRU SDUWLFLSDQWV (DFK SDUWLFLSDQW ZDV HYDOXDWHG GXULQJ D PLQ SHULRG DQG UHVXOWV VKRZHG WKDW WKH WDUJHW EHKDYLRU ZDV REVHUYHG DQG FOHDU GLIIHUHQWLDWLRQ LQ WKH GDWD SDWWHUQV HPHUJHG IRU b RI WKH SDUWLFLSDQWV ,Q D PRUH GLUHFW HYDOXDWLRQ RI WKH EULHI IXQFWLRQDO DQDO\VLV .DKQJ DQG ,ZDWD f FRPSDUHG WKH RXWFRPHV IURP IXQFWLRQDO DQDO\VHV FRQVLVWLQJ RI EULHI VLQJOHf DQG H[WHQGHG H[SRVXUHV WR WHVW DQG FRQWURO FRQGLWLRQV DQG IRXQG WKDW LQWHUSUHWDWLRQV IURP WKH EULHI DQG H[WHQGHG DQDO\VHV PDWFKHG IRU b RI WKH LQGLYLGXDOV 7KH GHYHORSPHQW RI WKH EULHI IXQFWLRQDO DQDO\VLV KDV EHHQ D VLJQLILFDQW FRQWULEXWLRQ WR WKH ILHOG RI EHKDYLRU DQDO\VLV E\ RIIHULQJ DQ HIILFLHQW DQG REMHFWLYH

PAGE 13

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f FRPSDUHG WKH UHVXOWV REWDLQHG IURP IXQFWLRQDO DQDO\VHV EDVHG RQ YDU\LQJ VHVVLRQ GXUDWLRQV LH PLQ PLQ DQG PLQf )XQFWLRQDO DQDO\VHV LQYROYLQJ DWWHQWLRQ GHPDQG DORQH WDQJLEOH DQG SOD\ FRQGLWLRQV ZHUH FRQGXFWHG LQ PXOWLHOHPHQW GHVLJQV )RUW\VL[ IXQFWLRQDO DQDO\VHV EDVHG RQ PLQ VHVVLRQV ZHUH XVHG WR JHQHUDWH VKRUWHU GDWD VHWV E\ VXEWUDFWLQJ WKH ODVW PLQ IRU WKH PLQ GXUDWLRQf DQG WKH ODVW PLQ IRU WKH PLQ GXUDWLRQf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

PAGE 14

PD\ JUHDWO\ UHGXFH WKH DPRXQW RI WLPH UHTXLUHG WR FRQGXFW DVVHVVPHQWV LQYROYLQJ UHSHDWHG H[SRVXUH WR WHVW FRQGLWLRQV $QRWKHU PHWKRG IRU LQFUHDVLQJ WKH HIILFLHQF\ RI IXQFWLRQDO DQDO\VHV LQYROYHV SURJUHVVLQJ IURP D EULHI IXQFWLRQDO DQDO\VLV WR D PRUH H[WHQGHG DQDO\VLV LI QHHGHG 9ROOPHU 0DUFXV 5LQJGDKO DQG 5RDQH f GHYHORSHG D PRGHO IRU FRQGXFWLQJ IXQFWLRQDO DQDO\VHV XVLQJ FRQGLWLRQV WKDW ZHUH EDVHG RQ WKRVH GHVFULEHG E\ ,ZDWD HW DO Df ,Q WKLV PRGHO D EULHI IXQFWLRQDO DQDO\VLV ZDV ILUVW FRQGXFWHG 3KDVH ,f LI GLIIHUHQWLDWHG UHVXOWV ZHUH REWDLQHG WKH DVVHVVPHQW ZDV FRPSOHWH ,I GLIIHUHQWLDWHG UHVXOWV ZHUH QRW REWDLQHG D PRUH H[WHQGHG PXOWLHOHPHQW IRUPDW ZDV XVHG 3KDVH ,,f ,I GLIIHUHQWLDWHG UHVXOWV ZHUH REVHUYHG LQ WKH PRUH H[WHQGHG DQDO\VLV WKH DVVHVVPHQW ZDV FRPSOHWH ,I XQGLIIHUHQWLDWHG UHVXOWV RFFXUUHG DQ H[WHQGHG QRLQWHUDFWLRQ FRQGLWLRQ ZDV FRQGXFWHG WR GHWHUPLQH ZKHWKHU EHKDYLRU SHUVLVWHG LQ WKH DEVHQFH RI VRFLDO FRQWLQJHQFLHV 3KDVH ,,,f ,I EHKDYLRU RFFXUUHG DW KLJK OHYHOV GXULQJ WKLV FRQGLWLRQ WKH DVVHVVPHQW ZDV FRPSOHWH LI EHKDYLRU H[WLQJXLVKHG GXULQJ WKH H[WHQGHG QRLQWHUDFWLRQ FRQGLWLRQ LQGLFDWLQJ WKDW EHKDYLRU PD\ EH VHQVLWLYH WR VRFLDO UHLQIRUFHPHQW WKH IXQFWLRQDO DQDO\VLV FRQGLWLRQV ZHUH SUHVHQWHG LQ D UHYHUVDO GHVLJQ 3KDVH ,9f 7KH IXQFWLRQ RI SDUWLFLSDQWVf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

PAGE 15

UHLQIRUFHPHQW ZH FRQGXFWHG D VHULHV RI DORQH VHVVLRQV WKDW ZHUH LQWHUVSHUVHG ZLWK DWWHQWLRQ DQG GHPDQG SUREHV ,I EHKDYLRU SHUVLVWHG GXULQJ WKH DORQH FRQGLWLRQ ZLWK ORZHU OHYHOV LQ WKH DWWHQWLRQ DQG GHPDQG FRQGLWLRQV WKLV ZRXOG YHULI\ RXU K\SRWKHVLV WKDW WKH SDUWLFLSDQWf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f )RU H[DPSOH *RK HW DO f SUHVHQWHG OHLVXUH LWHPV WR SDUWLFLSDQWV ZKR H[KLELWHG +0 PDLQWDLQHG E\ DXWRPDWLF UHLQIRUFHPHQW 5HVXOWV LQGLFDWHG WKDW 1&5 UHVXOWHG LQ UHGXFHG OHYHOV RI KDQG PRXWKLQJ DQG KLJK OHYHOV RI LWHP PDQLSXODWLRQ IRU RI WKH SDUWLFLSDQWV 7KHVH ILQGLQJV VXJJHVWHG WKDW VWLPXODWLRQ IURP WKH OHLVXUH LWHPV PLJKW KDYH VXEVWLWXWHG IRU WKH VWLPXODWLRQ IURP SDUWLFLSDQWVf +0 6KRUH HW DO f GHPRQVWUDWHG WKH HIIHFWLYHQHVV RI 1&5 LQ WKUHH LQGLYLGXDOV ZKR H[KLELWHG 6,% LQ WKH IRUP RI DUP UXEELQJ SDUWLFLSDQWf RU KDQG PRXWKLQJ SDUWLFLSDQWVf )XQFWLRQDO DQDO\VHV LQGLFDWHG WKDW 6,% RFFXUUHG PRVW RIWHQ LQ WKH DEVHQFH RI VRFLDO FRQWLQJHQFLHV LH WKH DORQH FRQGLWLRQf VXJJHVWLQJ SDUWLFLSDQWVf 6,% ZDV PDLQWDLQHG E\ DXWRPDWLF UHLQIRUFHPHQW 'XULQJ EDVHOLQH ZKHQ QR LQWHUDFWLRQ RU PDWHULDOV ZHUH GHOLYHUHG 6,% RFFXUUHG DW PRGHUDWHWRKLJK OHYHOV GXULQJ 1&5 ZKHQ SUHIHUUHG VWLPXOL EDVHG RQ D VWLPXOXV SUHIHUHQFH DVVHVVPHQWf ZHUH FRQWLQXRXVO\ DYDLODEOH 6,% ZDV JUHDWO\ UHGXFHG DQG LWHP FRQWDFW RFFXUUHG DW KLJK OHYHOV 7KHVH

PAGE 16

ILQGLQJV DOVR VXJJHVWHG WKDW WKH VWLPXODWLRQ IURP WKH OHLVXUH LWHPV PD\ KDYH FRPSHWHG ZLWK WKDW SURGXFHG E\ SDUWLFLSDQWVf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f DQGRU SURPSWLQJ KDYH EHHQ XVHG WR GHFUHDVH VWHUHRW\S\ DQG LQFUHDVH REMHFW PDQLSXODWLRQ )DYHOO 0F*LPVH\ DQG 6FKHOO f IRXQG WKDW '5$ HIIHFWLYHO\ UHGXFHG SDUWLFLSDQWVf SUREOHP EHKDYLRUV +0 H\H SRNLQJ DQGRU SLFDf DQG LQFUHDVHG DSSURSULDWH LWHP PDQLSXODWLRQ $IWHU 1&5 ZDV IRXQG LQHIIHFWLYH LQ VXSSUHVVLQJ +0 WR ORZ OHYHOV DQG LQFUHDVLQJ DGDSWLYH EHKDYLRUV HJ DSSURSULDWH VRFLDO LQWHUDFWLRQ DQG LWHP LQWHUDFWLRQf WR KLJK OHYHOV +RPHU f DVVHVVHG WKH HIIHFWLYHQHVV RI '5$ DQG REVHUYHG IXUWKHU GHFUHDVHV LQ 6,% DQG LQFUHDVHV LQ DGDSWLYH EHKDYLRUV 6LQJK DQG 0LOOLFKDPS f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f DVVHVVHG WKH HIIHFWV RI UHVSRQVH EORFNLQJ RQ RQH SDUWLFLSDQWfV

PAGE 17

+0 WKDW ZDV PDLQWDLQHG E\ DXWRPDWLF UHLQIRUFHPHQW %DVHOLQH DQG UHVSRQVH EORFNLQJ VHVVLRQV ZHUH FRQGXFWHG LQ D UHYHUVDO GHVLJQ %DVHOLQH VHVVLRQV ZHUH LGHQWLFDO WR WKH DORQH FRQGLWLRQ RI WKH IXQFWLRQDO DQDO\VLV LQ WKDW QR LQWHUDFWLRQ ZDV GHOLYHUHG 'XULQJ UHVSRQVH EORFNLQJ VHVVLRQV WKH WKHUDSLVW PDQXDOO\ EORFNHG VRPH RU DOO DWWHPSWV DW +0 5HVXOWV LQGLFDWHG WKDW ZKHQ +0 ZDV EORFNHG +0 DWWHPSWV GHFUHDVHG WR QHDU]HUR OHYHOV /LQGEHUJ ,ZDWD DQG .DKQJ f DVVHVVHG WKH HIIHFWLYHQHVV RI UHVSRQVH EORFNLQJ ZLWK WZR LQGLYLGXDOV ZKR H[KLELWHG VHYHUDO GLIIHUHQW WRSRJUDSKLHV RI 6,% KHDG RU ERG\ KLWWLQJ UXEELQJ ILQJHU SUHVVLQJ DJDLQVW IDFHf 5HVXOWV RI D IXQFWLRQDO DQDO\VLV LQGLFDWHG WKDW WKHLU EHKDYLRU ZDV QRW PDLQWDLQHG E\ VRFLDO UHLQIRUFHPHQW 7KH DXWKRUV LPSOHPHQWHG D SURJUHVVLYH VHULHV RI WUHDWPHQW FRPSRQHQWV 1&5 1&5 ZLWK SURPSWLQJ 1&5 DQG '5$ ZLWK SURPSWLQJ 1&5 DQG '5$ ZLWK SURPSWLQJ DQG UHVSRQVH EORFNLQJf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f DVVHVVHG WKH HIIHFWLYHQHVV RI EULHI PDQXDO UHVWUDLQW LQ WKUHH LQGLYLGXDOV ZLWK SURIRXQG PHQWDO UHWDUGDWLRQ 5HVXOWV RI D IXQFWLRQDO DQDO\VLV LQGLFDWHG WKDW SDUWLFLSDQWVf +0 ZDV QRW PDLQWDLQHG E\ VRFLDO

PAGE 18

UHLQIRUFHPHQW 1&5 DQG 1&5 ZLWK UHVWUDLQW VHVVLRQV ZHUH FRQGXFWHG LQ D UHYHUVDO GHVLJQ 'XULQJ 1&5 ZKHQ OHLVXUH LWHPV ZHUH FRQWLQXRXVO\ DYDLODEOH +0 RFFXUUHG DW KLJK OHYHOV DQG LWHP FRQWDFW RFFXUUHG DW ORZ OHYHOV GXULQJ 1&5 ZLWK UHVWUDLQW ZKHQ WKH WKHUDSLVW SODFHG WKH SDUWLFLSDQWfV KDQG LQ KHUKLV ODS IRU V IROORZLQJ HDFK +0 DWWHPSW +0 GHFUHDVHG WR ORZ OHYHOV DQG LWHP FRQWDFW LQFUHDVHG WR KLJK OHYHOV 7KRPSVRQ ,ZDWD &RQQHUV DQG 5RVFRH f GHPRQVWUDWHG WKH LPSRUWDQFH RI LQFOXGLQJ SUHIHUUHG LWHPV ZKHQ XVLQJ PDQXDO UHVWUDLQW SURFHGXUHV $IWHU FRQGXFWLQJ D IXQFWLRQDO DQDO\VLV RI SDUWLFLSDQWVf 6,% KDQG PRXWKLQJ KHDG DQG ERG\ KLWWLQJf WR UXOH RXW EHKDYLRUDO PDLQWHQDQFH E\ VRFLDO FRQWLQJHQFLHV DQ DVVHVVPHQW ZDV FRQGXFWHG WR LGHQWLI\ SDUWLFLSDQWVf SUHIHUUHG OHLVXUH LWHPV 1H[W WKH HIIHFWV RI EULHI PDQXDO UHVWUDLQW ZHUH HYDOXDWHG LQ UHYHUVDO RU PXOWLSOH EDVHOLQH GHVLJQV 'XULQJ EDVHOLQH DQG PDQXDO UHVWUDLQW FRQGLWLRQV SDUWLFLSDQWV ZHUH DOVR H[SRVHG WR UHLQIRUFHPHQW DQG QRUHLQIRUFHPHQW FRQGLWLRQV ZKLFK ZHUH DOWHUQDWHG LQ D PXOWLHOHPHQW GHVLJQ %DVHOLQH VHVVLRQV ZHUH LGHQWLFDO WR WKH QRLQWHUDFWLRQ FRQGLWLRQ WKH WKHUDSLVW GLG QRW XVH PDQXDO UHVWUDLQW GXULQJ WKLV FRQGLWLRQf RI WKH IXQFWLRQDO DQDO\VLV 'XULQJ PDQXDO UHVWUDLQW VHVVLRQV WKH WKHUDSLVW KHOG D SDUWLFLSDQWfV KDQGV LQ KHUKLV ODS RU DFURVV KHUKLV FKHVW IRU V FRQWLQJHQW XSRQ 6,% DWWHPSWV 'XULQJ QRUHLQIRUFHPHQW VHVVLRQV WKH SDUWLFLSDQW GLG QRW KDYH DFFHVV WR OHLVXUH LWHPV GXULQJ UHLQIRUFHPHQW VHVVLRQV SDUWLFLSDQWV KDG FRQWLQXRXV DFFHVV WR WKHLU WRS SUHIHUUHG OHLVXUH LWHP LGHQWLILHG LQ WKH SUHIHUHQFH DVVHVVPHQW 5HVXOWV LQGLFDWHG WKDW DOWKRXJK PDQXDO UHVWUDLQW ZLWK QR UHLQIRUFHPHQW UHVXOWHG LQ GHFUHDVHV RI SDUWLFLSDQWVf SUREOHP EHKDYLRU PDQXDO UHVWUDLQW ZDV PRUH HIIHFWLYH ZKHQ FRPELQHG ZLWK UHLQIRUFHPHQW 7KHVH ILQGLQJV OHQG VXSSRUW IRU WKH XVH RI PDQXDO UHVWUDLQW ZKHQ UHLQIRUFHPHQWEDVHG LQWHUYHQWLRQV SURYH LQHIIHFWLYH ,Q DGGLWLRQ WKH ILQGLQJV RI

