<%BANNER%>

Effects of Signals on the Maintenance of Alternative Behavior under Intermittent Reinforcement

Permanent Link: http://ufdc.ufl.edu/UFE0021351/00001

Material Information

Title: Effects of Signals on the Maintenance of Alternative Behavior under Intermittent Reinforcement
Physical Description: 1 online resource (55 p.)
Language: english
Creator: Dempsey, Carrie Melissa
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2007

Subjects

Subjects / Keywords: behavior, communication, differential, discrimination, extinction, functional, reinforcement, signal
Psychology -- Dissertations, Academic -- UF
Genre: Psychology thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: A common approach to the treatment of problem behavior is to strengthen an alternative, appropriate behavior. Although continuous reinforcement is required during acquisition, the practicalities of maintenance require intermittent reinforcement. Previous methods for thinning reinforcement for alternative behavior (i.e., delay fading), however, have resulted in response deterioration. Results of several basic studies have shown that exteroceptive stimuli (signals) may be used to control the rate and temporal distribution of responding when reinforcement is intermittent or delayed; however, only one applied study has examined the use of signals for maintaining alternative behavior during reinforcement-schedule thinning. The purpose of this study was to evaluate patterns of responding when reinforcer availability was signaled by visual stimuli and delivered according to a variable-interval (VI) schedule. Study 1 involved a basic demonstration of the effects of signaled and unsignaled reinforcer availability under a VI schedule. Study 2 involved a clinical evaluation of signaled, VI reinforcement as a maintenance procedure for alternative behavior in the context of treatment for severe problem behavior. Results of Study 1 revealed that the signaled-VI schedule was associated with moderate rates of responding, whereas the unsignaled schedule tended to produce either high or zero rates of responding. Results of Study 2 replicated and extended the results of Study 1,showing that the signaled-VI schedule produced low rates of problem behavior and moderate rates of alternative behavior as the reinforcement schedule was thinned.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Carrie Melissa Dempsey.
Thesis: Thesis (Ph.D.)--University of Florida, 2007.
Local: Adviser: Iwata, Brian.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2007
System ID: UFE0021351:00001

Permanent Link: http://ufdc.ufl.edu/UFE0021351/00001

Material Information

Title: Effects of Signals on the Maintenance of Alternative Behavior under Intermittent Reinforcement
Physical Description: 1 online resource (55 p.)
Language: english
Creator: Dempsey, Carrie Melissa
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2007

Subjects

Subjects / Keywords: behavior, communication, differential, discrimination, extinction, functional, reinforcement, signal
Psychology -- Dissertations, Academic -- UF
Genre: Psychology thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: A common approach to the treatment of problem behavior is to strengthen an alternative, appropriate behavior. Although continuous reinforcement is required during acquisition, the practicalities of maintenance require intermittent reinforcement. Previous methods for thinning reinforcement for alternative behavior (i.e., delay fading), however, have resulted in response deterioration. Results of several basic studies have shown that exteroceptive stimuli (signals) may be used to control the rate and temporal distribution of responding when reinforcement is intermittent or delayed; however, only one applied study has examined the use of signals for maintaining alternative behavior during reinforcement-schedule thinning. The purpose of this study was to evaluate patterns of responding when reinforcer availability was signaled by visual stimuli and delivered according to a variable-interval (VI) schedule. Study 1 involved a basic demonstration of the effects of signaled and unsignaled reinforcer availability under a VI schedule. Study 2 involved a clinical evaluation of signaled, VI reinforcement as a maintenance procedure for alternative behavior in the context of treatment for severe problem behavior. Results of Study 1 revealed that the signaled-VI schedule was associated with moderate rates of responding, whereas the unsignaled schedule tended to produce either high or zero rates of responding. Results of Study 2 replicated and extended the results of Study 1,showing that the signaled-VI schedule produced low rates of problem behavior and moderate rates of alternative behavior as the reinforcement schedule was thinned.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Carrie Melissa Dempsey.
Thesis: Thesis (Ph.D.)--University of Florida, 2007.
Local: Adviser: Iwata, Brian.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2007
System ID: UFE0021351:00001


This item has the following downloads:


Full Text





EFFECTS OF SIGNALS ON THE MAINTENANCE OF ALTERNATIVE BEHAVIOR
UNDER INTERMITTENT REINFORCEMENT




















By

CARRIE MELISSA DEMPSEY


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

2007

































O 2007 Carrie Melissa Dempsey


































To my grandparents, Ernie and Betty Moretti









ACKNOWLEDGMENTS

I would like to thank my advisor, Dr. Brian A. Iwata, whose enduring dedication, support,

and guidance have been instrumental to my professional development. Also, I would like to

thank the members of my committee: Drs. Scott Miller, Stephen Smith, and Timothy Vollmer,

for their input and direction during this process. I would like to thank my colleagues, particularly

Pam Neidert and Jessica Thomason, for shaping my clinical skills and for teaching me the

nuances of behavioral assessment and treatment. Finally, I would like to thank my parents and

grandparents, whose consistent, common sense approach to raising children cultured my basic

views on behavior, and predisposed me to a career in applied behavior analysis.












TABLE OF CONTENTS


page

ACKNOWLEDGMENTS .............. ...............4.....


LIST OF TABLES ................ ...............7............ ....


LIST OF FIGURES .............. ...............8.....


AB S TRAC T ......_ ................. ............_........9


CHAPTER


1 FUNCTIONAL COMMUNICATION TRAINING ................. .............. ......... .....11


Functi on-B ased Treatment. .............. ... .... ........ ...............11......

Applications of Functional Communication Training ................. ............... ......... ...12
Component Analyses of FCT .............. ...............13....

2 MAINTENANCE FOLLOWING FCT ................. ......... ...............15. ....


T he Natural Environment ................. ...............15................

D elay Fading................. ..... .............1
Fixed-Interval Schedule Thinning ................. ...............17................


3 SIGNALS .............. ...............18....


4 STUDY 1: COMPARISON OF UNSIGNALED VERSUS SIGNALED VI
SCHEDULES .............. ...............24....


M ethod ............... ... ........ ..... ...............24.......
Participants and Setting ................... ............ ...............24......
Response Measurement and Reliability .............. ...............24....
Stimulus Preference Assessment ................. ...............25........... ....
Procedure ................. ...............25.................
Re sults ................ ...............26.................


5 STUDY 2: CLINICAL EVALUATION OF SIGNALED VI SCHEDULE THINNING .....30


M ethod ............... ... ........ ..... ...............3.. 0....
Participants and Setting ................... ............ ...............3.. 0....
Response Measurement and Reliability .............. ...............31....
Safety Precautions .............. ...............33....
Procedure ............... .. ... .......... .. ........ .............3
Functional analysis (baseline) ................... ...............33..
Functional communication training (FCT) .............. ...............35....
Signal discrimination assessment............... ...............3













Signaled VI schedule thinning............... ...............36
Re sults............. ...... ._ ...............37...


6 DI SCUS SSION ............ _...... ...............44...


LIST OF REFERENCES ............_...... ...............51...


BIOGRAPHICAL SKETCH .............. ...............55....










LIST OF TABLES


Table


page


5-1 Operational definitions............... ..............4










LIST OF FIGURES


Figure page

4-1 Responses per minute on the arbitrary task under the unsignaled and signaled VI-30"
schedules ...._ ................. ...............29.......

5-1 Rate of problem behavior and communication during baseline (CR v. NCR), FCT,
and during signaled VI thinning for Amanda and Jennifer ................. ............ .........40

5-2 Rate of problem behavior and communication during baseline (CR v. NCR), FCT,
and during signaled VI thinning for Robert and Shelly ........................... ...............41

5-3 Rate of problem behavior and communication during baseline (CR v. NCR), FCT,
and during signaled VI thinning for Mandy and Elaine .............. ....................4









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

EFFECTS OF SIGNALS ON THE MAINTENANCE OF ALTERNATIVE BEHAVIOR
UNDER INTERMITTENT REINFORCEMENT

By

Carrie Melissa Dempsey

August 2007

Chair: Brian A. Iwata
Major: Psychology

A common approach to the treatment of problem behavior is to strengthen an alternative,

appropriate behavior. Although continuous reinforcement is required during acquisition, the

practicalities of maintenance require intermittent reinforcement. Previous methods for thinning

reinforcement for alternative behavior (i.e., delay fading), however, have resulted in response

deterioration. Results of several basic studies have shown that exteroceptive stimuli (signals)

may be used to control the rate and temporal distribution of responding when reinforcement is

intermittent or delayed; however, only one applied study has examined the use of signals for

maintaining alternative behavior during reinforcement-schedule thinning.

The purpose of this study was to evaluate patterns of responding when reinforcer

availability was signaled by visual stimuli and delivered according to a variable-interval (VI)

schedule. Study 1 involved a basic demonstration of the effects of signaled and unsignaled

reinforcer availability under a VI schedule. Study 2 involved a clinical evaluation of signaled, VI

reinforcement as a maintenance procedure for alternative behavior in the context of treatment for

severe problem behavior. Results of Study 1 revealed that the signaled-VI schedule was

associated with moderate rates of responding, whereas the unsignaled schedule tended to

produce either high or zero rates of responding. Results of Study 2 replicated and extended the










results of Study 1,showing that the signaled-VI schedule produced low rates of problem behavior

and moderate rates of alternative behavior as the reinforcement schedule was thinned.