PAGE 19

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f LW LV LPSRUWDQW WR FRQGXFW IXQFWLRQDO DQDO\VHV WR UXOH RXW D SRWHQWLDO VRFLDO IXQFWLRQ 7KLUG UHVHDUFKHUV RIWHQ GLG QRW XVH V\VWHPDWLF SUHIHUHQFH DVVHVVPHQWV SULRU WR LPSOHPHQWLQJ 1&5 DQG '5$ WUHDWPHQW FRPSRQHQWV %HFDXVH WKH UHLQIRUFHU IRU EHKDYLRU PDLQWDLQHG E\ DXWRPDWLF UHLQIRUFHPHQW LV XQNQRZQ LW LV LPSRUWDQW WR LGHQWLI\ SRWHQW UHLQIRUFHUV WKDW PD\ HIIHFWLYHO\ FRPSHWH ZLWK WKH VHQVRU\ VWLPXODWLRQ WKH SUREOHP EHKDYLRU SURGXFHV )LQDOO\ UHVHDUFKHUV KDYH RIWHQ XVHG LQWUXVLYH LQWHUYHQWLRQV VXFK DV UHVWUDLQW DQG SURWHFWLYH HTXLSPHQW ZLWKRXW ILUVW DVVHVVLQJ WKH HIIHFWLYHQHVV RI UHLQIRUFHPHQWEDVHG LQWHUYHQWLRQV 7KH SUHVHQW VWXG\ DWWHPSWHG WR H[WHQG FXUUHQW UHVHDUFK RQ WKH WUHDWPHQW RI +0 DQG EHKDYLRU PDLQWDLQHG E\ DXWRPDWLF UHLQIRUFHPHQW LQ WKH IROORZLQJ ZD\V Df $ IXQFWLRQDO DQDO\VLV ZDV FRQGXFWHG WR UXOH RXW EHKDYLRUDO VHQVLWLYLW\ WR VRFLDO FRQWLQJHQFLHV Ef V\VWHPDWLF SUHIHUHQFH DVVHVVPHQWV ZHUH FRQGXFWHG WR LGHQWLI\ SUHIHUUHG OHLVXUH PDWHULDOV WKDW PD\ FRPSHWH ZLWK +0 DQG WR LGHQWLI\ HGLEOH LWHPV WKDW PD\ UHLQIRUFH DSSURSULDWH SOD\ EHKDYLRUV Ff D SURJUHVVLYH

PAGE 20

VHULHV RI LQWHUYHQWLRQV ZDV XVHG EHJLQQLQJ ZLWK 1&5 DQG JUDGXDOO\ LQWURGXFLQJ PRUH LQWUXVLYH FRPSRQHQWV VXFK DV UHVSRQVH EORFNLQJ DQG EULHI PDQXDO UHVWUDLQ RQ DQ DV QHHGHG EDVLV

PAGE 21

678'< 35(9$/(1&( $1' 5,6. 2) +$1' 0287+,1* 0HWKRG 3DUWLFLSDQWV $OO LQGLYLGXDOV OLYLQJ LQ WZR VWDWH UHVLGHQWLDO IDFLOLWLHV IRU SHUVRQV ZLWK GHYHORSPHQWDO GLVDELOLWLHV RQH LQ )ORULGD 1 f DQG RQH LQ 7HQQHVVHH 1 f SDUWLFLSDWHG LQ WKH LQLWLDO VFUHHQLQJ VHH EHORZf 3DUWLFLSDQWVf DJHV UDQJHG IURP WR \HDUV WKHLU OHYHOV RI PHQWDO UHWDUGDWLRQ UDQJHG IURP PLOG WR SURIRXQG DQG WKHLU YHUEDO VNLOOV ZHUH H[WUHPHO\ YDULDEOH EXW JHQHUDOO\ YHU\ OLPLWHG $OO RU VRPH VXEVHW RI WKHVH LQGLYLGXDOV SDUWLFLSDWHG LQ WKH SURFHGXUHV QRWHG EHORZ 5HVSRQVH 'HILQLWLRQ 'XULQJ DOO DVVHVVPHQW SURFHGXUHV +0 ZDV GHILQHG DV SODFHPHQW RI WKH KDQG SDVW WKH SODQH RI WKH OLSV RU DV UHSHWLWLYH FRQWDFW EHWZHHQ WKH KDQG DQG WKH PRXWK ,Q DGGLWLRQ VWDII PHPEHUV ZHUH WROG WKDW SDUWLFLSDQWVf +0 VKRXOG EH UHSHWLWLYH DQG FKURQLF WR EH LQFOXGHG WR H[FOXGH VRFLDOO\ DFFHSWDEOH IRUPV RI +0 VXFK DV ZKHQ HDWLQJf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

PAGE 22

+0 $V D UHVXOW RI WKLV LQLWLDO VFUHHQLQJ SDUWLFLSDQWV ZHUH UHSRUWHG WR HQJDJH LQ +0 7KHVH LQGLYLGXDOV SDUWLFLSDWHG VHH 7DEOH IRU WKHLU GHPRJUDSKLF LQIRUPDWLRQf LQ DOO VXEVHTXHQW DVVHVVPHQWV XQOHVV RWKHUZLVH QRWHG 7KH SURSRUWLRQ RI WKH SDUWLFLSDQW 7DEOH 'HPRJUDSKLF &KDUDFWHULVWLFV RI 3DUWLFLSDQW 6DPSOH 1f 1XPEHU 3HUFHQWDJH 6H[ 0DOH b )HPDOH b $JH b b b b b b 'HJUHH RI UHWDUGDWLRQ 0LOGPRGHUDWH b VHYHUH b SURIRXQG b ,PSDLUPHQWV YLVXDO b DXGLWRU\ b SK\VLFDO b 5HFHSWLYH YHUEDO VNLOOV QRQH b PLQLPDO b IROORZV GLUHFWLRQV b ([SUHVVLYH YHUEDO VNLOOV QRQH b PLQLPDO b RQHZRUG XWWHUDQFHV b 7ZRWRWHQZRUG XWWHUDQFHV b ZRUG XWWHUDQFHV b

PAGE 23

VDPSOH FODVVLILHG DV KDYLQJ SURIRXQG PHQWDO UHWDUGDWLRQ ZDV YHU\ KLJK bf DQG QRW UHSUHVHQWDWLYH RI WKH WRWDO SRSXODWLRQ RI )ROORZLQJ WKH LQLWLDO VFUHHQLQJ PRUH H[WHQVLYH LQWHUYLHZV ZHUH FRQGXFWHG WR JDWKHU LQIRUPDWLRQ DERXW WKH LQGLYLGXDOV LGHQWLILHG $ JUDGXDWH VWXGHQW PHW ZLWK D VWDII PHPEHU ZKR ZRUNHG ZLWK WKH SDUWLFLSDQW LQ DQ RIILFH RU D TXLHW URRP DQG DVNHG WKH VWDII PHPEHU VSHFLILFDOO\ ZKHWKHU RU QRW D SDUWLFLSDQW H[KLELWHG +0 ,I +0 ZDV UHSRUWHG WR RFFXU WKH LQWHUYLHZHU DOVR DVNHG VWDII WR HVWLPDWH ZKHWKHU +0 RFFXUUHG DW D ORZ UDWH OHVV WKDQ RQFH SHU KRXUf RU DW D KLJKHU UDWH PRUH WKDQ RQFH SHU KRXUf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f +0 2QH KXQGUHG DQG WZR RI WKH LQGLYLGXDOV UHSRUWHG WR HQJDJH LQ +0 GXULQJ WKH LQGLUHFW DVVHVVPHQW SDUWLFLSDWHG :H ZHUH XQDEOH WR REWDLQ FRQVHQW IRU RI WKH LQGLYLGXDOV 6HVVLRQV ZHUH FRQGXFWHG LQ WKHUDS\ URRPV ORFDWHG LQ D FOLQLF IRU WKH DVVHVVPHQW DQG WUHDWPHQW RI 6,% RU ZHUH FRQGXFWHG LQ TXLHW URRPV LQ SDUWLFLSDQWVf KRPHV ,I SDUWLFLSDQWV ZRUH SURWHFWLYH HTXLSPHQW WKH HTXLSPHQW ZDV UHPRYHG SULRU WR FRQGXFWLQJ REVHUYDWLRQV 7UDLQHG REVHUYHUV UHFRUGHG LQVWDQFHV RI +0 XVLQJ HLWKHU KDQGKHOG FRPSXWHUV $VVLVWDQW PRGHO $67 f RU D GDWD VKHHW VHJPHQWHG LQWR V ELQV 6HVVLRQV ZHUH HLWKHU RU PLQ DQG WKH WRWDO REVHUYDWLRQ ZDV PLQ 'DWD ZHUH VXPPDUL]HG DV WKH

PAGE 24

SHUFHQWDJH RI V LQWHUYDOV GXULQJ ZKLFK +0 RFFXUUHG 5HOLDELOLW\ ZDV DVVHVVHG E\ KDYLQJ D VHFRQG REVHUYHU FROOHFW GDWD LQGHSHQGHQWO\ GXULQJ DQ DYHUDJH RI b RI WKH VHVVLRQV 2EVHUYHUVf UHFRUGV ZHUH FRPSDUHG RQ DQ LQWHUYDOE\LQWHUYDO EDVLV DQG SHUFHQWDJH DJUHHPHQW ZDV FDOFXODWHG E\ GLYLGLQJ WKH QXPEHU RI DJUHHPHQWV E\ WKH QXPEHU RI DJUHHPHQWV SOXV GLVDJUHHPHQWV DQG PXOWLSO\LQJ E\ b 7KH PHDQ SHUFHQW DJUHHPHQW ZDV b UDQJH b WR bf 7KH PDMRULW\ RI VHVVLRQV UHVHPEOHG WKH DORQH FRQGLWLRQ RI WKH IXQFWLRQDO DQDO\VLV LH LQGLYLGXDOV KDG QR PDWHULDOV DQG QR LQWHUDFWLRQ ZDV SURYLGHGf $OVR WR UXOH RXW WKH SRVVLELOLW\ WKDW +0 PD\ QRW RFFXU GXULQJ DQ DORQH FRQGLWLRQ DV D UHVXOW RI EHKDYLRUDO PDLQWHQDQFH E\ VRFLDO FRQWLQJHQFLHV DW OHDVW RQH DWWHQWLRQ DQG DW OHDVW RQH GHPDQG FRQGLWLRQ ZHUH FRQGXFWHG VHH EHORZ IXQFWLRQDO DQDO\VLVf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b RI WKH LQWHUYDOV ZHUH FODVVLILHG DV KDYLQJ ORZUDWH +0 ZKHUHDV LQGLYLGXDOV ZKR H[KLELWHG +0 GXULQJ b RU PRUH RI WKH LQWHUYDOV ZHUH FODVVLILHG DV KDYLQJ KLJKUDWH +0 6HYHULW\ 7KH SDUWLFLSDQWV UHSRUWHG WR HQJDJH LQ +0 GXULQJ WKH LQWHUYLHZ SDUWLFLSDWHG 7KH VHWWLQJV ZHUH LGHQWLFDO WR WKRVH XVHG IRU WKH GLUHFW DVVHVVPHQW 'XULQJ WKH VHYHULW\ DVVHVVPHQW HLWKHU D EHKDYLRU DQDO\VW RU D JUDGXDWH VWXGHQW H[DPLQHG D SDUWLFLSDQWfV KDQGV DQG UDWHG WKH H[WHQW RI LQMXU\ RQ D SRLQW VFDOH DFFRUGLQJ

PAGE 25

WR WKH IROORZLQJ FULWHULD Df $ ff ZDV VFRUHG LI QR FXWV EUXLVHV RU VZHOOLQJ ZHUH REVHUYHG Ef D ff ZDV VFRUHG LI UHGGHQHG DUHDV EXW QR FXWV EUXLVHV RU VZHOOLQJ ZHUH SUHVHQW Ff D ff ZDV VFRUHG LI WKHUH ZDV FXUUHQW LQMXU\ VXFK DV VFDEV VZHOOLQJ RU FUDFNV DQG Gf D ff ZDV VFRUHG LI WKH SDUWLFLSDQW ZRUH SURWHFWLYH HTXLSPHQW SUHVFULEHG WR SUHYHQW +0 +0 ZDV FODVVLILHG DV LQMXULRXV LI D VFRUH RI D RU D ZDV UHFRUGHG $ VHFRQG REVHUYHU FROOHFWHG GDWD LQGHSHQGHQWO\ RQ DQ DYHUDJH RI b RI WKHVH DVVHVVPHQWV ,Q FRPSDULQJ REVHUYHUVf UHFRUGV DQ DJUHHPHQW ZDV GHILQHG DV ERWK REVHUYHUV KDYLQJ UHFRUGHG WKH VDPH VFRUH IRU D JLYHQ SDUWLFLSDQW ,QWHUREVHUYHU DJUHHPHQW ZDV FDOFXODWHG E\ GLYLGLQJ DJUHHPHQWV E\ DJUHHPHQWV SOXV GLVDJUHHPHQWV DQG PXOWLSO\LQJ E\ b 7KH PHDQ SHUFHQWDJH DJUHHPHQW DFURVV SDUWLFLSDQWV ZDV b UDQJH b WR bf 5HVXOWV DQG 'LVFXVVLRQ 7DEOH VKRZV WKH UHVXOWV IURP WKH LQGLUHFW DVVHVVPHQW 2XW RI WKH HQWLUH SRSXODWLRQ LQGLYLGXDOVf bf ZHUH UHSRUWHG E\ VWDII WR HQJDJH LQ +0 DQG bf ZHUH UHSRUWHG WR HQJDJH LQ KLJKUDWH +0 JUHDWHU WKDQ RQFH SHU KRXUf $OWKRXJK WKH UHSRUWHG SUHYDOHQFH RI +0 ZDV PXFK OHVV WKDQ WKDW UHSRUWHG E\ 7URVWHU HW DO f LQVWDQFHV RI KLJK UDWH +0 ZHUH FRPSDUDEOH WR WKRVH UHSRUWHG E\ 7URVWHU HW DO 2XW RI WKH LQGLYLGXDOV UHSRUWHG WR HQJDJH LQ +0 GLUHFW REVHUYDWLRQV ZHUH FRQGXFWHG IRU 7R HVWLPDWH WKH QXPEHU RI LQGLYLGXDOV ZKR H[KLELWHG +0 RXW RI WKH ODUJHU VDPSOH VL]H Q f ZH ILUVW GHWHUPLQHG WKH SURSRUWLRQ RI LQGLYLGXDOV ZKR ZHUH REVHUYHG WR H[KLELW +0 RXW RI WKH VPDOOHU VDPSOH Q f 6L[W\IRXU RXW RI WKH RU b ZHUH REVHUYHG WR HQJDJH LQ +0 7KXV WKH HVWLPDWHG QXPEHU RI LQGLYLGXDOV ZKR H[KLELWHG +0 RXW RI WKH ODUJHU VDPSOH RI ZDV b RU 2XW RI WKH WRWDO SRSXODWLRQ RI LQGLYLGXDOV WKH HVWLPDWHG QXPEHU WKDW HQJDJHG LQ +0 ZDV RU b

PAGE 26

:H XVHG WKH VDPH PHWKRG WR HVWLPDWH WKH QXPEHU RI LQGLYLGXDOV ZKR HQJDJHG LQ KLJK UDWH +0 RU b RI WKH WRWDO SRSXODWLRQ RI f 7KH SHUFHQWDJH RI KLJK UDWH +0 JLYHQ WKDW +0 ZDV REVHUYHG ZDV b 7KXV UHVXOWV REWDLQHG IURP GLUHFW REVHUYDWLRQ \LHOGHG ORZHU ILJXUHV WKDQ WKRVH REWDLQHG IURP WKH LQGLUHFW DVVHVVPHQW IRU RYHUDOO SUHYDOHQFH DV ZHOO DV IRU SUHYDOHQFH RI KLJKUDWH +0 7KHVH ILQGLQJV DOVR VXJJHVW WKDW WKH PDMRULW\ RI LQGLYLGXDOV ZKR H[KLELW +0 HQJDJH LQ WKLV EHKDYLRU DW D KLJK UDWH 7DEOH ([SHULPHQW 3UHYDOHQFH )UHTXHQF\ DQG 6HYHULW\ 5HVXOWV 1XPEHU 3URSRUWLRQ ,QGLUHFW $VVHVVPHQW 1 f 2YHUDOO SUHYDOHQFH b +LJK UDWH +0 b 'LUHFW 2EVHUYDWLRQ 1 rf 2YHUDOO SUHYDOHQFH b +LJK UDWH +0 b +LJK UDWH +0$ REVHUYHG +0 b 6HYHULWY 5DWLQJV 1 2YHUDOO SUHYDOHQFH RI VHYHUH +0 b 6HYHUH +0$ REVHUYHG +0 b r 1XPEHUV UHIOHFW HVWLPDWHV IRU WKH LQGLYLGXDOV UHSRUWHG WR HQJDJH LQ +0 EDVHG RQ GLUHFW REVHUYDWLRQ RI RI WKHVH LQGLYLGXDOV $OO LQGLYLGXDOV UHSRUWHG WR HQJDJH LQ +0 UHFHLYHG D VHYHULW\ PHDVXUH WKURXJK GLUHFW REVHUYDWLRQ 7ZHQW\QLQH LQGLYLGXDOV RU b RI WKH WRWDO SRSXODWLRQ f ZHUH REVHUYHG WR KDYH VHYHUH +0 LH D RU D VHYHULW\ VFRUHf 7KH SURSRUWLRQ RI LQGLYLGXDOV ZKR HQJDJHG LQ VHYHUH +0 +0 WKDW SURGXFHG REVHUYDEOH WLVVXH GDPDJH RU UHTXLUHG SURWHFWLYH HTXLSPHQW WR SUHYHQW VXFK GDPDJHf JLYHQ WKDW +0 ZDV REVHUYHG ZDV b $ VWUHQJWK RI WKLV LQYHVWLJDWLRQ ZDV WKDW ERWK LQWHUYLHZ DQG GLUHFW REVHUYDWLRQ PHWKRGV ZHUH XVHG WR GHWHUPLQH WKH SUHYDOHQFH RI +0 5HVXOWV RI D UHFHQW VWXG\ E\

PAGE 27

0F*LOO +XJKHV 7HHU DQG 5\H f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

PAGE 28

(;3(5,0(17 )81&7,21$/ $1$/<6,6 2) +$1' 0287+,1* 0HWKRG 3DUWLFLSDQWV DQG 6HWWLQJ $OO LQGLYLGXDOV ZKR ZHUH GLUHFWO\ REVHUYHG WR HQJDJH LQ +0 1 f GXULQJ ([SHULPHQW SDUWLFLSDWHG LQ ([SHULPHQW 2EVHUYDWLRQV ZHUH FRQGXFWHG LQ D TXLHW URRP LQ WKH LQGLYLGXDOfV KRPH RU DW D GD\ SURJUDP IRU WKH DVVHVVPHQW DQG WUHDWPHQW RI WKHLU SUREOHP EHKDYLRUV 'DWD &ROOHFWLRQ DQG 5HOLDELOLW\ +0 ZDV GHILQHG DV HLWKHU FRQWDFW EHWZHHQ WKH KDQGILQJHUV DQG WKH PRXWK RU LQVHUWLRQ RI WKH ILQJHUV SDVW WKH SODQH RI WKH OLSV 7UDLQHG REVHUYHUV UHFRUGHG WKH SUHVHQFH RU DEVHQFH RI +0 GXULQJ FRQWLQXRXV V LQWHUYDOV XVLQJ D SHQFLO DQG VFRULQJ VKHHW RU D KDQGKHOG FRPSXWHU $VVLVWDQW PRGHO $67 f 'DWD ZHUH VXPPDUL]HG DV WKH SHUFHQWDJH RI LQWHUYDOV GXULQJ ZKLFK +0 RFFXUUHG ,QWHUREVHUYHU DJUHHPHQW ZDV DVVHVVHG GXULQJ D PHDQ RI b RI WKH VHVVLRQV 3HUFHQWDJH DJUHHPHQW VFRUHV ZHUH FDOFXODWHG EDVHG RQ DQ LQWHUYDOE\ LQWHUYDO FRPSDULVRQ RI WKH REVHUYHUVf UHFRUGV E\ GLYLGLQJ WKH QXPEHU RI DJUHHPHQWV E\ WKH WRWDO QXPEHU RI LQWHUYDOV DQG PXOWLSO\LQJ E\ b 0HDQ DJUHHPHQW ZDV b UDQJH b WR bf )XQFWLRQDO $QDO\VLV 3URFHGXUHV EDVHG RQ WKRVH GHVFULEHG E\ ,ZDWD HW DO Df ZHUH FRQGXFWHG WR LGHQWLI\ WKH IXQFWLRQDO SURSHUWLHV RI +0 DQG ZHUH PRGLILHG LQ VHYHUDO ZD\V WR LQFUHDVH WKH HIILFLHQF\ RI DVVHVVPHQW )LUVW LQ OLJKW RI GDWD UHSRUWHG E\ :DOODFH DQG ,ZDWD f

PAGE 29

LQGLFDWLQJ WKDW UHVXOWV IURP RU PLQ VHVVLRQV JHQHUDOO\ FRUUHVSRQG WR WKRVH REWDLQHG IURP PLQ VHVVLRQV DOO IXQFWLRQDO DQDO\VLV VHVVLRQV ZHUH HLWKHU RU PLQ LQ GXUDWLRQ ,Q DGGLWLRQ EDVHG RQ WKH *RK HW DO f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f DVVHVVPHQW ZDV WHUPLQDWHG +RZHYHU LI UHVXOWV ZHUH XQFOHDU DVVHVVPHQW ZDV H[WHQGHG XQWLO D FOHDU IXQFWLRQ HPHUJHG 7KH VSHFLILF FRQGLWLRQV H[WHQGHG GHSHQGHG XSRQ WKH SDWWHUQ RI UHVSRQGLQJ REWDLQHG GXULQJ WKH LQLWLDO VHVVLRQ DVVHVVPHQW )RU H[DPSOH LI UHVXOWV ZHUH XQGLIIHUHQWLDWHG ZLWK VOLJKWO\ ORZHU OHYHOV RI +0 LQ WKH DORQH FRQGLWLRQ UHODWLYH WR WKH GHPDQG FRQGLWLRQ DORQH DQG GHPDQG FRQGLWLRQV ZHUH H[WHQGHG DQG DOWHUQDWHG LQ D UDWLR +RZHYHU LI UHVXOWV ZHUH XQGLIIHUHQWLDWHG ZLWK VOLJKWO\ ORZHU OHYHOV RI +0 LQ WKH DORQH FRQGLWLRQ UHODWLYH WR DWWHQWLRQ DQG GHPDQG FRQGLWLRQV WKHQ DQ H[WHQGHG PXOWLHOHPHQW IXQFWLRQDO DQDO\VLV LQFOXGLQJ DORQH DWWHQWLRQ SOD\ DQG GHPDQG FRQGLWLRQVf ZDV FRQGXFWHG 5HVXOWV DQG 'LVFXVVLRQ )LJXUH VKRZV UHVXOWV RI WKH IXQFWLRQDO DQDO\VLV IRU SDUWLFLSDQWV (DFK SDUWLFLSDQW UHSUHVHQWV D SDWWHUQ RI UHVSRQGLQJ W\SLFDOO\ REVHUYHG ,Q WKH ILUVW SDWWHUQ +0 RFFXUUHG DW KLJK OHYHOV RQO\ LQ WKH DORQH FRQGLWLRQ ZLWK PXFK ORZHU OHYHOV RFFXUULQJ GXULQJ WKH DWWHQWLRQ DQG GHPDQG FRQGLWLRQV WRS SDQHOf 2XW RI LQGLYLGXDOV RU b H[KLELWHG UHVSRQGLQJ FKDUDFWHULVWLF RI WKLV SDWWHUQ ,Q WKH VHFRQG SDWWHUQ D KLJK

PAGE 30

3(5&(17$*( 2) ,17(59$/6 +$1' 0287+,1* !? a f f§ff§ $ORQH $WWHQWLRQ f§R 3OD\ B f§D 'HPDQG D 3DWWHUQ bf 3DWWHUQ bf )LJXUH 3HUFHQWDJH RI LQWHUYDOV KDQG PRXWKLQJ GXULQJ IXQFWLRQDO DQDO\VLV VHVVLRQV 7KH WRS JUDSK GHSLFWV SDWWHUQ WKH VHFRQG IURP WRS JUDSK GHSLFWV SDWWHUQ WKH VHFRQG IURP WKH ERWWRP JUDSK GHSLFWV SDWWHUQ DQG WKH ERWWRP JUDSK GHSLFWV SDWWHUQ 6HH WH[W IRU GHWDLOV

PAGE 31

OHYHO RI UHVSRQGLQJ RFFXUUHG DFURVV FRQGLWLRQV ZLWK VOLJKWO\ ORZHU OHYHOV LQ WKH DWWHQWLRQ DQG GHPDQG FRQGLWLRQV VHFRQG SDQHOf 2XW RI SDUWLFLSDQWV RU b H[KLELWHG UHVSRQGLQJ FKDUDFWHULVWLF RI WKLV SDWWHUQ ,Q WKH WKLUG SDWWHUQ +0 RFFXUUHG GXULQJ DOO FRQGLWLRQV EXW ZDV QRW GLIIHUHQWLDOO\ KLJK GXULQJ DQ\ FRQGLWLRQ WKLUG SDQHOf 2XW RI SDUWLFLSDQWV RU b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f LQGLFDWLQJ WKDW WKH EHKDYLRU ZDV PDLQWDLQHG E\ VRFLDO SRVLWLYH UHLQIRUFHPHQW LQ WKH IRUP RI DWWHQWLRQ ,Q VXPPDU\ RI RU b RI WKH IXQFWLRQDO DQDO\VHV UHYHDOHG SDWWHUQV LQGLFDWLQJ SDUWLFLSDQWVf +0 ZDV PDLQWDLQHG E\ DXWRPDWLF UHLQIRUFHPHQW ,Q ([SHULPHQW ZH XVHG DQ DEEUHYLDWHG IXQFWLRQDO DQDO\VLV GHVLJQ ZKRVH PRGLILFDWLRQV ZHUH EDVHG RQ SUHYLRXV UHVHDUFK :H K\SRWKHVL]HG WKDW WKH SDUWLFLSDQWVf +0 ZDV PRVW OLNHO\ PDLQWDLQHG E\ DXWRPDWLF UHLQIRUFHPHQW DQG DV D UHVXOW FRQGXFWHG DORQH VHVVLRQV SULPDULO\ ZKLFK ZHUH LQWHUVSHUVHG ZLWK DWWHQWLRQ DQG GHPDQG VHVVLRQV 7KLV GHVLJQ SURYHG XVHIXO LQ LGHQWLI\LQJ D EHKDYLRUDO IXQFWLRQ IRU DOO SDUWLFLSDQWV LQ D UHODWLYHO\ EULHI DPRXQW RI WLPH 7KH IXQFWLRQDO DQDO\VLV PHWKRG XVHG LQ WKLV VWXG\ H[WHQGV SUHYLRXV UHVHDUFK RQ EULHI IXQFWLRQDO DQDO\VHV 9ROOPHU 0DUFXV 5LQJGDKO DQG 5RDQH f DVVHVVHG WKH