CHAPTER 1
FUNCTIONAL COMMUNICATION TRAINING

Function-Based Treatment

The development of methods for identifying behavioral function, such as the functional

analysis (Iwata, Dorsey, Slifer, Bauman, & Richman, 1982), spurred a new generation of

treatment for problem behavior, commonly referred to as function-based treatment. In function-

based treatment, the variables that maintain problem behavior are identified and then are used to

inform the course of intervention. Function-based treatment may be applied in several ways. One

application involves extinction (EXT), in which the functional reinforcer is eliminated following

problem behavior. For example, Iwata, Kalsher, Cowdery, and Cataldo (1990) showed that the

self-injurious behavior (SIB) of 7 participants with developmental disabilities was maintained by

escape from task demands. Extinction was implemented by terminating escape following SIB,

and resulted in decreases in SIB to near zero levels.

Another application of function-based treatment involves noncontingent reinforcement

(NCR), in which the functional reinforcer is delivered on a time-based schedule that is

independent of responding. For example, Vollmer, Iwata, Zarcone, Smith, and Mazaleski (1993)

identified that the SIB of 3 participants with mental retardation was maintained by access to

attention. During NCR, attention was delivered initially on a dense, fixed-time schedule, which

was subsequently thinned. All participants showed immediate reductions in behavior, which

were sustained over time.

A third application of function-based treatment involves differential reinforcement of an

alternative behavior (DRA), in which the functional reinforcer is withheld following problem

behavior, and used to establish an alternative, competing response. DRA has been used to

increase compliance in participants with escape-maintained problem behavior (e.g., Lalli et al.,










1999) and leisure item engagement in participants who engage in automatically reinforced

behavior (Vollmer, Marcus, & LeBlanc, 1994). In addition, and of particular relevance to the

current study, DRA has been used to establish appropriate communication (e.g., Wacker et al.,

1990). In the context of communication, a common term given to DRA is Functional

Communication Training (FCT; Carr & Durand, 1985), although the basic structure of the

intervention remains unchanged. The following sections review topics relevant to the use and

effectiveness of FCT.

Applications of Functional Communication Training

Carr and Durand (1985) conducted the first application ofFCT with 4 participants who

engaged in disruption in a classroom setting. Antecedent variables were manipulated to generate

information regarding the events that reinforced problem behavior. Once reinforcers were

identified, participants were taught to emit two statements in response to an experimenter

prompt: a relevant statement that produced access to the reinforcer, and an irrelevant statement

that produced access to an arbitrary consequence. Throughout the evaluation, problem behavior

resulted in EXT. When relevant statements were reinforced, decreases in problem behavior and

increases in communication were observed for all participants. However, when irrelevant

statements produced access to an arbitrary consequence, treatment effects were not observed.

Since its initial application, the effectiveness of FCT has been demonstrated across

several behaviors including aggression (Thompson, Fisher, Piazza, & Kuhn, 1998), SBE (Wacker

et al. 1990), elopement (Tarbox, Wallace, & Williams, 2003), inappropriate sexual behavior

(Fyffe, Kahng, Fittro, & Russell, 2004), stereotypy (e.g., Durand & Carr, 1987), breath holding

(Kern, Mauk, Marder, & Mace, 1995), and off-task behavior (Meyer, 1999). Given the positive

treatment outcomes associated with FCT, studies have examined the components of FCT that

contribute to its effectiveness.










Component Analyses of FCT

Carr and Durand (1985) and Wacker et al. (1990) suggested that FCT differs from other

differential reinforcement procedures (e.g., DRO), in that it allows the participant to control the

delivery of reinforcement. That is, the participant may engage in communication to produce the

reinforcer at any time. Kahng, Iwata, Deleon, and Worsdell (1997) investigated the contributing

role of control over reinforcement by comparing responding under two procedures, one in which

participants' behavior influenced reinforcer delivery (FCT), and one in which participants'

behavior did not influence reinforcer delivery (NCR). The amount of reinforcement was yoked

across the two procedures. Both procedures produced comparable decreases in behavior,

indicating that control over reinforcement by a participant did not influence treatment effects.

In addition, investigators have examined the role of EXT in FCT. For example,

Hagopian, Fisher, Sullivan, Acquisto, and LeBlanc (1988) conducted a large-scale analysis of

FCT and found that, in the absence of EXT, problem behavior was not reduced by 90% for any

of 11 participants. Worsdell, Iwata, Hanley, Thompson, and Kahng (2000) evaluated the role of

EXT during the acquisition of alternative behavior for 5 participants whose problem behavior

was maintained by social positive reinforcement. When participants were exposed to an initial

phase in which both problem behavior and alternative behavior were reinforced, only 1

participant acquired the alternative response. The schedule of reinforcement for problem

behavior was made more intermittent for the remaining 4 participants to determine the point at

which more responding would be allocated to alternative behavior than to problem behavior.

Switching to alternative behavior did not occur until reinforcement for its occurrence was 2, 3,

and even 20 times more likely than for problem behavior.

Shirley, Iwata, Kahng, Mazeleski, and Lerman (1997) examined the influence of EXT

during both the acquisition and maintenance of manual signs during FCT for 3 participants.









When SBE continued to produce reinforcement, SBE remained at baseline rates, and

communication was not acquired. The addition of EXT, however, was associated with decreases

in SIB and increases in signing for all participants.

Collectively, these results suggest that EXT is a critical component of FCT; however,

some research suggests that FCT may also be conducted with high integrity in the absence of

EXT if the establishing operation for behavior (e.g., Smith, Iwata, Goh, & Shore, 1995), or

parameters of reinforcement for appropriate and inappropriate responses are manipulated

(Horner et al., 1980; Vollmer et al., 1999). Aside from examining the relative effects of FCT

components, a small number of studies have investigated methods for maintaining treatment

effects following FCT. These methods are reviewed briefly in Chapter 2.









CHAPTER 2
MAINTENANCE FOLLOWING FCT

The Natural Environment

During FCT sessions, communication is typically reinforced on a fixed-ratio 1 (FR-1)

schedule to facilitate response acquisition, which tends to engender high and stable rates of

communication (Carr & Durand, 1985). However, reinforcement for communication may be

delayed or intermittent in the natural environment. Some examples: A busy parent may have to

direct attention to other individuals or activities, some tangible items (e.g., edibles) are unhealthy

if provided in large quantities, and some demands (e.g., hygiene tasks) require compliance.

If communication occurs frequently but fails to contact reinforcement, it may extinguish.

Subsequently, other behavior that has produced reinforcement in the past (i.e., problem behavior)

may re-emerge. In support of this, Durand and Carr (1991) observed a breakdown in

communication and re-emergence in problem behavior when treatment was implemented in the

natural environment for 1 of 3 participants. Direct observations, which were conducted during

follow up, revealed that the teacher was not attending to the participant's requests for assistance.

Relatively few strategies have been designed to maintain communication under

conditions of intermittent or delayed reinforcement. Two such strategies are delay fading (Fisher,

Thompson, Hagopian, Bowman, & Krug, 2000) and fixed-interval (FI) schedule thinning

(Hanley, Iwata, & Thompson, 2001).

Delay Fading

Delay fading involves gradually lengthening the delay between communication and

reinforcement. Typically, interval increases are determined based on levels of responding during

preceding sessions. Three applied studies have examined the effects of delay fading on

responding following FCT and have obtained comparable results. Fisher et al. (2000) evaluated










progressively increasing delays to reinforcement for communication and found that additional

treatments (e.g., providing tasks during delay) were required to maintain treatment effects once

the delay interval was increased to 30 s. As part of a larger study, Hanley et al. (2001) examined

delay fading by systematically increasing the delay to reinforcement for communication from 1 s

to 25 s, and found that communication decreased to low rates under the 25-s delay, at which

point problem behavior increased.

Hagopian et al. (1998) conducted a large-scale study to assess the outcome of FCT alone

or in combination with other procedures (e.g., EXT) with 21 clients who engaged in problem

behavior such as SBE, aggression, and property destruction. Functional analyses revealed that

problem behavior was maintained by access to attention for 9 participants, by escape from

demands for 7 participants, by access to tangible items for 1 participant, and by multiple sources

of control for 4 participants. Demand fading or delay-to-reinforcement fading was conducted in

12 of 25 applications of FCT with EXT, but reductions in problem behavior were maintained

after fading in only 5 applications. Further, attempts to regain control over the behavior in the

remaining 7 cases (e.g., reducing delay interval or number of demands) were unsuccessful.

The results of these studies suggest that a newly acquired response may deteriorate under

delayed reinforcement, a conclusion that is also supported by basic research (Azzi, Fix, Keller, &

Rocha E Silva, 1964; Schaal & Branch, 1988; Sizemore & Lattal, 1978). Decrements in

responding under delay may reflect a decrease in contingency strength, which is determined by

the temporal contiguity between response and reinforcement (Baum, 1973) and the probability of

reinforcement for a response relative to the response-independent probability of reinforcement

(Hammond, 1980; Lattal, 1995). Under conditions of delay, the temporal contiguity between a

response and reinforcer is eroded. For example, Sizemore and Lattal (1978) showed that










response rates decrease systematically with obtained (actual) delays to reinforcement, in addition

to nominal (programmed) delays.

Fixed-Interval Schedule Thinning

Under an FI reinforcement schedule, a reinforcer is delivered for the first response that

occurs after an interval of time has elapsed since the last reinforced response (Ferster & Skinner,

1957). FI schedule thinning occurs by gradually lengthening the time interval across sessions,

based on levels of responding during preceding sessions (Hanley et al., 2001). In this

arrangement, the contingency between a response and reinforcer may be preserved given the

high probability of immediate reinforcement for communication.

Hanley et al. (2001) implemented FI thinning by gradually increasing the interval length

from 1 s to 58 s. Results showed that problem behavior remained low throughout thinning.

However, communication gradually increased to rates that were higher than those observed

during initial training, and only a small proportion of communication responses resulted in

reinforcement.