PAGE 32

HIIHFWLYHQHVV RI D EULHI IXQFWLRQDO DQDO\VLV LQ ZKLFK WR VHVVLRQV ZHUH FRQGXFWHG IRU HDFK FRQGLWLRQ DQG ZLWKLQVHVVLRQ UHVSRQVH SDWWHUQV ZHUH DQDO\]HG IRU HDFK RI WKH VHVVLRQV WR GHWHUPLQH IXQFWLRQ $VVHVVPHQWV WRRN WR KU DQG b RI WKH DVVHVVPHQWV \LHOGHG GLIIHUHQWLDWHG UHVXOWV GXULQJ WKLV EULHI DQDO\VLV 7KH SUHVHQW VWXG\ GLIIHUV IURP WKDW XVHG E\ 9ROOPHU HW DO LQ WKDW ZH XVHG UHSHDWHG PHDVXUHV RI UHVSRQGLQJ LQ DOO FRQGLWLRQV ZLWK SURSRUWLRQDOO\ PRUH DORQH VHVVLRQV UHODWLYH WR DWWHQWLRQ DQG GHPDQG VHVVLRQV +RZHYHU WKH WRWDO GXUDWLRQ RI WKH VHVVLRQ DVVHVVPHQW XVHG LQ WKH SUHVHQW VWXG\ ZDV FRPSDUDEOH WR WKDW UHSRUWHG E\ 9ROOPHU HW DO KU PLQ ZKHQ VHVVLRQV ZHUH PLQ KU PLQ ZKHQ VHVVLRQV ZHUH PLQf ,Q DGGLWLRQ WKH VHVVLRQ DVVHVVPHQW \LHOGHG GLIIHUHQWLDWHG UHVXOWV IRU RI WKH RU bf +DQOH\ ,ZDWD DQG 7KRPSVRQ f XVHG DQRWKHU SURFHGXUDO YDULDWLRQ ZKHQ FRQGXFWLQJ IXQFWLRQDO DQDO\VHV ,Q WKHLU DQDO\VLV WZR FRQGLWLRQV ZHUH FRPSDUHG $ WHVW FRQGLWLRQ HLWKHU DWWHQWLRQ RU WDQJLEOHf LQ ZKLFK D VXVSHFWHG UHLQIRUFHU ZDV GHOLYHUHG FRQWLQJHQW XSRQ SUREOHP EHKDYLRU DQG D FRQWURO FRQGLWLRQ LQ ZKLFK WKH VDPH UHLQIRUFHU ZDV DYDLODEOH QRQFRQWLQJHQWO\ 7KHUH DUH VHYHUDO QRWHZRUWK\ IHDWXUHV RI WKH DXWKRUVf SURFHGXUDO YDULDWLRQ RI WKH W\SLFDO IXQFWLRQDO DQDO\VLV PHWKRG )LUVW WKH PHWKRG ZDV YHU\ HIILFLHQW WDNLQJ XQGHU KRXUV WR LPSOHPHQWf 6HFRQG WKH DVVHVVPHQW DOORZHG IRU UHSHDWHG PHDVXUHV RI WKH WHVW FRQGLWLRQ DVVHVVHG 7KLUG WKH VSHFLILF WHVW FRQGLWLRQ XVHG ZDV EDVHG RQ UHSRUWV IURP VWDII DQG UHVXOWV IURP VWUXFWXUHG REVHUYDWLRQV )RXUWK DOO SDUWLFLSDQWV ZHUH VXFFHVVIXOO\ WUHDWHG E\ LQWHUYHQWLRQV EDVHG RQ WKH UHVXOWV RI WKHLU IXQFWLRQDO DQDO\VHV 2XU PHWKRG H[WHQGHG WKH +DQOH\ HW DO PHWKRG E\ Df WHVWLQJ IRU DXWRPDWLF UHLQIRUFHPHQW SULPDULO\ DQG Ef E\ FRQGXFWLQJ SUREHV RI DOO WHVW FRQGLWLRQV WR UXOH RXW EHKDYLRUDO PDLQWHQDQFH E\ VRFLDO UHLQIRUFHPHQW

PAGE 33

$ SRWHQWLDO OLPLWDWLRQ RI RXU PHWKRG ZDV RXU VHOHFWLYH XVH RI WHVW FRQGLWLRQV ,QFOXGLQJ D UHODWLYHO\ VPDOO QXPEHU RI DWWHQWLRQ DQG GHPDQG VHVVLRQV PD\ KDYH OLPLWHG WKH VHQVLWLYLW\ RI WKH DQDO\VLV LQ GHWHFWLQJ VRFLDO UHLQIRUFHPHQW IXQFWLRQV )XWXUH UHVHDUFK FRXOG FRPSDUH UHVXOWV REWDLQHG WKURXJK WKH PRGLILHG GHVLJQ XVHG LQ WKLV VWXG\ ZLWK UHVXOWV REWDLQHG E\ WKH PRUH H[WHQGHG GHVLJQ XVHG E\ ,ZDWD HW DO Df

PAGE 34

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

PAGE 35

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f RU D GDWD VKHHW VHJPHQWHG LQWR V ELQV 'DWD ZDV VXPPDUL]HG DV WKH SHUFHQWDJH RI LQWHUYDOV GXULQJ ZKLFK +0 RFFXUUHG 5HOLDELOLW\ ZDV FDOFXODWHG E\ KDYLQJ D VHFRQG REVHUYHU FROOHFW GDWD LQGHSHQGHQWO\ RQ DQ DYHUDJH RI b RI WKH VHVVLRQV 2EVHUYHUVf UHFRUGV ZHUH FRPSDUHG RQ DQ LQWHUYDOE\LQWHUYDO EDVLV DQG SHUFHQWDJH DJUHHPHQW ZDV FDOFXODWHG E\ GLYLGLQJ WKH QXPEHU RI DJUHHPHQWV E\ WKH QXPEHU RI DJUHHPHQWV SOXV GLVDJUHHPHQWV DQG PXOWLSO\LQJ E\ b 7KH PHDQ SHUFHQW DJUHHPHQW IRU +0 ZDV b UDQJH b WR bf 3KDVH )XQFWLRQDO $QDO\VLV %HFDXVH DOO SDUWLFLSDQWV UHFHLYHG D IXQFWLRQDO DQDO\VLV LQ WKH SUHYLRXV VWXG\ ZH GLG QRW FRQGXFW DQRWKHU IXQFWLRQDO DQDO\VLV )XQFWLRQDO DQDO\VHV IURP H[SHULPHQW IRU LQGLYLGXDOV SDUWLFLSDWLQJ LQ H[SHULPHQW DUH VKRZQ LQ )LJXUH 'HDQQD *LQD 'DQ 7HG $QG\ DQG %HWK WRS URZVf H[KLELWHG KLJK OHYHOV RI +0 GXULQJ WKH DORQH FRQGLWLRQ DQG

PAGE 36

3(5&(17$*( 2) ,17(59$/6 6(66,216 )LJXUH 3HUFHQWDJH RI LQWHUYDOV +0 GXULQJ WKH IXQFWLRQDO DQDO\VLV IURP ([SHULPHQW IRU DOO ([SHULPHQW SDUWLFLSDQWV 6HH WH[W IRU GHWDLOV

PAGE 37

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f 'HERUDK -DPHV .HUL 0LNH DQG /LVD H[KLELWHG XQGLIIHUHQWLDWHG UHVSRQGLQJ WKDW ZDV QRW GLIIHUHQWLDOO\ KLJK LQ WKH DORQH FRQGLWLRQ UHTXLULQJ D PRUH H[WHQGHG DQDO\VLV ERWWRP WZR URZVf 'HERUDK -DPHV .HUL DQG /LVD H[KLELWHG KLJK OHYHOV RI UHVSRQGLQJ GXULQJ ERWK

PAGE 38

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f +0 GXULQJ 1&5 VHVVLRQV %HFDXVH DOO SDUWLFLSDQWV ZHUH H[SRVHG WR WKH 1&5 WUHDWPHQW DOO SDUWLFLSDQWV UHFHLYHG D OHLVXUH LWHP DVVHVVPHQW /HLVXUH LWHPV SUHVXPHG WR SURYLGH GLIIHUHQW W\SHV RI VHQVRU\ VWLPXODWLRQ WDFWLOH DXGLWRU\ ROIDFWRU\f ZHUH VHOHFWHG IRU XVH GXULQJ WKH OHLVXUH LWHP DVVHVVPHQW ,WHP FRQWDFW ZDV GHILQHG DV DQ\ FRQWDFW EHWZHHQ WKH SDUWLFLSDQWfV KDQG DQG WKH OHLVXUH LWHP 2EVHUYHUV UHFRUGHG WKH GXUDWLRQ RI LWHP FRQWDFW XVLQJ D VWRSZDWFK SDSHU DQG SHQFLO 3URFHGXUHV ZHUH EDVHG RQ WKRVH GHVFULEHG E\ 'H/HRQ ,ZDWD &RQQHUV DQG :DOODFH f 7HQWR LWHPV ZHUH VLQJO\ SUHVHQWHG (DFK LWHP ZDV SUHVHQWHG IRU D PLQ WULDO ZKLFK ZDV UHSHDWHG WLPHV IRU D WRWDO GXUDWLRQ RI PLQ SHU LWHP 'DWD ZHUH VXPPDUL]HG DV WKH WRWDO GXUDWLRQ RI LWHP FRQWDFW ,QWHUREVHUYHU DJJUHHPHQW ZDV DVVHVVHG GXULQJ D PHDQ RI b RI WKH WULDOV 0HDQ LQWHUREVHUYHU DJUHHPHQW IRU GXUDWLRQ RI LWHP FRQWDFW ZDV b UDQJH b WR bf

PAGE 39

)RRG ,WHP $VVHVVPHQW 7KH SXUSRVH RI WKH IRRG LWHP DVVHVVPHQW ZDV WR LGHQWLI\ D SUHIHUUHG HGLEOH WKDW PLJKW IXQFWLRQ DV D UHLQIRUFHU IRU VKDSLQJ LWHP FRQWDFW GXULQJ '5$ VHVVLRQV VHH EHORZf )LYH LQGLYLGXDOV ZKR FRQWLQXHG WR HQJDJH LQ KLJK OHYHOV RI +0 DQG ORZ OHYHOV RI LWHP FRQWDFW GXULQJ 1&5 VHVVLRQV VHH EHORZf UHFHLYHG D IRRG SUHIHUHQFH DVVHVVPHQW EHIRUH SDUWLFLSDWLQJ LQ WKH '5$ LQWHUYHQWLRQ 7KH DVVHVVPHQW SURFHGXUH ZDV VLPLODU WR WKDW GHVFULEHG E\ 3DFH ,YDQFLF (GZDUGV ,ZDWD DQG 3DJH f 7HQ HGLEOHV ZHUH SUHVHQWHG VLQJO\ WKUHH WLPHV HDFK 2EVHUYHUV UHFRUGHG ZKHWKHU RU QRW DQ DSSURDFK UHVSRQVH RFFXUUHG IRU HDFK WULDO RI IRRG SUHVHQWDWLRQ $SSURDFK UHVSRQVHV ZHUH GHILQHG DV WKH SDUWLFLSDQWfV SODFHPHQW RI WKH HGLEOH LQWR KHUKLV PRXWK :H DOWHUHG WKLV GHILQLWLRQ VOLJKWO\ IRU LQGLYLGXDOV ZKR FRXOG QRW HDW LQGHSHQGHQWO\ LH DSSURDFK ZDV GHILQHG DV VZDOORZLQJ WKH LWHP DIWHU LW ZDV SODFHG LQ WKH SDUWLFLSDQWfV PRXWKf 'DWD ZHUH VXPPDUL]HG DV SHUFHQWDJH RI DSSURDFK UHVSRQGLQJ E\ GLYLGLQJ WKH QXPEHU RI WULDOV GXULQJ ZKLFK DQ DSSURDFK UHVSRQVH RFFXUUHG IRU DQ LWHP E\ WKH WRWDO QXPEHU RI WULDOV GXULQJ ZKLFK WKH LWHP ZDV SUHVHQWHG DQG PXOWLSO\LQJ E\ b ,QWHUREVHUYHU DJUHHPHQW ZDV DVVHVVHG GXULQJ D PHDQ RI b RI WKH WULDOV 0HDQ LQWHUREVHUYHU DJUHHPHQW IRU DSSURDFK UHVSRQGLQJ ZDV b UDQJH b WR bf 5HVXOWV DQG 'LVFXVVLRQ 5HVXOWV RI WKH OHLVXUH LWHP DVVHVVPHQW LQGLFDWHG D KLHUDUFK\ RI SUHIHUUHG OHLVXUH LWHPV IRU RI WKH SDUWLFLSDQWV WKH UHPDLQLQJ SDUWLFLSDQWV LQWHUDFWHG ZLWK DOO LWHPV b RI WKH WLPHf )LJXUH WRS SDQHOf VKRZV GDWD IURP /\QQfV SUHIHUHQFH DVVHVVPHQW ZKLFK DUH UHSUHVHQWDWLYH GDWD RI WKH SDUWLFLSDQWV ZKR H[KLELWHG GLIIHUHQWLDO UHVSRQGLQJ GXULQJ WKH DVVHVVPHQW 7KH WKUHH LWHPV DVVRFLDWHG ZLWK WKH KLJKHVW GXUDWLRQ RI LWHP

PAGE 40

)22' ,7(06 )LJXUH 'XUDWLRQ RI FRQWDFW ZLWK HDFK RI WKH LWHPV SUHVHQWHG GXULQJ /\QQfV OHLVXUH LWHP DVVHVVPHQW WRS SDQHOf DQG SHUFHQWDJH RI WULDOV DSSURDFK IRU HDFK LWHP SUHVHQWHG GXULQJ 0LNHfV IRRG DVVHVVPHQW ERWWRP SDQHOf 3(5&(17$*( 2) 75,$/6 $3352$&+f '85$7,21 2) ,7(0 &217$&7 > L L L L R .! 2 .! 2 X! R /\L

PAGE 41

FRQWDFW IRU /\QQ ZHUH D VWULQJ RI EHDGV D ULEERQ DQG D SODVWLF ULQJ $W OHDVW RQH RI WKHVH WKUHH LWHPV ZDV XVHG GXULQJ /\QQfV 1&5 VHVVLRQV 5HVXOWV RI WKH IRRG LWHP DVVHVVPHQW VXJJHVWHG VHYHUDO SUHIHUUHG IRRG LWHPV IRU DOO SDUWLFLSDQWV 'DWD IRU WKUHH RI WKH ILYH SDUWLFLSDQWV LQGLFDWHG D KLHUDUFK\ RI SUHIHUUHG OHLVXUH LWHPV GDWD IRU WKH UHPDLQLQJ WZR SDUWLFLSDQWV LQGLFDWHG b DSSURDFK UHVSRQGLQJ DFURVV DOO LWHPV )LJXUH ERWWRP SDQHOf VKRZV 0LNHfV GDWD IURP WKLV DVVHVVPHQW ZKLFK DUH UHSUHVHQWDWLYH RI WKRVH REWDLQHG IRU SDUWLFLSDQWV ZKR H[KLELWHG GLIIHUHQWLDO UHVSRQGLQJ GXULQJ WKLV DVVHVVPHQW 7KH WKUHH HGLEOHV DVVRFLDWHG ZLWK WKH KLJKHVW SHUFHQWDJH RI DSSURDFK UHVSRQGLQJ IRU 0LNH ZHUH )LJ 1HZWRQV JUDKDP FUDFNHUV DQG FRRNLHV $W OHDVW RQH RI WKHVH HGLEOHV ZDV GHOLYHUHG GXULQJ 0LNHf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fV KDQG DQG D OHLVXUH LWHP 'XULQJ VHVVLRQV LQ ZKLFK UHVSRQVH EORFNLQJ RU UHVWUDLQW ZDV LPSOHPHQWHG WKH SHUFHQWDJH RI LQWHUYDOV LQ ZKLFK +0 DWWHPSWV RFFXUUHG ZDV UHFRUGHG ,QWHUREVHUYHU DJUHHPHQW ZDV DVVHVVHG GXULQJ D PHDQ RI b RI WKH VHVVLRQV 0HDQ LQWHUREVHUYHU DJUHHPHQW ZDV b UDQJH b WR bf DQG b UDQJH b WR bf IRU KDQG PRXWKLQJ DQG LWHP FRQWDFW UHVSHFWLYHO\