Elevated rates of communication during FI schedule thinning suggest that the contingency

was preserved. On the other hand, a large proportion of communication responses occurred when

reinforcement was not available. Persistence of high rates of unreinforced responding over time

may degrade the contingency between the response and reinforcement, causing a break down in

responding (Hammond, 1980).

The results of studies on delay fading and FI schedule thinning illustrate that

communication may not occur at appropriate levels under intermittent reinforcement if

reinforcement is delayed or difficult to discriminate. As such, the identification of strategies that

enhance reinforcer immediacy and discrimination is warranted. Research relevant to the

development of one strategy is addressed in the next chapter










CHAPTER 3
SIGNALS

In basic and applied research, exteroceptive stimuli, or signals, have been used to control

the occurrence of responding under conditions in which reinforcement is delayed (Kelley, 2003;

Lattal, 1984; Morgan, 1972; Pierce, Hanford, & Zimmerman, 1972; Richards, 1981; Schaal &

Branch, 1988, 1990; Williams, 1976) or intermittent (Fisher, Kuhn, & Thompson, 1998; Hanley

et al., 2001).

Several basic studies have shown that response rates may be maintained under delayed

reinforcement when the response that begins the delay also produces a change in visual

stimulation, such as a blackout or flashing light. For example, Azzi et al. (1964) trained three rats

on a schedule of continuous reinforcement before exposing them to unsignaled reinforcement

delays of 1, 3, 5, 7.5, 10, 15, and 20 s. Results showed that response rates decreased

systematically for all 3 subj ects. Subj ects were then exposed to a reinforcement delay of 20 s and

then 30 s for several days. Under each reinforcement delay, half of the sessions involved an

unsignaled reinforcement delay, and half of the sessions involved signaled reinforcement delay,

in which responses produced a blackout period (chamber light off) that endured until

reinforcement was delivered. The signaled delay was associated with high response rates at both

interval values, whereas the unsignaled delay was associated with low response rates.

Richards (1981) examined the influence of signals using two reinforcement schedules that

were expected to produce similar response rates but disparate reinforcer rates under

reinforcement delay: a VI 60-s schedule and a differential-reinforcement-of-low-rate (DRL) 20-s

schedule. Four pigeons were trained on the VI-60 s schedule, and 5 pigeons were trained on the

DRL schedule, following which all subj ects were exposed to 48, 1-hour sessions with the

following reinforcement delays: 10 s, 5 s, 2.5 s, 1 s, .5 s. Half of the subjects were exposed to









signaled delays (blackout) in the first half of sessions and unsignaled delays in the second half of

sessions. The other subj ects were exposed to the reverse order. Although higher rates of

responding were observed under unsignaled rather than signaled delays at short delay values

(e.g., 1 s), large decreases in response rates under unsignaled delays and little reduction relative

to baseline under signaled delays was observed at longer delay values.

Schaal and Branch (1988) compared the effects of briefly signaled versus continuously

signaled delay intervals with three pigeons trained on a VI-60 s schedule. Under briefly signaled

delays, a .5 s change in key light color from red to green occurred immediately following a key

peck that initiated the delay. Under continuously signaled delays, the key light color changed to

green, but remained green throughout the delay and during delivery of the reinforcer. Brief

signals maintained response rates under delays up to 9 s; however, continuous signals were

required to maintain responding at delays of 27 s. In a follow-up study, Schaal and Branch

(1990) showed that key-pecking rate was an increasing function of the duration of the key light

that signaled a 27-s delay.

Kelley (2003) extended findings on signaled reinforcement delay from basic research to

application with 3 participants diagnosed with developmental disabilities who were taught

communicative responses. During signaled delays, a visual (timer that tracked the delay) or

auditory (shaking a can of money) stimulus was presented continuously during the delay interval.

During unsignaled delays, no stimuli were presented. Communication maintained under

progressively longer delays when the delay period was signaled but not when it was unsignaled.

Vollmer, Borrero, Lalli, and Daniel (1999) investigated the effects of signals within a

choice paradigm in which, following FCT, problem behavior produced access to a smaller,

immediate reinforcer (e.g., one chip immediately after aggression), whereas communication










produced access to a larger, delayed reinforcer (e.g. three chips after 10 s). When the delay to the

large reinforcer was unsignaled, participants made impulsive choices (chose the smaller,

immediate reinforcer); when the delay to the large reinforcer was signaled (e.g., placement of

timer in front of the participant), participants engaged in self-control (chose the larger, delayed

reinforcer).

The mechanisms) by which signals maintain responding during reinforcement delay has

not been identified definitively. However, it is often speculated that response maintenance may

be a function of either conditioned reinforcement or discrimination (Richards, 1981; Schaal &

Branch, 1990). From a conditioning standpoint, a signal may acquire reinforcing value due to its

close temporal relationship with the delivery of food reinforcement. From the standpoint of

discrimination, nonresponding comes under discriminative control of the delay signal, and is

therefore less likely to influence responding at other points in time.

In applied research, signals have been programmed during periods of reinforcement and

EXT to facilitate discrimination of changing conditions. Specifically, signals function as

discriminative stimuli due to a differential correlation of reinforcement in their presence relative

to their absence (Michael, 1982). Fisher et al. (1998), for example, trained participants to use

communication responses to access two different sources of positive reinforcement (e.g., toys) in

the context of FCT. During training, each response was paired with a distinct stimulus (e.g.,

picture of a person playing with toys). Following training, reinforcement contingencies and

associated stimuli were alternated every 30 min, and participants displayed only the response that

would be reinforced during that period, suggesting that the signals had become discriminative for

each response.









Hanley et al. (2001) compared the effects of a multiple schedule, in which different

stimuli signaled periods of reinforcement and EXT, to those of a mixed schedule, which

contained similar, but unsignaled periods, to evaluate the effects of thinning reinforcement for

communication following FCT. Under the multiple schedule, a white card was present when the

reinforcement component was in effect and communication produced 10-s access to a preferred

item on a FR-1 schedule, whereas a red card was present when the EXT component was in effect

and communication produced no programmed reinforcement. Initially, the component durations

were 45 s and 15 s for FR 1 and EXT, respectively. The reinforcement schedule was then thinned

to 1 min, whereas the EXT schedule was eventually increased to 4 min. Schedule thinning

progressed when problem behavior remained at or below 85% of the baseline mean for two

consecutive sessions. Throughout the assessment, problem behavior produced no programmed

consequences.

Under the mixed schedule, indiscriminate responding (i.e., high across both schedule

components) occurred across components for both participants, and problem behavior occurred

at high rates for 1 participant. Under the multiple schedule, differential responding occurred

across schedule components (i.e., low rates during EXT and moderately high rates under FR-1),

and problem behavior was maintained at low rates for both participants. The authors concluded

that discriminative stimuli, which were not present in the mixed schedule, contributed to

differential responding observed under the multiple schedule.

Collectively, this research suggests that signaled procedures may facilitate control over

the rate and temporal distribution of responding under conditions of delayed or intermittent

reinforcement. Only one study (Hanley et al., 2001) has examined the influence of signals on

responding during intermittent reinforcement following FCT. However, signals may be










particularly important in maintaining communication under such conditions. Specifically, signals

may increase the discrimination of reinforcer availability so that communication is more likely to

occur when reinforcement is available, and therefore, more likely to contact reinforcement. If the

maj ority of communication responses result in reinforcement, the contingency between

communication and reinforcement is likely to remain strong, resulting in the maintenance of

communication over time.

Moreover, signals are common components of the natural environment. For example, the

green light at the intersection indicates that accelerating through the intersection is likely to result

in obtaining safe passage to the other side. Walking through a door labeled "exit" is correlated

with successfully vacating a building. Further, the presence of a child's mother is likely to

engender a number of attention-seeking behaviors, due to the probability by which her mother

has delivered attention in the past. As such, the programming of signals to control appropriate

communication following FCT appears to be a pragmatic approach to maintaining

communication under intermittent reinforcement.

Interval reinforcement schedules, whether fixed or variable, often characterize the way in

which adaptive behavior is reinforced in the natural environment. As such, the use of interval

schedules in maintaining communication following FCT seems logical. Hanley et al. (2001)

showed that a FI schedule of reinforcement was associated with excessively high rates of

communication; however, more moderate rates of communication might have occurred if

reinforcer availability was signaled.

No previous studies have investigated the use of a VI schedule during maintenance

following FCT. Nevertheless, VI schedules reflect the manner in which several important









reinforcers are provided to individuals. For example, caregivers often provide attention following

varying durations of time given the influence of other schedule requirements.

In the current study, the effectiveness of a signaled VI schedule in maintaining

appropriate responding under intermittent reinforcement was examined. Study 1 involved a basic

operant comparison of performance under two VI-30 s schedules: one in which the availability of

reinforcement was signaled, and one in which the availability of reinforcement was unsignaled.

Demonstrating the effects of the VI schedule using arbitrary responses within a basic

arrangement prior to using the schedule clinically carries two potential advantages. First, the

responses selected for study could be made comparable so that differences obtained across the

schedules would be more likely to reflect the role of the signal, as opposed to other sources of

uncontrolled variability. Second, a basic demonstration would allow for the assessment of the VI

schedule with the signal present and absent, so that a clinical evaluation of responding under

both types of schedule would be unnecessary. As such, the risk of exposing participants to an

ineffective or even counter-therapeutic procedure might be minimized. Once a basis for using the

signaled VI schedule was established in Study 1, Study 2 was conducted, which involved a

clinical evaluation of signaled reinforcer availability under a VI schedule following response

acquisition with FCT.









CHAPTER 4
STUDY 1: COMPARISON OF UNSIGNALED VERSUS SIGNALED VI SCHEDULES

Method

Participants and Setting

Four individuals who attended a local special education school participated. Katy was a

21-year-old female who had been diagnosed with mental retardation and who had a limited

vocabulary. Charlotte was a 19-year-old nonverbal female who had been diagnosed with mental

retardation and who communicated through gestures. Lesley was a 40-year-old male, who was

diagnosed with mental retardation and communicated through vocalizations and gestures. All

participants could follow one-step instructions and were capable of performing simple tasks.

George was a 15-year-old male diagnosed with Down's syndrome, who engaged in speech at

home but not at school (i.e., was selectively mute).

Sessions were conducted in isolated areas of a treatment room located at the school.

Session areas were equipped with a table, two chairs, and target task materials (see below).

Sessions lasted for five min and were conducted 2 to 5 times per day, 3 to 5 days per week.

Response Measurement and Reliability

Target behaviors consisted of arbitrary responses that were selected based on each

participant's level of adaptive functioning. The task for Katy and Lesley consisted of pressing a

switch, which was defined by the onset of a light. Charlotte's task involved placing an index card

into a slotted bin, which was defined as the moment at which the card left Charlotte's hand.

George's task was touching his nose, and was scored when his index finger contacted any part of

his nose. Data were collected by trained observers on laptop computers during continuous 10-s

intervals, and were converted to responses per minute. A second observer simultaneously and

independently collected data during at least 27% of sessions for each participant. Interobserver










agreement was calculated based on interval-by-interval comparisons of observers' records. The

smaller number of responses scored in each interval was divided by the larger number of

responses; these fractions were then average and multiplied by 100%. Mean agreement for the

arbitrary task was 89.6% (range, 85%-100%), 93.5% (range, 83.3%-100%/), 90.7% (range,

71.7%-100%/), 91.6% (range, 70%-100%) for Katy, Charlotte, George, and Lesley, respectively.

Stimulus Preference Assessment

Paired-stimulus preference assessments (Fisher et al., 1992) were conducted to identify a

highly preferred edible item to use as a reinforcer for task completion. Preference for 9 edible

items was assessed. Prior to the assessment, participants were allowed to sample a small portion

of each item. During the assessment, items were presented in pairs, and the participant was

prompted to select one item. Each item was paired once with every other item, in a quasi-random

fashion, for a total of 36 trials. The item selected most often, and on at least 80% of trials, was

selected to serve as the reinforcer.

Procedure

The effects of unsignaled and signaled VI schedules on response rates were evaluated

using a combined multiple baseline and reversal design.

No-reinforcement baseline. An initial baseline was conducted to ensure that participants did

not engage in the target task in the absence of reinforcement. Participants were seated at a table

facing the therapist with the task positioned directly in front of them. Immediately prior to

session, the participant was prompted to engage in the response using a three-step prompting

procedures consisting of verbal, model, and physical prompts, and no programmed consequence

occurred. During the session, target responses resulted in no programmed consequences.

Fixed-ratio 1(FR-1) baseline. The FR-1 baseline served as an "acquisition" condition to ensure

that the item identified in the preference assessment functioned as a reinforcer for target










responses and to establish a baseline of responding for comparing the effects of unsignaled and

signaled VI schedules. This condition was identical to baseline except that the pre-session

prompt resulted in the delivery of the reinforcer, and target responses during the session resulted

in the delivery of the reinforcer on an FR-1 schedule.

Unsignaled VI schedule. During this condition, the seating arrangement and task position were

identical to the FR-1 baseline; however, target responses resulted in the delivery of the reinforcer

according to a VI-30 s schedule. The first response in a session was reinforced. Thereafter, a

reinforcer was delivered for the first response that occurred after a pre-determined interval of

time had elapsed since the last response was reinforced. The programmed time intervals varied

around an average value of 30 s and ranged from 15 s to 45 s. The VI 30-s schedule was

generated using Microsoft Excel RAND function, and a different schedule was used for each

session.

Signaled VI schedule. This condition was identical to the unsignaled VI schedule condition,

except that a signal (e.g., colored card or light) was presented in front of the participant as soon

as a time interval had elapsed, and remained present until the participant engaged in a target

response.

Results

Figure 4-1 shows the rate of responding on the arbitrary task during the comparison of the

unsignaled and signaled VI schedules. Katy (top panel) emitted no responses during the no-

reinforcement baseline but high rates of responding during the first FR-1 condition. Higher and

more variable rates of responding were observed under the unsignaled VI schedule, sometimes as

high as 140 rpm. Response rates stabilized during the second FR-1 phase. When the availability

of reinforcement was signaled during the second VI 30-s phase, response rates gradually

decreased until they became low and stable. An additional measure, the percentage of target









responses that resulted in reinforcement during the last ten sessions of the unsignaled and

signaled VI schedules, was calculated to determine the extent to which the signal exerted control

over responding. This measure was calculated by dividing the number of reinforcers earned in

each session by the number of responses that occurred in each session and multiplying by 100%.

The mean percentage of responses resulting in reinforcement for Katy was 5.9% in the

unsignaled VI condition and 56.7% in the signaled VI condition.

Charlotte (second panel) emitted no responses during the no-reinforcement baseline and

moderate rates of responding during the initial FR-1 phase. Her responding increased initially

under the unsignaled VI schedule but then fell to near zero rates and ceased completely during

the final nine sessions of the condition. The second FR-1 phase was associated with moderate,

variable rates of responding. During the Signaled VI Schedule condition, her responding

increased initially but then decreased. In contrast to the unsignaled VI condition, however,

Charlotte's responding decreased, but maintained at a low rate as the condition progressed. The

mean percentage of responses resulting in reinforcement was 10% in the unsignaled condition

and 98.6% in the signaled condition.

Lesley (third panel) emitted very few responses during the no-reinforcement baseline. His

responding increased during the first FR-1 phase and then decreased immediately with the

introduction of the unsignaled VI schedule, but gradually increased across the condition and

continued to increase during the second FR-1 phase. Lesley's responding dropped to moderate,

stable, rates during the signaled VI condition. The mean percentage of responses resulting in

reinforcement was 53.7% in the unsignaled condition and 100% in the signaled condition.

George (bottom panel) emitted very few responses during the no-reinforcement baseline.

His responding increased steeply and remained at high rates during both FR-1 phases.









Decreased rates of responding were observed under both the unsignaled and signaled VI

schedules; however, response rates were somewhat lower under the signaled VI schedule relative

to the unsignaled VI schedule. George's mean percentage of responses resulting in reinforcement

was 40.1% in the unsignaled condition and 67.4% in the signaled condition.

In summary, idiosyncratic patterns of responding were observed across participants

during the unsignaled VI schedule condition, and none of the participants engaged in stable,

moderate rates of responding. Further, a relatively small percentage of responses resulted in

reinforcement under the unsignaled VI schedule. By contrast, response patterns became

moderate and stable for all participants, and a large percentage of responses resulted in

reinforcement, under the signaled VI schedule condition.





























Charlotte











Lesley













George


Unsignaled
BL FR-1 VI30"


Signaled
VI 30"


150-
125-
100-
75-
50-
25-



20-

15-

10-

0 -



20 -


****


5 1 1'5 20 25 3'0 3'5 40 45 50 5'5 60

SESSIONS


Figure 4-1. Responses per minute on the arbitrary task under the unsignaled and signaled VI-30"
schedules









CHAPTER 5
STUDY 2: CLINICAL EVALUATION OF SIGNALED VI SCHEDULE THINNING

Method

The results of Study 1 indicated that the signaled VI schedule may be an effective way to

maintain desirable response patterns (low rates) under conditions of intermittent reinforcement.

Given these positive results in the context of establishing and maintaining an arbitrary response,

a clinical extension seemed warranted. Thus, Study 2 involved an evaluation of signaled VI-

schedule thinning as a maintenance procedure for adaptive behavior following its initial

acquisition as a replacement for problem behavior.

Participants and Setting

Six individuals who attended either a special education school or outpatient clinic and who

were referred for the assessment and treatment of problem behavior participated. Amanda was a

verbal 18-year-old female who was diagnosed with Prader-Willi syndrome. Her problem

behavior consisted of aggression, which staff reported to occur more frequently when preferred

items were removed or denied. Jennifer was a verbal 15-year-old female, also diagnosed with

Prader-Willi Syndrome, who engaged in severe SIB, aggression, and disruption. In several cases,

Jennifer's SBE produced open wounds and bleeding that resulted in her being temporarily

removed from the school. Robert was a verbal 15-year-old male who was diagnosed with autism

and who engaged in severe aggression directed toward teachers and other students. Shelly was a

verbal 13-year-old female who engaged in several topographies of severe SBE, aggression, and

disruption, which had caused her to be permanently removed from five previous educational

settings. Mandy was a verbal 10-year-old female who was diagnosed with Pervasive

Developmental Disorder. Her problem behaviors consisted of aggression and disruption, which

were reported to occur more often when asked to do something she did not like to do (e.g.,









schoolwork). Elaine was a nonverbal 10-year-old female with Down's syndrome who had a

history of aggression toward adults during work situations. Sessions were conducted in an

isolated room at the special education school or in a room equipped with a one-way mirror at the

outpatient clinic. Sessions were 10 min in length and were conducted 3 to 6 times a day, 1 to 5

times per week.

Response Measurement and Reliability

Data were collected on participants' problem behaviors, alternative communication

responses, and the presentation of a signal. Amanda's aggression was defined as hitting.