PAGE 42

([SHULPHQWDO 6HTXHQFH DQG 'HVLJQ $OO SDUWLFLSDQWV ZHUH H[SRVHG WR EDVHOLQH DQG RQH RU PRUH WUHDWPHQW FRQGLWLRQV (YHU\RQH ZDV H[SRVHG WR 1&5 VHH EHORZf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f $OO SDUWLFLSDQWV ZHUH H[SRVHG WR WKLV FRQGLWLRQ 'XULQJ 1&5 VHVVLRQV SDUWLFLSDQWV KDG FRQWLQXRXV DFFHVV WR RQH RU PRUH RI WKHLU WRS WKUHH SUHIHUUHG OHLVXUH LWHPV LGHQWLILHG GXULQJ WKH OHLVXUH LWHP SUHIHUHQFH DVVHVVPHQW 7KH LWHPVf WKDW ZHUH XVHG ZHUH KHOG FRQVWDQW DFURVV DOO 1&5 VHVVLRQV 7KDW LV LI WKH WRS SUHIHUUHG LWHP ZDV WKH RQO\ LWHP XVHG LW ZDV WKH RQO\ LWHP GXULQJ DOO 1&5 VHVVLRQV DQG LI WKH VHFRQG PRVW SUHIHUUHG LWHP ZDV XVHG WKLV LWHP ZDV XVHG GXULQJ DOO 1&5 VHVVLRQV $ WKHUDSLVW ZDV SUHVHQW KRZHYHU QR LQWHUDFWLRQ ZDV GHOLYHUHG 'LIIHUHQWLDO UHLQIRUFHPHQW RI DOWHUQDWLYH EHKDYLRU '5$f ,QGLYLGXDOV ZKR H[KLELWHG ORZ OHYHOV RI LWHP FRQWDFW GXULQJ 1&5 VHVVLRQV ZHUH H[SRVHG WR WKLV FRPSRQHQW LQ DQ DWWHPSW WR LQFUHDVH WKHLU HQJDJHPHQW ZLWK OHLVXUH LWHPV ,Q DGGLWLRQ RQH SDUWLFLSDQW ZKR H[KLELWHG KLJK OHYHOV RI LWHP FRQWDFW DQG KLJK OHYHOV RI +0f ZDV DOVR H[SRVHG WR WKLV WUHDWPHQW FRPSRQHQW 'XULQJ WKHVH VHVVLRQV SDUWLFLSDQWV KDG FRQWLQXRXV DFFHVV WR WKHLU WRS SUHIHUUHG OHLVXUH LWHPV DQG ZHUH SK\VLFDOO\ SURPSWHG WR HQJDJH LQ LWHP FRQWDFW $ WKUHHVWHS SURPSWLQJ VHTXHQFH YRFDO PRGHO DQG SK\VLFDO

PAGE 43

JXLGDQFHf ZDV LQLWLDWHG HYHU\ V XQOHVV WKH SDUWLFLSDQW ZDV DOUHDG\ HQJDJLQJ LQ LWHP FRQWDFW DW WKH EHJLQQLQJ RI WKH LQWHUYDO 7KH SURPSWLQJ VHTXHQFH EHJDQ ZLWK D YRFDO SURPSW HJ f3OD\ ZLWK WKH LWHPff ,I WKH SDUWLFLSDQW GLG QRW EHJLQ PDQLSXODWLQJ WKH LWHP ZLWKLQ V WKHQ WKH WKHUDSLVW PRGHOHG WKH DSSURSULDWH EHKDYLRU ZKLOH UHSHDWLQJ WKH YRFDO SURPSW ,I WKH SDUWLFLSDQW GLG QRW HQJDJH ZLWK WKH LWHP ZLWKLQ V RI WKH PRGHO SURPSW WKH WKHUDSLVW SK\VLFDOO\ JXLGHG KHUKLP WR GR VR 3DUWLFLSDQWVf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f +0 DWWHPSWV RQ D IL[HGUDWLR VFKHGXOH 7KH WKHUDSLVW EORFNHG SDUWLFLSDQWVf KDQGV IURP HQWHULQJ WKHLU PRXWKV E\ SODFLQJ WKH SDOP RI WKH WKHUDSLVWfV KDQG DERXW FP LQ IURQW RI WKH SDUWLFLSDQWVf PRXWKV +0 DWWHPSWV ZHUH VFRUHG HDFK WLPH D SDUWLFLSDQWfV KDQG FRQWDFWHG WKH EDFN RI WKH WKHUDSLVWfV KDQG %ULHI PDQXDO UHVWUDLQW 2QH LQGLYLGXDO $YHU\ ZKRVH +0 ZDV QRW UHVSRQVLYH WR RWKHU LQWHUYHQWLRQV ZDV H[SRVHG WR EULHI PDQXDO UHVWUDLQW 'XULQJ WKHVH VHVVLRQV $YHU\ KDG FRQWLQXRXV DFFHVV WR KHU WRS SUHIHUUHG OHLVXUH LWHPV &RQWLQJHQW XSRQ +0 DWWHPSWV WKH WKHUDSLVW KHOG $YHU\fV KDQGV LQ KHU ODS IRU V 6HVVLRQ WLPH GLG QRW LQFOXGH WLPH GXULQJ ZKLFK $YHU\ ZDV LQ EULHI UHVWUDLQW

PAGE 44

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b RI WKH LQWHUYDOV DFURVV DOO 1&5 VHVVLRQV 6L[ RI WKH LQGLYLGXDOV ZKR UHFHLYHG WKH 1&5 FRPSRQHQW ZHUH VXFFHVVIXOO\ WUHDWHG E\ WKLV LQWHUYHQWLRQ 7KH LQGLYLGXDOV ZKR ZHUH QRW VXFFHVVIXOO\ WUHDWHG GXULQJ 1&5 UHFHLYHG RQH RU PRUH RI WKH VXEVHTXHQW WUHDWPHQW FRPSRQHQWV 1&5 '5$ 5HVSRQVH %ORFNLQJ )LJXUH VKRZV WUHDWPHQW UHVXOWV IRU $QG\ .HUL %HWK 0LNH DQG .DWH 'XULQJ EDVHOLQH SDUWLFLSDQWV H[KLELWHG PRGHUDWHWRKLJK OHYHOV RI +0 'XULQJ 1&5 .DWHfV +0 GLG QRW GHFUHDVH DQG FRQWLQXHG WR RFFXU DW KLJK OHYHOV %HWKfV 0LNHfV DQG $QG\fV +0 GHFUHDVHG LQLWLDOO\ DQG WKHQ LQFUHDVHG DQG PDLQWDLQHG DW KLJK OHYHOV DQG .HULfV +0 GHFUHDVHG DQG PDLQWDLQHG DW ORZ OHYHOV $OO SDUWLFLSDQWV H[FHSW IRU 0LNH H[KLELWHG ORZ OHYHOV RI LWHP FRQWDFW GXULQJ 1&5 VHVVLRQV ,Q DQ DWWHPSW WR UHGXFH +0 IRU %HWK 0LNH &DWH DQG $QG\f DQG WR LQFUHDVH LWHP FRQWDFW %HWK .HUL .DWH DQG $QG\f SDUWLFLSDQWV ZHUH H[SRVHG WR WKH 1&5 '5$ 5HVSRQVH EORFNLQJ LQWHUYHQWLRQ 'XULQJ WKLV

PAGE 45

3(5&(17$*( 2) ,17(59$/6 )LJXUH 3HUFHQWDJH RI LQWHUYDOV +0 RU +0 DQG LWHP FRQWDFW GXULQJ %/ DQG 1&5 VHVVLRQV IRU 'HEUD OHIW WRS SDQHOf 'HDQQD OHIW PLGGOH SDQHOf *LQD OHIW ERWWRP SDQHOf 7HG ULJKW WRS SDQHOf 'DQ ULJKW PLGGOH SDQHOf DQG -DPHV ULJKW ERWWRP SDQHOf

PAGE 46

7 Wf§ff§ff§f ff§7 Q 3HUFHQWDJH RI LQWHUYDOV +0 RU +0 DQG LWHP FRQWDFW GXULQJ %/ 1&5 DQG 1&5 '5$ UHVSRQVH EORFNLQJ VHVVLRQV IRU %HWK WRS SDQHOf .HUL VHFRQG SDQHOf 0LNH WKLUG SDQHOf .DWH IRXUWK SDQHOf DQG $QG\ ERWWRP SDQHOf

PAGE 47

FRQGLWLRQ +0 GHFUHDVHG WR PXFK ORZHU OHYHOV %HWK 0LNH .DWH DQG $QG\f RU FRQWLQXHG WR RFFXU DW ORZ OHYHOV .HULf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f ZKR H[KLELWHG KLJK OHYHOV RI LWHP FRQWDFW GXULQJ 1&5 ZH PRGLILHG WKH WUHDWPHQW SDFNDJH E\ MXVW FRQGXFWLQJ 1&5 EORFNLQJ IRU DOO VXEVHTXHQW SDUWLFLSDQWV H[KLELWLQJ KLJK OHYHOV RI LWHP FRQWDFW GXULQJ 1&5 VHH EHORZf 1&5 UHVSRQVH EORFNLQJ )LJXUH VKRZV WUHDWPHQW UHVXOWV IRU /\QQ DQG /LVD 'XULQJ EDVHOLQH WKH\ H[KLELWHG PRGHUDWHWRKLJK OHYHOV RI +0 'XULQJ 1&5 /\QQ LQLWLDOO\ H[KLELWHG PRGHUDWH OHYHOV RI LWHP FRQWDFW KRZHYHU LWHP FRQWDFW LQFUHDVHG WR KLJK OHYHOV DFURVV VXEVHTXHQW VHVVLRQV ,Q DGGLWLRQ +0 FRQWLQXHG WR RFFXU DW PRGHUDWHWRKLJK OHYHOV 'XULQJ 1&5 /LVD LPPHGLDWHO\ H[KLELWHG KLJK OHYHOV RI LWHP FRQWDFW WKDW PDLQWDLQHG WKURXJKRXW WKH 1&5 SKDVH /LVDfV +0 RFFXUUHG DW YDULDEOH OHYHOV RIWHQ RFFXUULQJ DW YHU\ ORZ OHYHOV IRU VHYHUDO FRQVHFXWLYH 1&5 VHVVLRQV DQG WKHQ UHHPHUJLQJ DW YHU\ KLJK OHYHOV %HFDXVH /\QQfV DQG /LVDfV LWHP FRQWDFW RFFXUUHG DW KLJK OHYHOV GXULQJ 1&5 WKHUH ZDV QR QHHG IRU WKH '5$ FRPSRQHQW 7KHUHIRUH /\QQ DQG /LVD ZHUH H[SRVHG WR 1&5 UHVSRQVH EORFNLQJ RQO\ 'XULQJ 1&5 UHVSRQVH EORFNLQJ /\QQfV +0

PAGE 48

3(5&(17$*( 2) ,17(59$/6 %/ 1&5 1&55 EORFN 6(66,216 )LJXUH 3HUFHQWDJH RI LQWHUYDOV +0 RU +0 DQG LWHP FRQWDFW GXULQJ EDVHOLQH 1&5 DQG 1&5 UHVSRQVH EORFNLQJ ZLWKRXW '5$f IRU /\QQ WRS SDQHOf DQG /LVD ERWWRP SDQHOf

PAGE 49

LPPHGLDWHO\ GHFUHDVHG EXW UHPDLQHG YDULDEOH RQ VHYHUDO VHVVLRQV XQWLO HYHQWXDOO\ UHDFKLQJ ORZ OHYHOV ,Q DGGLWLRQ DIWHU WKH ILUVW VHVVLRQV /\QQfV LWHP FRQWDFW GHFUHDVHG WR YHU\ ORZ OHYHOV KRZHYHU GXULQJ WKH WK VHVVLRQ LWHP FRQWDFW UHFRYHUHG DQG UHPDLQHG KLJK GXULQJ DOO VXEVHTXHQW VHVVLRQV 'XULQJ 1&5 UHVSRQVH EORFNLQJ /LVDfV +0 GHFUHDVHG WR ORZHU OHYHOV ZLWK OHVV YDULDELOLW\ WKDQ WKDW REVHUYHG GXULQJ WKH 1&5 FRPSRQHQW ,Q DGGLWLRQ /LVDfV LWHP FRQWDFW UHPDLQHG DW YHU\ KLJK OHYHOV 1&5 EULHI PDQXDO UHVWUDLQW $YHU\fV UHVXOWV DUH VKRZQ LQ )LJXUH 'XULQJ EDVHOLQH $YHU\ H[KLELWHG KLJK OHYHOV RI +0 'XULQJ 1&5 +0 GHFUHDVHG EXW FRQWLQXHG WR RFFXU DW PRGHUDWH OHYHOV DQG LWHP FRQWDFW RFFXUUHG DW KLJK OHYHOV 'XULQJ WKH 1&5 UHVSRQVH EORFNLQJ LQWHUYHQWLRQ $YHU\ FRQWLQXHG WR H[KLELW PRGHUDWH OHYHOV RI +0 DQG KLJK OHYHOV RI LWHP FRQWDFW %HFDXVH 1&5 UHVSRQVH EORFNLQJ GLG QRW VXFFHVVIXOO\ UHGXFH $YHU\fV +0 ZH DVVHVVHG WKH HIIHFWLYHQHVV RI 1&5 EULHI PDQXDO UHVWUDLQW 'XULQJ D UHWXUQ WR EDVHOLQH $YHU\fV +0 UHWXUQHG WR KLJK OHYHOV VLPLODU WR WKDW REVHUYHG GXULQJ WKH SUHYLRXV %/ FRQGLWLRQ 'XULQJ 1&5 PDQXDO UHVWUDLQW +0 LPPHGLDWHO\ GHFUHDVHG WR YHU\ ORZ OHYHOV 'XULQJ WKH WK VHVVLRQ RI WKH 1&5 PDQXDO UHVWUDLQW FRQGLWLRQ +0 LQFUHDVHG VOLJKWO\ IRU VHVVLRQV DQG WKHQ GHFUHDVHG DJDLQ WR ORZ OHYHOV ,WHP FRQWDFW RFFXUUHG DW KLJK OHYHOV WKURXJKRXW WKH 1&5 PDQXDO UHVWUDLQW FRQGLWLRQ 'XULQJ D VHFRQG UHWXUQ WR EDVHOLQH $YHU\ H[KLELWHG KLJK OHYHOV RI +0 'XULQJ WKH ILQDO 1&5 UHVWUDLQW FRQGLWLRQ +0 LPPHGLDWHO\ GHFUHDVHG WR QHDU]HUR OHYHOV DQG LWHP FRQWDFW RFFXUUHG DW KLJK OHYHOV 7KHVH OHYHOV RI +0 DQG LWHP FRQWDFW PDLQWDLQHG WKURXJKRXW WKH ILQDO 1&5 EULHI PDQXDO UHVWUDLQW SKDVH 7DEOH VKRZV D VXPPDU\ RI WKH QXPEHU RI SDUWLFLSDQWV ZKR UHFHLYHG HDFK WUHDWPHQW FRPSRQHQW DQG WKH QXPEHU ZKR ZHUH VXFFHVVIXOO\ WUHDWHG E\ WKDW FRPSRQHQW