Jennifer's SBE consisted of skin picking, scratching, self-biting, and tampering with bandages

that covered existing wounds. Jennifer's aggression was defined as hitting, kicking, biting, and

throwing obj ects at others. Robert's aggression was defined as hitting, kicking, choking, and head

butting. Like Jennifer, Shelly's topographies of SBE included skin picking, scratching, self-

biting, and bandage tampering; however, Shelly also engaged in head banging, head hitting, teeth

banging, and nose picking. Shelly's aggression consisted of hitting, kicking, biting the body or

clothing of others, and throwing objects at others. Her disruption included throwing, breaking,

hitting, and kicking obj ects, and overturning furniture. Mandy's aggression was defined as

hitting, kicking, and biting, and her disruption was defined as throwing or breaking obj ects,

hitting or kicking surfaces, and drawing on walls. Elaine's aggression was defined as hair pulling,

scratching, and pinching. See Table 5-1 for a detailed listing of operational definitions for

problem behaviors.

Alternative responses were selected for participants based on their level of adaptive

functioning. The verbal response, "Can I have my toys back, please?" was selected for Amanda,

Robert, and Shelly. The verbal response "Can we talk?" was selected for Jennifer, and the verbal

response "Break please?" was selected for Mandy. A voice-output microswitch, which produced









"Break please," was selected as the alternative response for Elaine. All verbal responses were

scored immediately following emission.

Signals were identified individually for participants based on a signal discrimination

assessment (see details in procedure section). Signal presentation was defined as the introduction

of the signal into the participant's immediate environment. Signals consisting of colored cards

were presented either directly in front of a participant (Mandy) or on a wall (using Velcro) to the

participant's immediate left (Amanda, Robert). Colored badges were presented by attaching the

badge to the therapist's shirt (Jennifer, Shelly). The light panel, which was affixed to a wall to the

participant's immediate left, was presented by activating a switch that turned on the light.

Data were collected by trained observers on laptop computers during continuous 10-s

intervals. Data on problem behavior and communication responses were converted to a rate

measure; communication responses also were summarized as the percentage of responses that

occurred in the presence of the signal. Reliability was assessed by having a second observer

simultaneously and independently collect data during at least 27% of sessions for each

participant. Agreement percentages were calculated based on interval-by-interval comparisons of

observers' records. The smaller number of responses was divided by the larger number of

responses in each interval; these fractions were then averaged and multiplied by 100%. Mean

agreement for problem behavior was 99.6% (range, 95%-100%), 99% (range, 90%-100%),

99.3% (range, 88.1%-100%), 99.5% (range, 97.2%-100%/), 98.8% (range, 93.3%-100%/), and

98.6% (range, 92.5%-100%) for Amanda, Jennifer, Robert, Shelly, Mandy, and Elaine,

respectively. Mean reliability for alternative communication responses was 98.8% (range,

96.7%-100%/), 97.1% (range, 91.6%-100%/), 99% (range, 96.7%-100%/), 97.8% (range, 86.7%-










100%), 98.1% (range, 93.3%-100%/), and 98.9% (96.7%-100%) for Amanda, Jennifer, Robert,

Shelly, Mandy, and Elaine, respectively.

Safety Precautions

Certain measures were taken to prevent participants who engaged in SIB from harm

throughout the course of the study. First, attempts at severe SIB were physically blocked by the

therapist in the session. Second, criteria for (a) terminating sessions, and (b) seeking additional

medical assistance were set for each participant prior to admission into the study. If the

participant met the session termination criterion (e.g., presence of blood, bruising, or swelling)

during the study, he or she was immediately removed from the session area and first aid was

administered (e.g., compress or cold pack applied). If first aid did not result in a decrease in

injury, as assessed by the experimenter, the participant was brought to the school or clinic nurse

for additional medical assistance.

Procedure

A functional analysis, FCT intervention, and evaluation of maintenance under a signaled

VI schedule were conducted for each participant. Responding during the functional analysis was

evaluated within a multielement design; the effects of FCT and VI reinforcement were evaluated

in a multiple baseline design across participants.

Functional analysis (baseline)

The functional analysis was conducted using a pairwise format (Iwata, Duncan, Lerman,

Iwata, & Shore, 1994) during which responding in one test condition and one control condition

was compared and contrasted. In the test condition, the reinforcer presumed to maintain behavior

was provided contingent on the occurrence of problem behavior (contingent reinforcement; CR).

Specifically, test conditions consisted of contingent access to preferred items (tangible condition)

for Andrea, Robert, and Shelly, attention (attention condition) for Jennifer, and escape from task









demands (demand condition) for Mandy and Elaine. During the attention condition, the therapist

provided moderately preferred toys to the participant, and instructed the participant to play with

the toys while the therapist completed work. The therapist then sat down and ignored the

participant unless the participant engaged in problem behavior. Contingent on problem behavior,

the therapist provided approximately 10-15 s of attention in the form of response interruption,

reprimands (e.g., "don't do that") and statements of concern. During the demand condition, the

therapist instructed the participant to perform a variety of educational and vocational tasks using

a three-step graduated prompting procedure (i.e., verbal, model, physical). Contingent on

problem behavior, the therapist terminated task instructions, moved away from the participant,

and did not interact with the participant for 30 s. During the tangible condition, the participant

was given 2 min of continuous access to preferred tangible items (e.g., toys) prior to the session,

after which, the therapist removed the preferred items, and the session began. The therapist

returned the items to the participant for 30 s following each occurrence of problem behavior, and

ignored all other behavior.

In the control condition, the same reinforcer was available continuously and irrespective

of the occurrence of problem behavior noncontingentt reinforcement; NCR). Each condition

was conducted in a separate area of a classroom (school) or a different treatment room (clinic),

and was associated with the therapist wearing a distinctly colored t-shirt. The functional analysis

continued until a pattern of responding emerged in which high rates of problem behavior during

the CR condition relative to the NCR condition verified that the presumed reinforcer was

responsible for behavioral maintenance. The CR condition of the pairwise FA served as a

baseline for evaluating treatment effects during the FCT and signaled VI schedule conditions.









Functional communication training (FCT)

During the FCT condition, the contingency was reversed from baseline so that

appropriate communication, as opposed to problem behavior, produced access to the reinforcer

identified for each participant. At the outset of session, the therapist associated with the CR

condition implemented antecedent events in a manner identical to the CR condition. That is,

preferred items were removed from Andrea, Robert, and Shelly, Mandy and Elaine were

instructed to engage in tasks, and Jennifer was "ignored". During session, each communication

response resulted in 30-s access to the reinforcer identified during baseline, whereas, problem

behavior resulted in extinction (no reinforcement). To facilitate the emission of communication,

a second therapist verbally prompted the participant to appropriately communicate for the

reinforcer at the beginning of each session. If the participant appropriately requested the

reinforcer, no additional prompting occurred. If the participant did not request the reinforcer

independently following the initial verbal prompt, additional prompts were given (initially at an

interval 25% less than the interresponse time for problem behavior during baseline) and then

were eliminated using a prompt-fading procedure, in which the prompt interval was increased by

50%. Additional prompting was required for the first session with Amanda and Robert, and for

the first 6 sessions with Elaine. Fading was accomplished by gradually increasing the latency to

the verbal prompt over time. The initial delay to prompt the communication response was set at

25% below the average inter-response time between occurrences of problem behavior during the

last three CR sessions for Amanda and Robert, and 25% below the average latency to engage in

problem behavior following an instruction during the last 3 baseline CR sessions for Elaine. The

prompt delay was increased by 50% following two sessions at which problem behavior was at or

below a 90% reduction from baseline. The FCT condition was terminated when a participant









engaged in (a) Eive consecutive sessions with problem behavior at or below a 90% reduction

from baseline, and (b) independent communication.

Signal discrimination assessment

An assessment was conducted to identify a signal that was likely to be discriminated by

each participant. During the assessment, visual stimuli were presented in random locations near

the participant, with each presentation constituting a trial, and attending behavior (i.e., eyes

directed toward stimulus) within 10 s of stimulus presentation was measured. Each stimulus was

presented on 10 separate trials. A colored card or badge represented the first stimulus assessed

for all participants, with the assumption that the presentation of a colored card would be easy to

use in the natural environment. If the participant attended to the card on 80% of the trials, the

colored card was selected as the signal. If the participant did not attend to the card on at least

80% of trials, additional stimuli were assessed. Five of the 6 participants attended to the colored

card (Amanda, Robert, Maggie) or badge (Jennifer, Shelly) during the assessment; Elaine

required assessment with 5 additional stimuli before a tungsten light panel was identified.

Signaled VI schedule thinning

During this condition, communication continued to produce access to the functional

reinforcer, and problem behavior produced no programmed consequences; however,

communication was reinforced according to a VI schedule that was thinned gradually over time.

According to the VI schedule in effect, reinforcement was delivered for the first communication

response that occurred after an interval of time had elapsed since the last communication

response had been reinforced. Within the VI schedule, the availability of reinforcement for

communication was signaled by the presentation of the stimulus identified during the signal

assessment. The signal was presented to the participant at the precise second that the VI interval

elapsed and remained present until the participant engaged in communication and received the









reinforcer. Once the signal was removed, the next interval began. Throughout this condition,

each session began with the immediate presentation of the signal.

VI schedules were generated and randomized for each session using the Microsoft Excel

RAND function, and schedule values ranged from 50% below to 50% above the VI schedule

value used within that session. For example, if communication was reinforced according to a VI

70-s schedule, 50% of the VI values included interval lengths between 3 5-70 s, and 50% of the

VI values would include interval lengths between 70 and 105 s. The initial VI value was set at

25% below (a) the average interresponse time between communication responses during the last

three FCT sessions (Amanda, Jennifer, Robert, and Shelly) or (b) the average latency to engage

in communication following an instruction during the last three FCT sessions (Mandy and

Elaine).

The VI schedule was thinned by increasing the interval value by 50% following two

consecutive sessions in which (a) the rate of problem behavior did not increase and (b)

communication maintained. Thinning was terminated following five consecutive sessions in

which the rate of problem behavior was at or below a 90% reduction from baseline,

communication continued to occur, and a large proportion of communication responses occurred

in the presence of the signal under a terminal schedule of VI 240 s.