PAGE 50

%/ 1&5 1&55 %/2&. %/ 1&55(675$,17 %/ )LJXUH 3HUFHQWDJH RI LQWHUYDOV +0 RU +0 DQG LWHP FRQWDFW GXULQJ EDVHOLQH 1&5 1&5 UHVSRQVH EORFNLQJ DQG 1&5 EULHI UHVWUDLQW FRQGLWLRQV IRU $YHU\

PAGE 51

1&5 ZDV LPSOHPHQWHG ZLWK DOO SDUWLFLSDQWV DQG ZHUH VXFFHVVIXOO\ WUHDWHG E\ WKLV LQWHUYHQWLRQ 1&5 '5$ UHVSRQVH EORFNLQJ ZDV LPSOHPHQWHG ZLWK LQGLYLGXDOV DQG DOO ZHUH VXFFHVVIXOO\ WUHDWHG E\ WKLV LQWHUYHQWLRQ 1&5 UHVSRQVH EORFNLQJ ZLWKRXW '5$f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f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

PAGE 52

LQGLYLGXDOV ZKR H[KLELWHG KLJK OHYHOV RI +0 DQGRU ORZ OHYHOV RI LWHP FRQWDFW GXULQJ 1&5 ZKHUHDV UHVSRQVH EORFNLQJ ZLWKRXW WKH '5$ FRPSRQHQWf ZDV LPSOHPHQWHG ZLWK LQGLYLGXDOV ZKR H[KLELWHG KLJK OHYHOV RI +0 DQG KLJK OHYHOV RI LWHP FRQWDFW GXULQJ 1&5 ,Q DGGLWLRQ WKH PRVW LQWUXVLYH LQWHUYHQWLRQ EULHI PDQXDO UHVWUDLQW ZDV UHTXLUHG IRU RQO\ RQH SDUWLFLSDQW 7KH DGGLWLYH WUHDWPHQW GHVLJQ XVHG LQ WKLV LQYHVWLJDWLRQ H[WHQGV WKDW XVHG E\ /LQGEHUJ HW DO f ZKR SURJUHVVLYHO\ LQWURGXFHG WUHDWPHQW FRPSRQHQWV LQ DQ DGGLWLYH IDVKLRQ IRU SUREOHP EHKDYLRUV PDLQWDLQHG E\ DXWRPDWLF UHLQIRUFHPHQW )RU H[DPSOH WKH\ VHTXHQWLDOO\ LQWURGXFHG 1&5 SURPSWLQJ '5$ UHVSRQVH EORFNLQJ DQG SURWHFWLYH HTXLSPHQW ZLWK WZR SDUWLFLSDQWV ZKR H[KLELWHG VWHUHRW\S\ 5HVXOWV LQGLFDWHG WKDW UHVSRQVH EORFNLQJ RU SURWHFWLYH HTXLSPHQW ZDV QHFHVVDU\ IRU UHGXFLQJ SDUWLFLSDQWVf VWHUHRW\S\ 7KH SUHVHQW VWXG\ H[WHQGV WKDW XVHG E\ /LQGEHUJ HW DO LQ WZR ZD\V Df ZH LPSOHPHQWHG WKH WUHDWPHQW GHVLJQ ZLWK D JUHDWHU QXPEHU RI SDUWLFLSDQWV DQG Ef ZH REWDLQHG VXFFHVVIXO UHVXOWV ZLWK HDFK RI WKH GLIIHUHQW WUHDWPHQW FRPSRQHQWV XVHG 7KH WUHDWPHQW GHVLJQ XVHG LQ WKH SUHVHQW VWXG\ DOVR H[WHQGV WKDW XVHG E\ &RRSHU HW DO f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f ZHUH LQ

PAGE 53

HIIHFW 1H[W RQH FRPSRQHQW HJ QRQFRQWLQJHQW DFFHVV WR WR\V DQG VRFLDO LQWHUDFWLRQf ZDV UHPRYHG ZKLOH WKH RWKHU FRPSRQHQWV UHPDLQHG 2XU GHVLJQ GLIIHUV IURP WKDW XVHG E\ &RRSHU HW DO LQ WKDW ZH HYDOXDWHG WKH HIIHFWLYHQHVV RI D VSHFLILF FRPSRQHQW LQLWLDOO\ LH 1&5f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bf 2QH SRVVLEOH H[SODQDWLRQ IRU WKH OLPLWHG HIIHFWLYHQHVV RI 1&5 PD\ KDYH EHHQ WKH IDFW WKDW VHYHUDO RI WKH SDUWLFLSDQWV GLG QRW KDYH LWHP PDQLSXODWLRQ LQ WKHLU FXUUHQW UHSHUWRLUHV ,Q DGGLWLRQ IRU LQGLYLGXDOV ZKR GLG KDYH LWHP FRQWDFW LQ WKHLU UHSHUWRLUHV WKH OHLVXUH LWHPV PD\ QRW KDYH HIIHFWLYHO\ FRPSHWHG ZLWK WKH VHQVRU\ VWLPXODWLRQ SURGXFHG E\ +0 )RU H[DPSOH DOWKRXJK 0LNH HQJDJHG LQ LWHP FRQWDFW GXULQJ 1&5 KLV +0 GLG QRW GHFUHDVH VXJJHVWLQJ WKDW WKH VWLPXODWLRQ SURGXFHG E\ WKH OHLVXUH LWHPV GLG QRW HIIHFWLYHO\ FRPSHWH ZLWK 0LNHfV +0 )XWXUH UHVHDUFK FRXOG LGHQWLI\ ZKDW IDFWRUV ZHUH UHVSRQVLEOH IRU WKH LQHIIHFWLYHQHVV RI 1&5 ODFN RI LWHP FRQWDFW LQ UHSHUWRLUH RU ODFN RI FRPSHWLWLRQf E\ FRQGXFWLQJ 1&5 DORQH ERWK EHIRUH DQG DIWHU '5$ ,I 1&5 LV LQHIIHFWLYH EHFDXVH SDUWLFLSDQWV GR QRW KDYH LWHP FRQWDFW LQ WKHLU UHSHUWRLUH 1&5 DORQH PLJKW PDLQWDLQ KLJK OHYHOV RI LWHP FRQWDFW DQG ORZ OHYHOV RI

PAGE 54

SUREOHP EHKDYLRU RQFH LWHP FRQWDFW LV VKDSHG GXULQJ '5$ ,Q DGGLWLRQ IXWXUH UHVHDUFK LV QHHGHG WR GHWHUPLQH IDFWRUV WKDW PD\ LQFUHDVH WKH OLNHOLKRRG RI LGHQWLI\LQJ OHLVXUH LWHPV WKDW PD\ FRPSHWH ZLWK +0 2QH ZD\ WR LQFUHDVH WKH SUREDELOLW\ WKDW OHLVXUH LWHPV XVHG GXULQJ 1&5 ZLOO FRPSHWH ZLWK LWHP FRQWDFW LV WR PHDVXUH +0 GXULQJ WKH SUHIHUHQFH DVVHVVPHQW DQG RQO\ XVH LWHPV FRUUHODWHG ZLWK ERWK KLJK OHYHOV RI LWHP FRQWDFW DQG ORZ OHYHOV RI +0 5RVFRH ,ZDWD t *RK f $QRWKHU PHWKRG WKDW PD\ XVHG WR LQFUHDVH WKH OLNHOLKRRG WKDW LWHPV XVHG GXULQJ 1&5 ZLOO FRPSHWH ZLWK +0 LV WR LGHQWLI\ LWHPV WKDW SURGXFH VWLPXODWLRQ VLPLODU WR WKDW SURGXFHG E\ WKH EHKDYLRU 3LD]]D $GHOLQLV +DQHO\ *RK t 'HOLD f 7KH RQO\ WUHDWPHQW FRPSRQHQW WKDW VXFFHVVIXOO\ UHGXFHG +0 DQGRU LQFUHDVHG LWHP FRQWDFW IRU DOO SDUWLFLSDQWV 1 f ZDV WKH '5$ LQWHUYHQWLRQ 7KHUH DUH WKUHH SRVVLEOH UHDVRQV IRU WKH KLJK GHJUHH RI HIIHFWLYHQHVV RI '5$ LQ WKH SUHVHQW LQYHVWLJDWLRQ )LUVW WKH UHLQIRUFHUV XVHG GXULQJ '5$ ZHUH LGHQWLILHG WKURXJK D IRUPDO SUHIHUHQFH DVVHVVPHQW 6HFRQG LWHP FRQWDFW ZKHQ VKDSHG HIIHFWLYHO\ FRPSHWHG ZLWK SDUWLFLSDQWVf +0 7KLUG '5$ ZDV DOZD\V FRPELQHG ZLWK UHVSRQVH EORFNLQJ ZKLFK PD\ KDYH FRQWULEXWHG WR WKH HIIHFWV REWDLQHG )XWXUH UHVHDUFK FRXOG GHWHUPLQH XQGHU ZKDW FRQGLWLRQV '5$ LV PRVW HIIHFWLYH LH ZKHQ LPSOHPHQWHG DORQH RU ZKHQ FRPELQHG ZLWK UHVSRQVH EORFNLQJf WR DVVHVV WKH H[WHQW WR ZKLFK '5$ GHSHQGV RQ EHLQJ FRPELQHG ZLWK UHVSRQVH EORFNLQJ

PAGE 55

',6&866,21 +DQG PRXWKLQJ KDV EHHQ UHIHUUHG WR DV D IRUP RI ERWK VWHUHRW\SLF DQG VHOI LQMXULRXV EHKDYLRU DQG LW KDV EHHQ GHVFULEHG DV EHKDYLRU QRW PDLQWDLQHG E\ VRFLDO FRQWLQJHQFLHV +RZHYHU WKH FODVVLILFDWLRQ RI KDQG PRXWKLQJ LQWR DQ\ RQH RI WKHVH FDWHJRULHV VHHPV TXHVWLRQDEOH LQ WKH DEVHQFH RI DSSURSULDWH DVVHVVPHQW PHWKRGV 'LUHFW REVHUYDWLRQV RI SDUWLFLSDQWVf LQMXU\ DVVRFLDWHG ZLWK +0 LQGLFDWHG WKDW +0 ZKHQ LW GRHV RFFXU LV LQMXULRXV DERXW D WKLUG RI WKH WLPH %HFDXVH +0 PD\ RU PD\ QRW EH LQMXULRXV LW PD\ EH EHVW QRW WR UHIHU WR +0 DV H[FOXVLYHO\ D VWHUHRW\SLF RU D VHOILQMXULRXV EHKDYLRU +RZHYHU LQ WKH SUHVHQW VWXG\ UHVXOWV RI IXQFWLRQDO DQDO\VHV LQGLFDWHG WKDW +0 ZDV RYHUZKHOPLQJO\ PDLQWDLQHG E\ DXWRPDWLF UHLQIRUFHPHQW 1HYHUWKHOHVV JLYHQ WKDW RQH SDUWLFLSDQWfV +0 ZDV PDLQWDLQHG E\ VRFLDO UHLQIRUFHPHQW DWWHQWLRQf LW LV LPSRUWDQW WR FRQGXFW D IXQFWLRQDO DQDO\VLV WR GHILQLWHO\ UXOH RXW EHKDYLRUDO PDLQWHQDQFH E\ VRFLDO FRQWLQJHQFLHV ,I UHVXOWV RI WKH IXQFWLRQDO DQDO\VLV LQGLFDWH WKDW WKH EHKDYLRU LV PDLQWDLQHG E\ VRFLDO FRQWLQJHQFLHV LQWHUYHQWLRQV IRXQG HIIHFWLYH IRU EHKDYLRU PDLQWDLQHG E\ VRFLDO UHLQIRUFHPHQW VKRXOG EH XVHG ,W LV XQFOHDU ZK\ VXFK D KLJK SURSRUWLRQ RI SDUWLFLSDQWVf +0 ZDV PDLQWDLQHG E\ DXWRPDWLF UHLQIRUFHPHQW 2QH H[SODQDWLRQ PD\ EH WKDW XQOLNH RWKHU IRUPV RI 6,% WKDW SURGXFH PRUH LPPHGLDWH GDPDJH HJ KHDG EDQJLQJf +0 PD\ HPHUJH LQLWLDOO\ DV D IDLUO\ EHQLJQ EHKDYLRU $V D UHVXOW LW PD\ QRW SURGXFH LPPHGLDWH VRFLDO FRQVHTXHQFHV DQG DV D UHVXOW PD\ EH OHVV OLNHO\ WKDQ PRUH VHYHUH IRUPV RI 6,% WR DFTXLUH D VRFLDO

PAGE 56

IXQFWLRQ 7KXV ZKHQ +0 HYHQWXDOO\ SURGXFHV LQMXULHV DQG UHVXOWV LQ VRFLDO FRQVHTXHQFHVf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f $ VWUHQJWK RI WKLV VWXG\ ZDV WKDW WKH XWLOLW\ RI WKH IXQFWLRQDO DQDO\VLV GHVLJQ ZDV DVVHVVHG E\ FRQGXFWLQJ WUHDWPHQW SURFHGXUHV EDVHG RQ WKH ILQGLQJV REWDLQHG GXULQJ WKH IXQFWLRQDO DQDO\VLV *LYHQ WKDW DOO SDUWLFLSDQWV ZHUH VXFFHVVIXOO\ WUHDWHG E\ LQWHUYHQWLRQV VSHFLILFDOO\ GHVLJQHG IRU EHKDYLRU PDLQWDLQHG E\ DXWRPDWLF UHLQIRUFHPHQW WKH YDOLGLW\ RI WKH PRGLILHG IXQFWLRQDO DQDO\VLV GHVLJQ ZDV VXSSRUWHG %HFDXVH PRVW SDUWLFLSDQWVf +0 ZDV PDLQWDLQHG E\ DXWRPDWLF UHLQIRUFHPHQW WKH XWLOLW\ RI WKH VSHFLILF IXQFWLRQDO DQDO\VLV VHTXHQFH WKDW ZH XVHG PD\ QRW EH DSSURSULDWH IRU EHKDYLRU VXVSHFWHG WR EH PDLQWDLQHG E\ VRFLDO FRQWLQJHQFLHV %HFDXVH SUHYLRXV UHVHDUFK LQGLFDWHG WKDW KDQG PRXWKLQJ LV RIWHQ PDLQWDLQHG E\ QRQVRFLDO YDULDEOHV ZH FRQGXFWHG DORQH VHVVLRQV LQ D UDWLR ZLWK DWWHQWLRQ DQG GHPDQG VHVVLRQV %\ FRQWUDVW IRU UHVSRQVH WRSRJUDSKLHV SUHVXPHG WR EH PDLQWDLQHG E\ VRFLDO YDULDEOHV HJ