Results

Figures 5-1 and 5-2 depict the results of the clinical evaluation of the signaled VI schedule

for the participants whose problem behavior was maintained by social positive reinforcement.

During baseline, Amanda (Figure 5-1, top panel) engaged in aggression during the CR condition

but not during the NCR condition, indicating that her aggression was maintained by positive

(tangible) reinforcement. During FCT, when requests ("Can I have my toys back please")

produced access to toys, and aggression was placed on EXT, Amanda immediately began









making requests, while her aggression decreased to zero. During thinning, when the schedule of

reinforcement for requests was increased from VI-30 s to VI-240 s, her requests gradually

decreased and maintained at low rates, with no observed increase in problem behavior. The

percentage of requests that resulted in reinforcement during thinning was 80%.

Jennifer (Figure 5-1, bottom panel) engaged in problem behavior (SIB, aggression, and

disruption) during the CR condition in baseline but not during the NCR condition, indicating that

these problem behaviors were maintained by positive reinforcement (access to attention). The

implementation of FCT resulted in an immediate increase in requests for attention ("Can we

talk?"), and an immediate decrease in problem behavior. Thinning reinforcement from a VI-30 s

to a VI-240 s was associated with an immediate decrease in requests, which decreased further as

the condition progressed. With the exception of a few sessions, thinning was associated with low

rates of problem behavior relative to the baseline CR condition. In addition, 98% of Jennifer's

communication responses resulted in access to reinforcement.

Robert (Figure 5-2, top panel) engaged in aggression during the CR condition in baseline

but not during the NCR condition, indicating that his aggression was maintained by positive

(tangible) reinforcement. During FCT, when reinforcement was delivered following appropriate

requests ("Can I have my toys back, please?") but not following aggression, requests increased

immediately, and aggression decreased to zero. Thinning the reinforcement schedule to the

terminal value of VI-240 s was accomplished with gradual reductions in requests while

maintained zero rates of aggression. In addition, 90% of Robert' s requests resulted in

reinforcement during thinning.

Shelly (Figure 5-2, bottom panel) engaged in SIB and aggression during the CR condition

in baseline but not during the NCR condition of baseline, indicating that these behaviors were









maintained by positive (tangible) reinforcement. FCT was associated with an increase in requests

("Can I have my toys back please?") and a decrease in problem behavior to zero. Signaled VI

thinning was associated with an immediate decrease in requests and further decreases throughout

the condition. Problem behavior increased temporarily when the schedule was first increased to a

VI 30-s but subsequently decreased to zero for the remainder of the condition except for one

session. Further, 100% of her requests resulted in reinforcement during thinning.

Figure 5-3 shows data for the 2 participants whose problem behavior was maintained by

social negative reinforcement (escape from task demands). Mandy (top panel) and Elaine

(bottom panel) both engaged in problem behavior under the CR condition but not in the NCR

condition of baseline, verifying that problem behavior was maintained by escape. When FCT

was implemented for Mandy, appropriate requests ("Break please") increased to variable rates,

whereas aggression and property destruction showed noticeable variability initially, before

decreasing to zero. Schedule thinning produced an immediate and then more gradual decrease in

requests and low to zero rates of problem behavior. For Mandy, 97% of communication

responses resulted in reinforcement during thinning. During Elaine's FCT condition, voice-

output microswitch presses emerged somewhat slowly but then increased steadily. Her problem

behavior showed some variability throughout the first part of the FCT condition but eventually

decreased to zero. During schedule thinning, Elaine's microswitch press decreased gradually

throughout the condition, and problem behavior remained at or near zero except for three

sessions. In addition, 97% of Elaine' s microswitch press resulted in reinforcement during the

thinning phase.












BL FCT SIGNALED VI THINNING


AGG
E-0.5


0 130

AGG oO
(NCR) o

0-co ~ ^-------o see 0.0 C



eP Jennifer





Combined
10- ~Problem Behavior -

304






5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80

SESSIONS




Figure 5-1. Rate of problem behavior and communication during baseline (CR v. NCR), FCT,
and during signaled VI thinning for Amanda and Jennifer






























ill


I ---~---------------


5 10 15 20 25 30 35 40 45 50 55 60

SESSIONS




Figure 5-2. Rate of problem behavior and communication during baseline (CR v. NCR), FCT,
and during signaled VI thinning for Robert and Shelly


C~~C~C~-C~C~C~Cb~~~~~~~~~


BL


AGG
(CR)








AGG
(NCR)


SIGNALED VI THINNING


Robert


Shelly ri U


Communication

30

70105
240


AGG


Combined
Problem Behavior







20
45 70 20
10516
30


-~












BL FCT SIGNALED VI THINNING
6- -3
a Mandy
5-
Z Communication
4- AGG -2
--(CR)

S3-( I Id R 1Problem
Behavior



1- AGG
S(NCR) 1 24




Elaine


OO




0 20 304 1 1 0


105

5 10 15 20 25 30 35 40 45 50 55 60 65 70

SESSIONS



Figure 5-3. Rate of problem behavior and communication during baseline (CR v. NCR), FCT,
and during signaled VI thinning for Mandy and Elaine













Table 5-1. Operational definitions
Participant Target Operational Definitions
Amanda Aggression Hitting- Forceful contact between hand and the body of another individual. Scoring required hand to travel at least 6 and to produce e an audible
sound upon conta ct
Jennifer SIB Skin picking: fingernail to skin contact with movement associated with skin depression or lifting
Scratching: movement of a fingernail across the skin
Sel-iig closure of teeth on any part of participant's body
Bandage tamperin g: finger to bandage contact with movement of bandage
Aggression Hitting: forceful contact between hand and body of another individual. Scoring required hand to travel at least 6" and to produce an audible sound
upon contact
Kicking- forceful contact between foot and body of another individual. Same scoring criteria as with aggression
Bitng closure of teach on any part of another individual's body
Thoigodects at others: object must pass within 2' of individual
Disruption Ripping/tearing objects: ripping at least 1 or breaking objects such that tear or break in material is visible
Throwing obi ects: obj ect must travel at least 2'
Robert Aggression Hitting- Forceful contact between hand and the body of another individual. Scoring required hand to travel at least 6 and to produce e an audible
sound upon conta ct
Kicking- forceful contact between foot and body of another individual. Same scoring criteria as with aggression
Head butting: Forceful contact between head and the body of another individual
Choking- paeent around neck of another individual with prsuedirected inward
Shelly SIB Skin picking: fingernail to skin contact with movement associated with skin depression or lifting
Scratching: movement of a fingernail across the skin
Sel-iig closure of teeth on any part of participant's body
Bandage tamperin g: finger to bandage contact with movement of bandage
Head banging: forceful contact between the head and any surface from a distance of 6" or greater with audible sound.
Head hitting: forceful contact between the hand and any part of the head from a distance of 6" or greater with audible sound
Teeth banging: forceful contact between teeth and any surface
Nose picking: insertion of a finger into the nose
Aggression Hitting: forceful contact between hand and body of another individual. Scoring required hand to travel at least 6" and to produce an audible sound
upon contact
Kicking- forceful contact between foot and body of another individual. Same scoring criteria as with aggression
Biting- closure of teach on any part of another individual's body or clothes
Throwing obj ects at others: object must pass within 2' of individual
Disruption Throwing obj ects: obj ect must travel at least 2'
Breaking objects: tear or break in material must be visible
Hitting/Kicking objects: forceful contact between hand/foot and any surface. Body part must travel at least 6" and produ ce audible sound
Overturning furniture: furniture must be displaced 45 degrees from upright
Mandy Aggression Hitting: forceful contact between hand and body of another individual. Scoring required hand to travel at least 6" and to produce an audible sound
upon contact
Kicking- forceful contact between foot and body of another individual. Same scoring criteria as with aggression
Biting- closure of teach on any part of another individual's body or clothes
Property Breaking objects: tear or break in material must be visible
Destruction Hitting/Kicking objects: forceful contact between hand/foot and any surface. Body part must travel at least 6" and produ ce audible sound
Writing on walls
Elaine Aggression Hair pulling: grasping hair with any part of hand and pulling in direction away from head
Scratching: movement of fingernail across skin
Pinchin g- closure of at least two fingers around the skin with pressure directed toward th e skin









CHAPTER 6
DISCUSSION

The present studies provide both a basic demonstration and a clinical extension of the

effects of signaled VI reinforcement on responding. Data from both studies indicated that this

type of schedule may facilitate maintenance when relatively low rates of responding are

desirable following initial response acquisition.

In Study 1, response rates were compared under two VI-30 s schedules that were

identical except that the end of the VI interval (i.e., the availability of reinforcement) was

unsignaled in one schedule but signaled by the presentation of a visual stimulus in the other.

Under the unsignaled VI-30 s schedule, only 1 of the 4 participants engaged in moderate

response rates (George), and a small percentage of responses resulted in reinforcement across

participants. By contrast, responding became moderate and efficient for all participants under the

Signaled VI 30-s schedule, occurring at approximately 2.0 rpm at the end of the condition.

Further, a large proportion of responses resulted in reinforcement across participants.

As the unsignaled and signaled conditions differed only with respect to the presence and

absence of the signal, differences in responding observed between the two conditions can be

attributed to the effects of the signal. In the absence of the signal, it is likely that participants

could not discriminate the availability of reinforcement (the end of the interval), particularly

given the variable nature of the schedule. Hence, other factors, such as those discussed below,

may have contributed to the variability in responding observed.