PAGE 57

DJJUHVVLRQf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f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f )XWXUH UHVHDUFK LV QHHGHG WR GHWHUPLQH ZKLFK RI WKH YDULRXV WUHDWPHQW FRPSRQHQWV LV PRVW HIIHFWLYH ZKHQ FRPSDUHG LQ LVRODWLRQ

PAGE 58

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

PAGE 59

5()(5(1&(6 %DOO 7 6 &DPSEHOO 5 t %DUNHPH\HU 5 f $LU VSOLQWV DSSOLHG WR FRQWURO VHOILQMXULRXV ILQJHU VXFNLQJ LQ SURIRXQGO\ UHWDUGHG LQGLYLGXDOV -RXUQDO RI %HKDYLRU 7KHUDS\ DQG ([SHULPHQWDO 3V\FKLDWU\ %RGILVK :f &UDZIRUG 7 :f 3RZHOO 6 %f 3DUNHU (f *ROGHQ 5 1 t /HZLV 0 + f &RPSXOVLRQ LQ DGXOWV ZLWK PHQWDO UHWDUGDWLRQ 3UHYDOHQFH SKHQRPHQRORJ\ DQG FRPRUELGLW\ ZLWK VWHUHRW\S\ DQG VHOILQMXU\ $PHULFDQ -RXUQDO RQ 0HQWDO 5HWDUGDWLRQ &RRSHU / :DFNHU 3 7KXUVE\ 3ODJPDQQ / $ +DUGLQJ 0LOODUG 7 t 'HUE\ 0 f $QDO\VLV RI WKH HIIHFWV RI WDVN SUHIHUHQFHV WDVN GHPDQGV DQG DGXOW DWWHQWLRQ RQ FKLOG EHKDYLRU LQ RXWSDWLHQW DQG FODVVURRP VHWWLQJV -RXUQDO RI $SSOLHG %HKDYLRU $QDO\VLV &RRSHU / :DFNHU 3 0F&RPDV %URZQ 3HFN 6 0 5LFKPDQ 'UHZ )ULVFKPH\HU 3 t 0LOODUG 7 f 8VH RI FRPSRQHQW DQDO\VHV WR LGHQWLI\ DFWLYH YDULDEOHV LQ WUHDWPHQW SDFNDJHV IRU FKLOGUHQ ZLWK IHHGLQJ GLVRUGHUV -RXUQDO RI $SSOLHG %HKDYLRU $QDO\VLV 'H/HRQ ,ZDWD % $ &RQQHUV t :DOODFH 0 f ([DPLQDWLRQ RI DPELJXRXV VWLPXOXV SUHIHUHQFHV ZLWK GXUDWLRQEDVHG PHDVXUHV -RXUQDO RI $SSOLHG %HKDYLRU $QDO\VLV 'HUE\ 0 :DFNHU 3 6DVVR 6WHHJH 0 1RUWKXS &LJUDQG t $VPXV f %ULHI IXQFWLRQDO DVVHVVPHQW WHFKQLTXHV WR HYDOXDWH DEHUUDQW EHKDYLRU LQ DQ RXWSDWLHQW VHWWLQJ $ VXPPDU\ RI FDVHV -RXUQDO RI $SSOLHG %HKDYLRU $QDO\VLV 'RUVH\ 0 ) ,ZDWD % $ 2QJ 3 t 0F6ZHHQ 7 ( f 7UHDWPHQW RI VHOILQMXULRXV EHKDYLRU XVLQJ D ZDWHU PLVW ,QLWLDO UHVSRQVH VXSSUHVVLRQ DQG JHQHUDOL]DWLRQ -RXUQDO RI $SSOLHG %HKDYLRU $QDO\VLV 'XUD -5 0XOLFN -$ t 5DVQDNH /. f 3UHYDOHQFH RI VWHUHRW\S\ DPRQJ LQVWLWLRQDOL]HG QRQDPEXODWRU\ SURIRXQGO\ PHQWDOO\ UHWDUGHG SHRSOH $PHULFDQ -RXUQDO RI 0HQWDO 'HILFLHQF\

PAGE 60

'XUDQG 9 0 t &ULPPLQV % f ,GHQWLI\LQJ WKH YDULDEOHV PDLQWDLQLQJ VHOILQMXULRXV EHKDYLRU -RXUQDO RI $XWLVP DQG 'HYHORSPHQWDO 'LVRUGHUV )DYHOO ( 0F*LPVH\ ) t 6FKHOO 5 0 f 7UHDWPHQW RI VHOILQMXU\ E\ SURYLGLQJ DOWHUQDWH VHQVRU\ DFWLYLWLHV $QDO\VLV DQG ,QWHUYHQWLRQ LQ 'HYHORSPHQWDO 'LVDELOLWLHV *RK + t ,ZDWD % $ f %HKDYLRUDO SHUVLVWDQFH DQG YDULDELOLW\ GXULQJ H[WLQFWLRQ RI VHOILQMXU\ PDLQWDLQHG E\ HVFDSH -RXUQDO RI $SSOLHG %HKDYLRU $QDO\VLV *RK + ,ZDWD % $ 6KRUH % $ 'H/HRQ /HUPDQ & 8OULFK 6 0 t 6PLWK 5 f $Q DQDO\VLV RI WKH UHLQIRUFLQJ SURSHUWLHV RI KDQG PRXWKLQJ -RXUQDO RI $SSOLHG %HKDYLRU $QDO\VLV *ULIILQ & :LOOLDPV ( 6WDUN 0 7 $OWPH\HU % t 0DVRQ 0 f 6HOILQMXULRXV EHKDYLRU $ VWDWHZLGH SUHYDOHQFH VXUYH\ RI WKH H[WHQW DQG FLUFXPVWDQFHV $SSOLHG 5HVHDUFK LQ 0HQWDO 5HWDUGDWLRQ +RPHU 5 f 7KH HIIHFWV RI DQ HQYLURQPHQWDO HQULFKPHQW SURJUDP RQ WKH EHKDYLRU RI LQVWLWXWLRQDOL]HG SURIRXQGO\ UHWDUGHG FKLOGUHQ -RXUQDO RI $SSOLHG %HKDYLRU $QDO\VLV ,ZDWD % $ 'RUVH\ 0 ) 6OLIHU %DXPDQ ( t 5LFKPDQ 6 Df 7RZDUG D IXQFWLRQDO DQDO\VLV RI VHOILQMXU\ -RXUQDO RI $SSOLHG %HKDYLRU $QDO\VLV 5HSULQWHG IURP $QDO\VLV DQG ,QWHUYHQWLRQ LQ 'HYHORSPHQWDO 'LVDELOLWLHV f ,ZDWD % $ 3DFH 0 'RUVH\ 0 ) =DUFRQH 5 9ROOPHU 7 5 6PLWK 5 5RGJHUV 7 $ /HUPDQ & 6KRUH % $ 0D]DOHVNL / *RK + &RZGHU\ ( .DOVKHU 0 t :LOOLV Ef 7KH IXQFWLRQV RI VHOILQMXULRXV EHKDYLRU $Q H[SHULPHQWDOHSLGHPLRORJLFDO DQDO\VLV -RXUQDO RI $SSOLHG %HKDYLRU $QDO\VLV .DKQJ 6 t ,ZDWD % $ f &RUUHVSRQGHQFH EHWZHHQ RXWFRPHV RI EULHI DQG H[WHQGHG IXQFWLRQDO DQDO\VHV -RXUQDO RI $SSOLHG %HKDYLRU $QDO\VLV /D*URZ 6 t 5HSS $ & f 6WHUHRW\SLF UHVSRQGLQJ $ UHYLHZ RI LQWHUYHQWLRQ UHVHDUFK $PHULFDQ -RXUQDO RI 0HQWDO 'HILFLHQF\ /HUPDQ & t ,ZDWD % $ f $ PHWKRGRORJ\ IRU GLVWLQJXLVKLQJ EHWZHHQ H[WLQFWLRQ DQG SXQLVKPHQW HIIHFWV DVVRFLDWHG ZLWK UHVSRQVH EORFNLQJ -RXUQDO RI $SSOLHG %HKDYLRU $QDO\VLV

PAGE 61

/LQGEHUJ 6 ,ZDWD % $ t .DKQJ 6 f 2Q WKH UHODWLRQ EHWZHHQ REMHFW PDQLSXODWLRQ DQG VWHUHRW\SLF VHOILQMXULRXV EHKDYLRU -RXUQDO RI $SSOLHG %HKDYLRU $QDO\VLV /RFNZRRG t :LOOLDPV ( f 7UHDWPHQW DQG H[WHQGHG IROORZXS RI FKURQLF KDQG PRXWKLQJ -RXUQDO RI %HKDYLRU 7KHUDS\ DQG ([SHULPHQWDO 3V\FKLDWU\ 0DXULFH 3 t 7UXGHO f 6HOILQMXULRXV EHKDYLRU 3UHYDOHQFH DQG UHODWLRQVKLSV WR HQYLURQPHQWDO HYHQWV ,Q + +ROOLV t & ( 0H\HUV (GVf /LIH WKUHDWHQLQJ EHKDYLRU $QDO\VLV DQG LQWHUYHQWLRQ SS f :DVKLQJWRQ $PHULFDQ $VVRFLDWLRQ RQ 0HQWDO 'HILFLHQF\ 0D]DOHVNL / ,ZDWD % $ 5RGJHUV 7 $ 9ROOPHU 7 5 t =DUFRQH 5 f 3URWHFWLYH HTXLSPHQW DV WUHDWPHQW IRU VWHUHRW\SLF KDQG PRXWKLQJ 6HQVRU\ H[WLQFWLRQ RU SXQLVKPHQW HIIHFWV" -RXUQDO RI $SSOLHG %HKDYLRU $QDO\VLV 0F&OXUH 7 0RVV 5 $ 0F3HWHUV : t .LUNSDWULFN 0 $ f 5HGXFWLRQ RI KDQG PRXWKLQJ E\ D ER\ ZLWK SURIRXQG PHQWDO UHWDUGDWLRQ 0HQWDO 5HWDUGDWLRQ 0F*LOO 3 +XJKHV 7HHU t 5\H / f 9DULDELOLW\ LQ VWDII UHSRUWV RI WKH IUHTXHQF\ RI FKDOOHQJLQJ EHKDYLRU 5HVHDUFK LQ 'HYHORSPHQWDO 'LVDELOLWLHV 1RUWKXS :DFNHU 6DVVR 6WHHJH 0 &LJUDQG &RRN t 'H5DDG $ f $ EULHI IXQFWLRQDO DQDO\VLV RI DJJUHVVLYH DQG DOWHUQDWLYH EHKDYLRU LQ DQ RXWFOLQLF VHWWLQJ -RXUQDO RI $SSOLHG %HKDYLRU $QDO\VLV 3DFH 0 ,YDQFLF 0 7 (GZDUGV / ,ZDWD % $ t 3DJH 7 f $VVHVVPHQW RI VWLPXOXV SUHIHUHQFH DQG UHLQIRUFHU YDOXH ZLWK SURIRXQGO\ UHWDUGHG LQGLYLGXDOV -RXUQDO RI $SSOLHG %HKDYLRU $QDO\VLV 3DLVH\ 7 + :KLWQH\ 5 % t :DLQF]DN 6 0 f &DVH VWXG\ 1RQLQYDVLYH EHKDYLRUDO WUHDWPHQW RI VHOILQMXULRXV KDQG VWHUHRW\S\ LQ D FKLOG ZLWK 5HWW V\QGURPH %HKDYLRUDO 5HVLGHQWLDO 7UHDWPHQW 3LD]]D & & $GHOLQLV +DQOH\ 3 *RK + t 'HOLD 0 f $Q HYDOXDWLRQ RI WKH HIIHFWV RI PDWFKHG VWLPXOL RQ EHKDYLRUV PDLQWDLQHG E\ DXWRPDWLF UHLQIRUFHPHQW -RXUQDO RI $SSOLHG %HKDYLRU $QDO\VLV 5DVW t -DFN 6 f 0RXWKLQJ ,Q ($ .RQDUVNL -( )DYHOO (GVf 0DQXDO IRU WKH DVVHVVPHQW DQG WUHDWPHQW RI WKH EHKDYLRU GLVRUGHUV RI SHRSOH ZLWK PHQWDO UHWDUGDWLRQ SS f 0RUJDQWRQ 1& :HVWHUQ &DUROLQD &HQWHU )RXQGDWLRQ

PAGE 62

5HSS $ & 6LQJK 1 1 2OLQJHU ( t 2OVRQ 5 f 7KH XVH RI IXQFWLRQDO DQDO\VHV WR WHVW FDXVHV RI VHOILQMXULRXV EHKDYLRU 5DWLRQDOH FXUUHQW VWDWXV DQG IXWXUH GLUHFWLRQV -RXUQDO RI 0HQWDO 'HILFLHQF\ 5HVHDUFK 5LQJGDKO ( 9ROOPHU 7 5 0DUFXV % $ t 5RDQH + 6 f $Q DQDORJXH HYDOXDWLRQ RI HQYLRPPHQWDO HQULFKPHQW 7KH UROH RI VWLPXOXV SUHIHUHQFH -RXUQDO RI $SSOLHG %HKDYLRU $QDO\VLV 5RMDKQ f 6HOILQMXULRXV DQG VWHUHRW\SLF EHKDYLRU RI QRQLQVWLWXWLRQDOL]HG PHQWDOO\ UHWDUGHG SHRSOH 3UHYDOHQFH DQG FODVVLILFDWLRQ $PHULFDQ -RXUQDO RI 0HQWDO 'HILFLHQF\ 5RVFRH ( 0 ,ZDWD % $ t *RK + f $ FRPSDULVRQ RI QRQFRQWLQJHQW UHLQIRUFHPHQW DQG VHQVRU\ H[WLQFWLRQ DV WUHDWPHQWV IRU VHOILQMXULRXV EHKDYLRU -RXUQDO RI $SSOLHG %HKDYLRU $QDO\VLV 6FKURHGHU 6 5 6FKURHGHU & 6 6PLWK % t 'DOOGRUI f 3UHYDOHQFH RI VHOILQMXULRXV EHKDYLRUV LQ D ODUJH VWDWH IDFLOLW\ IRU WKH UHWDUGHG $ WKUHH \HDU IROORZXS VWXG\ -RXUQDO RI $XWLVP DQG &KLOGKRRG 6FKL]RSKUHQLD 6KRUH % $ ,ZDWD % $ 'H/HRQ .DKQJ 6 t 6PLWK 5 f $Q DQDO\VLV RI UHLQIRUFHU VXEVWLWXWDELOLW\ XVLQJ REMHFW PDQLSXODWLRQ DQG VHOILQMXU\ DV FRPSHWLQJ UHVSRQVHV -RXUQDO RI $SSOLHG %HKDYLRU $QDO\VLV 6LQJK 1 1 t 0LOOLFKDPS & f ,QGHSHQGHQW DQG VRFLDO SOD\ DPRQJ SURIRXQGO\ PHQWDOO\ UHWDUGHG DGXOWV 7UDLQLQJ JHQHUDOL]DWLRQ DQG ORQJWHUP IROORZXS -RXUQDO RI $SSOLHG %HKDYLRU $QDO\VLV 7KRPSVRQ 5 + ,ZDWD % $ &RQQHUV t 5RVFRH ( 0 f (IIHFWV RI UHLQIRUFHPHQW IRU DOWHUQDWLYH EHKDYLRU GXULQJ SXQLVKPHQW RI VHOILQMXU\ -RXUQDO RI $SSOLHG %HKDYLRU $QDO\VLV 7URVWHU + %UDPEULQJ 0 t %HHOPDQQ $ f 3UHYDOHQFH DQG VLWXDWLRQDO FDXVHV RI VWHUHRW\SHG EHKDYLRUV LQ EOLQG LQIDQWV DQG SUHVFKRROHUV -RXUQDO RI $EQRUPDO &KLOG 3V\FKRORJ\ 7XUQHU : 5HDORQ 5 ( ,UYLQ t 5RELQVRQ ( f 7KH HIIHFWV RI LPSOHPHQWLQJ SURJUDP FRQVHTXHQFHV ZLWK D JURXS RI LQGLYLGXDOV ZKR HQJDJHG LQ VHQVRU\ PDLQWDLQHG KDQG PRXWKLQJ 5HVHDUFK LQ 'HYHORSPHQWDO 'LVDELOLWLHV 9DXJKDQ 0 ( t 0LFKDHO / f $XWRPDWLF UHLQIRUFHPHQW $Q LPSRUWDQW EXW LJQRUHG FRQFHSW %HKDYLRULVP

PAGE 63

9ROOPHU 7 5 0DUFXV % $ mIH /H%ODQF / f 7UHDWPHQW RI VHOILQMXU\ DQG KDQG PRXWKLQJ IROORZLQJ LQFRQFOXVLYH IXQFWLRQDO DQDO\VHV -RXUQDO RI $SSOLHG %HKDYLRU $QDO\VLV 9ROOPHU 7 5 0DUFXV % $ 5LQJGDKO ( IH 5RDQH + 6 f 3URJUHVVLQJ IURP EULHI DVVHVVPHQWV WR H[WHQGHG H[SHULPHQWDO DQDO\VHV LQ WKH HYDOXDWLRQ RI DEHUUDQW EHKDYLRU -RXUQDO RI $SSOLHG %HKDYLRU $QDO\VLV :DFNHU 6WHHJH 0 1RUWKXS 5HLPHUV 7 %HUJ : mIH 6DVVR f 8VH RI IXQFWLRQDO DQDO\VLV DQG DFFHSWDELOLW\ PHDVXUHV WR DVVHVV DQG WUHDW VHYHUH EHKDYLRU SUREOHPV $Q RXWSDWLHQW FOLQLF PRGHO ,Q $ & 5HSS mIH 1 1 6LQJK (GVf 3HUVSHFWLYHV RQ WKH XVH RI QRQDYHUVLYH DQG DYHUVLYH LQWHUYHQWLRQV IRU SHUVRQV ZLWK GHYHORSPHQWDO GLVDELOLWLHV SS f 6\FDPRUH ,/ 6\FDPRUH 3XEOLVKLQJ &R :DOODFH 0 t ,ZDWD % $ f (IIHFWV RI VHVVLRQ GXUDWLRQ RQ IXQFWLRQDO DQDO\VLV RXWFRPHV -RXUQDO RI $SSOLHG %HKDYLRU $QDO\VLV

PAGE 64

%,2*5$3+,&$/ 6.(7&+ (LOHHQ 5RVFRH ZDV ERP LQ 0LDPL )/ 6KH HQWHUHG WKH 8QLYHUVLW\ RI )ORULGD LQ DQG JUDGXDWHG LQ ZLWK D GHJUHH LQ SV\FKRORJ\ 'XULQJ KHU XQGHUJUDGXDWH VWXGLHV VKH HQUROOHG LQ WKH DSSOLHG EHKDYLRU DQDO\VLV ODE DQG JDLQHG KHU ILUVW H[SRVXUH WR EHKDYLRU DQDO\VLV DQG LQGLYLGXDOV GLDJQRVHG ZLWK GHYHORSPHQWDO GLVDELOLWLHV )ROORZLQJ WKLV VKH VHUYHG DV D YROXQWHHU DW WKH )ORULGD &HQWHU RQ 6HOI,QMXU\ ZKHUH VKH JDLQHG H[SHULHQFH FRQGXFWLQJ DVVHVVPHQWV DQG WUHDWPHQWV RI VHYHUH EHKDYLRU GLVRUGHUV +HU H[SHULHQFH DW WKH )ORULGD &HQWHU RQ 6HOI,QMXU\ OHG KHU WR VHHN JUDGXDWH WUDLQLQJ LQ WKH ILHOG RI EHKDYLRU DQDO\VLV ,Q WKH )DOO RI VKH HQUROOHG LQ WKH EHKDYLRU DQDO\VLV SURJUDP DW WKH 8QLYHUVLW\ RI )ORULGD WR SXUVXH D GRFWRUDO GHJUHH LQ SV\FKRORJ\ 6KH ZRUNHG DV D JUDGXDWH UHVHDUFK DVVLVWDQW DW WKH )ORULGD &HQWHU RQ 6HOI,QMXU\ IURP WR XQGHU WKH GLUHFWLRQ RI 'U %ULDQ ,ZDWD ([SHFWLQJ JUDGXDWLRQ LQ $XJXVW (LOHHQ KDV DFFHSWHG DQ DSSRLQWPHQW DV D SRVWGRFWRUDO IHOORZ DW WKH 0DUFXV ,QVWLWXWH D KRVSLWDO ORFDWHG LQ $WODQWD *$ DQG DIILOLDWHG ZLWK (PRU\ 8QLYHUVLW\ 6KH SODQV WR IXUWKHU KHU FDUHHU LQ DSSOLHG EHKDYLRU DQDO\VLV E\ FRQGXFWLQJ FOLQLFDO ZRUN DQG UHVHDUFK LQ WKH DSSOLFDWLRQ RI EHKDYLRUDO SULQFLSOHV WR WKH WUHDWPHQW RI EHKDYLRU GLVRUGHUV

PAGE 65

, FHUWLI\ WKDW KDYH UHDG WKLV VWXG\ DQG WKDW LQ P\ RSLQLRQ LW FRQIRUPV WR DFFHSWDEOH VWDQGDUGV RI VFKRODUO\ SUHVHQWDWLRQ DQG LV IXOO\ DGHTXDWH LQ VFRSH DQG TXDOLW\ DV D GLVVHUWDWLRQ IRU WKH GHJUHH RI 'RFWRU RI 3KLORVRSK\ 3URIHVVRU RI 3V\FKRORJ\ FHUWLI\ WKDW KDYH UHDG WKLV VWXG\ DQG WKDW LQ P\ RSLQLRQ LW FRQIRUPV WR DFFHSWDEOH VWDQGDUGV RI VFKRODUO\ SUHVHQWDWLRQ DQG LV IXOO\ DGHTXDWH LQ VFRSH DQG TXDOLW\ DV D GLVVHUWDWLRQ IRU WKH GHJUHH RI 'RFWRU RI 3KLORVRSK\ \ ? RWKFA -HQQLIHU 0 $VPXV $VVLVWDQW 3URIHVVRU RI (GXFDWLRQDO 3V\FKRORJ\ FHUWLI\ WKDW KDYH UHDG WKLV VWXG\ DQG WKDW LQ P\ RSLQLRQ LW FRQIRUPV WR DFFHSWDEOH VWDQGDUGV RI VFKRODUO\ SUHVHQWDWLRQ DQG LV IXOO\ DGHTXDWH LQ VFRSH DQG TXDOLW\ DV D GLVVHUWDWLRQ IRU WKH GHJUHH RI 'RFWRU RI 3KLORVRSK\ 8QL 7LP 9VV SWK\ MLL NFLDWH 3URI U79?@ NHU VVVR 9IFHUJ U RI 3V\FKRORJ\ FHUWLI\ WKDW KDYH UHDG WKLV VWXG\ DQG WKDW LQ P\ RSLQLRQ LW FRQIRUPV WR DFFHSWDEOH VWDQGDUGV RI VFKRODUO\ SUHVHQWDWLRQ DQG LV IXOO\ DGHTXDWH LQ VFRSH DQG TXDOLW\ DV D GLVVHUWDWLRQ IRU WKH GHJUHH RI 'RFWRU RI 3KLORVRSK\ 0DUN + /HZLV 3URIHVVRU RI 3V\FKRORJ\

PAGE 66

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

PAGE 67

/' RMBB 81,9(56_7< RI )/25,'$


xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID E905C6YRL_P0GJ78 INGEST_TIME 2014-05-24T00:14:04Z PACKAGE AA00020463_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES


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.
11

TABLE OF CONTENTS
page
ACKNOWLEDGMENTS 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
Method 16
Results and Discussion 20
EXPERIMENT 2: FUNCTIONAL ANALYSIS OF HAND MOUTHING 23
Method 23
Results and Discussion 24
EXPERIMENT 3: TREATMENT 29
Participants and Setting 30
Data Collection and Reliability 30
Phase I: Functional Analysis 30
Phase II: Preference Assessments 33
Results and Discussion for Phase II 34
Phase III: Treatment Evaluation 36
Results and Discussion for Phase III 39
DISCUSSION 50
REFERENCES 54
BIOGRAPHICAL SKETCH 59
in

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.
IV

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 participant(s), respectively.
v

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
1

2
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 discriminations 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

3
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)

4
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, l-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.

5
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

6
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,

7
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

8
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

9
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

10
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 HM 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

11
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

12
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

13
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

14
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
16

17
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
0
0.0%
11-20
3
2.3%
21-30
21
16.2%
31-40
37
28.5%
41-50
49
37.7%
51+
20
15.4%
Degree of retardation
Mild/moderate
4
3.1%
severe
2
1.5%
profound
124
95.4%
Impairments
visual
43
33.1%
auditory
38
29.2%
physical
47
36.2%
Receptive verbal skills
none
15
11.5%
minimal
69
53.1%
follows directions
46
35.4%
Expressive verbal skills
none
103
79.2%
minimal
4
3.1%
one-word utterances
9
6.9%
Two-to-ten-word utterances
8
6.2%
>10 word utterances
6
4.6%

18
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

19
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

20
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%.

21
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/A11 observed HM
50/82
61.0%
Severitv Ratings (N=1301
Overall prevalence of severe HM
29/830
3.5%
Severe HM/A11 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

22
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
verily 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=r64) 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)
23

24
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

PERCENTAGE OF INTERVALS HAND MOUTHING
25
10
15
1>\ /~-
- •
—•— Alone
Attention
—o Play
_
—a Demand
^ D A
1 1 1
Pattern 2 (45.3%)
Pattern 4 (1.6% 1
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

26
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

27
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. (198271994a).

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
29

30
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

PERCENTAGE OF INTERVALS
31
SESSIONS
Figure 2: Percentage of intervals HM during the functional analysis from Experiment 2
for all Experiment 3 participants. See text for details.

32
Table 3
FA patterns for all participants during the functional analysis from Experiment 2
Participant FA Pattern
Deanna
Gina
Dan
Ted
Andy
Beth
Kate
Lynn
Avery
Deborah
James
Keri
Mike
Lisa
High in alone, lower levels in attention and demand
High in alone, lower levels in attention and demand
High in alone, lower levels in attention and demand
High in alone, lower levels in attention and demand
High in alone, lower levels in attention and demand
High in alone, lower levels in attention and demand
High and undifferentiated across all conditions
High and undifferentiated across all conditions
High and undifferentiated across all conditions
Extended alone and demand sessions; not differentially high in demand
Extended alone and demand sessions; not differentially high in demand
Extended alone and demand sessions; not differentially high in demand
Extended alone, attention, play, and demand sessions; undifferentiated
across conditions
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

33
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 aggreement 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%).

34
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

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)
PERCENTAGE OF TRIALS (APPROACH)
o
o
o
FIG NEWTON
GRAHAM SNACK ^
COOKIE w
2
§
DURATION OF ITEM CONTACT
KJ
O
-L-
O
L
u>
o
J
BEADS

36
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 and 2.
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.

37
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

38
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.

39
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

PERCENTAGE OF INTERVALS
40
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).

80-
60-
40-
20-
0-
100
80
60
40
20
0
100
80
60
40
20
0
100
80
60
40
20
0
100
80
60
40
20
0
41
BL NCR
NCR + DRA + Blocking
•—•—•—•
1
T
5
H
25
: 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)

42
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

PERCENTAGE OF INTERVALS
43
BL NCR
NCR+R block
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)

44
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.

45
BL NCR NCR+R BLOCK BL NCR+RESTRAINT BL
Figure 6: Percentage of intervals HM or HM and item contact during baseline, NCR,
NCR + response blocking, and NCR + brief restraint conditions for Avery

46
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

47
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 (noncontingent access to toys and social interaction,
escape extinction, praise and a sip of liquid contingent on acceptance of food) were in

48
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 identity 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

49
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
50

51
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.,

52
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.

53
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.
54

55
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 persistance 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.

56
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.

57
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 enviommental 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.

58
Vollmer, T. R., Marcus, B. A., «fe 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., 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., Use of functional analysis and acceptability measures to assess and treat severe behavior
problems: An outpatient clinic model. In A. C. Repp «fe 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 bom 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.
59

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.
Brian A. Iwata, (/nair
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.
y \ oth.c^
Jennifer 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.
/
Uni
A
Tim
Vss
pthy D. jii
kriate Prof
iiker
ssso
Vfcerg /
r of Psychology1
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.
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
1780
20oj__
.IWSL
UNIVERSITY OF FLORIDA
3 1262 08555 0308