The presence of the signal may have increased the likelihood that participants could

discriminate the availability of reinforcement, such that responding occurred primarily in the

presence of the signal, when it was likely to contact reinforcement. Given the high probability of









reinforcement for responding, the response-reinforcer contingency was likely to remain strong,

resulting in low but persistent rates of behavior.

The results of the signaled VI condition are consistent with those reported by Azrin (1958),

who examined the effects of an auditory stimulus, noise, on the FI observing responses of 80

soldiers. A meter needle was deflected every 3 min, and soldiers were trained to release a switch

to bring the needle back to its normal position. The room was kept dark, and presses on a button

illuminated the needle for 0. 1 s; the sight of the needle deflecting was considered the reinforcer.

During baseline, responses occurred rarely during the first 2 min of the FI schedule; however,

high response rates occurred throughout the 3rd min of the FI until the needle was detected and

restored. When a 110-decibel white noise was continuously present throughout the first 165 s of

the interval but was discontinued for the 15 s preceding a needle deflection, observing responses

occurred only during the 15-s period of quiet and never during the initial 165 s of noise.

The temporal relation of noise to the target was then reversed so that 165 s of quiet

preceded a 15-s period of noise. Soldiers responded by making observing responses during the

165 s of quiet and no observing responses during the 15 s of noise. Recovery occurred around

interval 14, and soldiers began to respond only during the 15 s of noise. At that point, the

temporal relation of noise to the target was again reversed to the initial conditions, and soldiers

again responded exclusively during periods of quiet. Azrin (1958) concluded that the noise (or

quiet) had acquired discriminative control over observing responses to the exclusion of control

by the temporal properties of the FI schedule.

Previous research with nonhuman subj ects has shown that unsignaled VI schedules tend to

produce low to moderate, stable response rates (Ferster & Skinner, 1957). In the current study,

however, only 1 participant engaged in moderate response rates throughout the unsignaled









condition (George), and the responding of the other participants was characterized by a high

degree of variability. High response rates, which stabilized toward the end of the condition, were

observed for 1 participant (Katy) whereas moderate, but increasing response rates were observed

for Lesley. Initially high response rates, that decreased to low to zero rates after four sessions,

were observed for Charlotte.

Responding that is inconsistent with a current reinforcement schedule, such as that

observed with Katy, Charlotte, and Lesley, may reflect the influence of historical variables such

as prior experience with a different reinforcement schedule (Weiner, 1969). Freeman and Lattal

(1992), for example, trained three pigeons on an FR-1 schedule before exposing them to a VI

schedule, while discriminative stimuli were held constant across the schedules. Following

training, all pigeons engaged in high response rates during the initial 8 sessions of the VI

condition, and one pigeon engaged in high response rates across all sessions of the VI condition.

Historical control, such as that identified in the study by Freeman and Lattal (1992) may

have accounted for the persistent high rate responding observed under the VI schedule for Katy

and Lesley. In Charlotte's case, however, high response rates may have interacted with the

temporal delivery of reinforcement under the VI schedule to produce the steep decline in

responding after four sessions. More specifically, if high response rates were emitted under the

VI schedule, and reinforcement was not delivered for an average of 30 s, several responses

would have been unreinforced, resulting in EXT of the response. In support of this, Charlotte's

data under the VI schedule are characteristic of responding during EXT. That is, a slight increase

in responding resembling an extinction-induced burst was observed immediately after Charlotte

switched schedules, followed by systematic decreases in response rate to zero.










The generally decreasing trend observed across the unsignaled and signaled VI conditions

for George may suggest the occurrence of an order effect. That is, responding under the

unsignaled VI schedule may have influenced responding under the signaled VI schedule.

However, the greater percentage of responses that produced reinforcement under the signaled VI

schedule (67.4%) relative to the unsignaled VI schedule (40.1%) suggested that the signal, as

opposed to a previous history with the VI schedule, accounted for the moderate and efficient

responding observed under the signaled VI schedule for George.

To summarize, the signaled VI schedule was effective in maintaining moderate and

efficient response rates, which provided a basis for evaluating the schedule in an applied context

in Study 2.

Study 2 involved a clinical evaluation of the signaled VI schedule as a maintenance

procedure following acquisition of an alternative response to problem behavior. A pairwise

functional analysis similar to that used by Hanley et al. (2001) identified a maintaining

contingency for each participant's problem behavior. Although the pairwise assessment included

only one test condition and did not rule out other environmental influences on problem behavior,

it was sufficient to develop an intervention strategy based on the identified function. The

differential reinforcement procedures implemented as FCT were effective in decreasing problem

behavior and increasing communication for all participants. These results were consistent with

those from a large body of research supporting the use of FCT as treatment for problem behavior

maintained by social consequences (e.g., Hagopian et al., 1998; Wacker et al. 1990).

The vast maj ority of studies on FCT involve continuous schedules of reinforcement for

alternative behavior (e.g., Carr & Durand, 1985). Although continuous reinforcement may be

advantageous during response acquisition, a comparable reinforcement schedule is not likely to









exist in the natural environment, where several factors preclude the immediate delivery of

reinforcement (e.g., parent attending to another child). To preserve the integrity of FCT

outcomes under naturalistic conditions, strategies for maintaining communication under

conditions of intermittent or delayed reinforcement are necessary.

A survey of the basic and applied literature revealed that response maintenance is related to

preservation of the response-reinforcer contingency (Hanley et al., 2001; Sizemore & Lattal,

1978), and that contingency preservation may be facilitated through the use of signals that are

correlated with the availability of reinforcement (Hanley et al., 2001). In the current study,

signaled VI thinning, in which alternative behavior was reinforced less frequently over time

according to an increasing VI schedule that included a signal when the interval terminated, was

used as the maintenance strategy. The schedule was effective in maintaining treatment effects

under increasingly intermittent reinforcement. That is, as the reinforcement schedule was thinned

to the terminal value of 240 s, problem behavior remained low, and alternative behavior

decreased to rates that reflected the availability of reinforcement.

Data indicating that a high percentage of alternative responses resulted in reinforcement

during thinning provides evidence that the signal had acquired stimulus control over appropriate

behavior. At any given interval value during thinning, there are limited opportunities for

reinforcement. Under a VI-70 s schedule, for example, there are approximately 8 potential

reinforcer deliveries within a 10-min session. For a high percentage of responses to contact

reinforcement, a participant would have to respond when, and only when reinforcement was

available. Across participants, 93.7% of communication responses resulted in reinforcement,

indicating that participants discriminated the availability of reinforcement and responded

accordingly.









Results of the current study are consistent with those of Hanley et al. (2001), showing

that signaled reinforcer availability produces moderate rates of communication that occur at

appropriate points in time (i.e., when reinforcement was available) and low rates of problem

behavior. The differences between the multiple-schedule used by Hanley et al. (2001) and the VI

schedule in the current study suggest different types of application. When only one response is

sufficient for reinforcement, but reinforcement will be delayed or available only once in awhile,

the signaled VI schedule may be preferred. Most relevant to the use of the signaled VI schedule

are instructional or other demand-related situations in which only one request for a break or task

termination is required and appropriate; however, completion of the task is mandatory (e.g.,

brushing teeth). By contrast, when higher rates of behavior are either acceptable or desirable, but

reinforcement will be unavailable for periods of time, the multiple-schedule may be

advantageous. For example, repeated requests for attention are considered appropriate during

periods of conversation or play, and may be accommodated nicely within a multiple schedule.

During reinforcement thinning, several sessions were conducted before alternative

behavior became moderate and efficient, which suggests that discriminative control over

responding may require time to develop. Future investigations may examine procedures for

facilitating the establishment of stimulus control prior to exposing newly acquired behavior to

conditions of intermittent or delayed reinforcement. For example, presenting a stimulus during

the initial reinforcement phase (FR-1) would establish a longer history of reinforcement in the

presence of the signal and may be an efficient way to generate stimulus control over responding

prior to using the signal clinically.

The structure of the signaled VI procedure is potentially susceptible to producing

inadvertent reinforcement of problem behavior. More specifically, the onset of the signal is time-










based, and if problem behavior occurs temporally close to the onset of the signal, it may become

associated with the signal, and the subsequent production of reinforcement. The terminal result

of this process is that the signal may become discriminative for the occurrence of problem

behavior (as opposed to alternative behavior), engendering high rates of problem behavior in its

presence. Although not a problem in the current investigation, therapeutic use of the signaled VI

procedure may incorporate a changeover delay contingent on the occurrence of problem

behavior. A changeover delay involves a brief delay to reinforcement (or onset of the signal)

following problem behavior, such that inadvertent reinforcement of problem behavior can be

reduced.

Future research may also examine additional therapeutic uses of signals. In some

situations, it is not possible or acceptable to use extinction during the treatment of problem

behavior, and discriminative control over appropriate responding may decrease the likelihood

that problem behavior will occur and inadvertently contact reinforcement. For example, severe

SBE may be maintained by attention; however, the delivery of attention following self-inflicted

injury may be unavoidable. If, however, appropriate requests for attention are brought under the

control of signals, appropriate requests may occur to the exclusion of problem behavior.










LIST OF REFERENCES


Azrin, N. H. (1958). Some effects of noise on human behavior. Journal of the Experimental
Analysis ofBehavior, 1, 183-200.

Azzi, R., Fix, D. S. R., Keller, F. S., & Rocha E Silva, M. I. (1964). Exteroceptive control of
response under delayed reinforcement. Journal of the Experimental Analysis ofBehavior,
7, 159-162.

Baum, W. M. (1973). The correlation-based law of effect. Journal of the Experimental Analysis
ofBehavior, 20, 137-153.

Carr, E. G., & Durand, V. M. (1985). Reducing behavior problems through functional
communication training. Journal ofApplied Behavior Analysis, 18, 1 11-126.

Durand, V. M., & Carr, E. G. (1987). Social influences on "self-stimulatory" behavior: Analysis
and treatment application. Journal ofApplied Behavior Analysis, 20, 1 19-132.

Durand, V. M., & Carr, E. G. (1991). Functional communication training to reduce challenging
behavior: Maintenance and application in new settings. Journal ofAppliedBehavior
Analysis, 24, 251-264.

Ferster, C. B., & Skinner, B. F. (1957). Schedules ofreinforcement. New York: Appleton-
Century-Crofts .

Fisher, W., Piazza, C. C., Bowman, L. G., Hagopian, L. P., Owens, J. C., Slevin I. (1992) A
comparison of two approaches for identifying reinforcers for persons with severe and
profound disabilities. Journal ofApplied Behavior Analysis, 5, 491-498

Fisher, W. W., Thompson, R. H., Hagopian, L. P., Bowman, L. G., Krug, A. (2000). Facilitating
tolerance of delayed reinforcement during Functional Communication Training. Journal
of the Experimental Analysis ofBehavior, 24, 3-29.

Freeman, T. J., & Lattal, K. A. (1992). Stimulus control of behavioral history. Journal of the
ExperimentalAnalysis ofBehavior, 57, 5-15.

Fyffe, C. C., Kahng, S., Fittro, E., & Russell, D. (2004). Functional analysis and treatment of
inappropriate sexual behavior. Journal ofApplied Behavior Analysis, 3 7, 401-404.

Hagopian, L. P., Fisher, W. W., Sullivan, M. T., Acquisto, J., & LeBlanc, L. A. (1998).
Effectiveness of functional communication training with and without extinction and
punishment: A summary of 21 inpatient cases. Journal ofApplied Behavior Analysis, 31,
211-235.

Hammond, L. J. (1980). The effect of contingency upon the appetitive conditioning of free-
operant behavior. Journal of the Experimental Analysis ofBehavior, 34, 297-304.










Hanley, G. P., Iwata, B. A., & Thompson, R. H. (2001). Reinforcement schedule thinning
following treatment with functional communication training. Journal ofAppliedBehavior
Analysis, 34, 17-37.

Horner, R. D., & Day, H. M. (1980). The effects of response efficiency on functionally
equivalent competing behaviors. Journal ofApplied Behavior Analysis, 24, 719-732.

Iwata, B. A., Dorsey, M. F., Slifer, K. J., Bauman, K. E., & Richman, G. S. (1982). Toward a
functional analysis of self-injury. Analysis and'Intervention in Developmental
Disabilities, 192, 193-120.

Iwata, B. A., Pace, G. M., Kalsher, M. J., Cowdery, G. E., & Cataldo, M. F. (1990).
Experimental analysis and extinction of self-injurious escape behavior. Journal of
Applied Behavior Analysis, 23, 1 1-27.

Kahng, S., Iwata, B. A., DeLeon, I. G., & Worsdell, A. S. (1997). Evaluation of the "control over
reinforcement" component in functional communication training. Journal of Applied
Behavior Analysis, 30, 267-277.

Kelley, M. E. (2003). The effects of signals on responding during delayed reinforcement.
(Doctoral Dissertation, Louisiana State University, 2003). Dissertation Abstracts
International, 64, 1478.

Kern, L., Mauk, J. E., Marder, T. J., & Mace, F. C. (1995). Functional analysis and intervention
for breath holding. Journal ofApplied Behavior Analysis, 28, 339-340.

Lalli, J. S., Vollmer, T. R., Progar, P. R., Wright, C., Borrero, J., Daniel, D., et al. (1999).
Competition between positive and negative reinforcement in the treatment of escape
behavior. Journal ofApplied Behavior Analysis, 32, 285-296.

Lattal, K. A. (1984). Signal functions in delayed reinforcement. Journal of the Experimental
Analysis ofBehavior, 42, 239-253.

Lattal, K. A. (1995). Contingency and behavior analysis. The Behavior Analyst, 18, 204-224.

Meyer, K. A. (1999). Functional analysis and treatment of problem behavior exhibited by
elementary school children. Journal ofApplied Behavior Analysis, 32, 229-232.

Michael, J. (1982). Distinguishing between discriminative and motivational functions of stimuli.
Journal of the Experimental Analysis ofBehavior, 37, 149-155.
Pierce, C.H., Hanford, P.V., & Zimmerman, J. (1972). Effects of different delay of reinforcement
procedures on variable-interval responding. Journal of the Experimental Analysis of
Behavior, 18, 141-146.

Richards, R.W. (1981). A comparison of signaled and unsignaled delay of reinforcement.
Journal of the Experimental Analysis ofBehavior, 35, 145- 152.











Schaal, D. W. & Branch, M. N. (1988). Responding of pigeons under variable-interval
schedules of unsignaled, briefly signaled, and completely signaled delays to
reinforcement. Journal of the Experimental Analysis ofBehavior, 50, 33-54.

Shirley, M. J., Iwata, B. A., Kahng, S., Mazaleski, J. L., & Lerman, D. C. (1997). Does
functional communication training compete with ongoing contingencies of
reinforcement? An analysis during response acquisition and maintenance. Journal of
Applied Behavior Analysis, 30, 93-104.

Sizemore, O.J., & Lattal, K.A. (1978). Unsignaled delay of reinforcement in variable-interval
schedules. Journal of the Experimental Analysis ofBehavior, 30, 169-175.

Smith, R. G., Iwata, B. A., Goh, H., & Shore, B. A. (1995). Analysis of establishing operations
for self-injury maintained by escape. Journal ofAppliedBehavior Analysis, 28, 515-535.

Tarbox, R. S. F., Wallace, M. D., & Williams, L. (2003). Assessment and treatment of
elopement: A replication and extension. Journal ofAppliedBehavior Analysis, 36, 239-
244.

Thompson, R. H., Fisher, W. W., Piazza, C. C., & Kuhn, D. E. (1998). The evaluation and
treatment of aggression maintained by attention and automatic reinforcement. Journal of
Applied Behavior Analysis, 31, 103-1 16.

Vollmer, T. R., Borrero, J. C., Lalli, J. S., & Daniel, D. (1999). Evaluation self-control and
impulsivity in children with severe behavior disorders. Journal ofAppliedBehavior
Analysis, 32, 451-466.

Vollmer, T. R., Iwata, B. A., Zarcone, J. R., Smith, R. G., & Mazaleski, J. L. (1993). The role of
attention in the treatment of attention-maintained self-injurious behavior: Noncontingent
reinforcement and differential reinforcement of other behavior. Journal ofApplied
Behavior Analysis, 26, 9-21.

Vollmer, T. R., Marcus, B. A., & LeBlanc, L. (1994). Treatment of self-injury and hand
mouthing following inconclusive functional analyses. Journal ofAppliedBehavior
Analysis, 27, 331-344.

Vollmer, T. R., Roane, H. S., & Ringdahl, J. E., & Marcus, B. A. (1994). Evaluation treatment
challenges with differential reinforcement of alternative behavior. Journal of Applied
Behavior Analysis, 32, 9-23.

Wacker, D. P., Steege, M. W., Northrup, J., Sasso, G., Berg, W., Reimers, T., et al. (1990). A
component analysis of functional communication training across three topographies of
severe behavior problems. Journal ofApplied Behavior Analysis, 23, 417-429.










Williams, B.A. (1976). The effects of unsignaled delayed reinforcement. Journal of the
ExperimentalAnalysis ofBehavior, 26, 441-449.

Worsdell, A. S., Iwata, B. A., Hanley, G. P., Thompson, R. H., & Kahng, S. (2000). Effects of
continuous and intermittent reinforcement for problem behavior during functional
communication training. Journal of Applied Behavior Analysis, 33, 167-179.









BIOGRAPHICAL SKETCH

Carrie Dempsey was born in Santa Rosa, California (CA), in 1975. She graduated with a

Bachelor of Arts degree in psychology and Latin American & Iberian studies from the University

of California-Santa Barbara in 1995. Following graduation, Carrie pursued higher education at

the University of the Pacific in Stockton, CA, where she was first exposed to behavioral

psychology and became inspired to develop as a behavior analyst. Carrie entered graduate school

at the University of Florida in 2002 under the supervision of Brian A. Iwata. Upon graduation,

she intends to return to CA to establish a research-oriented treatment center for severe problem

behavior. In her spare time, Carrie enj oys being in the outdoors, kayaking, traveling, and

learning about various aspects of computer technology.





PAGE 1

1 EFFECTS OF SIGNALS ON THE MAINTEN ANCE OF ALTERNATIVE BEHAVIOR UNDER INTERMITTENT REINFORCEMENT By CARRIE MELISSA DEMPSEY A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2007

PAGE 2

2 2007 Carrie Melissa Dempsey

PAGE 3

3 To my grandparents, Ernie and Betty Moretti

PAGE 4

4 ACKNOWLEDGMENTS I would like to thank my advisor, Dr. Brian A. Iwata, whose enduri ng dedication, support, and guidance have been instrumental to my pr ofessional development. Also, I would like to thank the members of my committee: Drs. Sco tt Miller, Stephen Smit h, and Timothy Vollmer, for their input and direction during this process. I would like to th ank my colleagues, particularly Pam Neidert and Jessica Thomason, for shaping my clinical skills and for teaching me the nuances of behavioral assessment and treatment. Finally, I would like to thank my parents and grandparents, whose consistent, common sense appr oach to raising children cultured my basic views on behavior, and predisposed me to a career in applied behavior analysis.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS...............................................................................................................4 LIST OF TABLES................................................................................................................. ..........7 LIST OF FIGURES................................................................................................................ .........8 ABSTRACT....................................................................................................................... ..............9 CHAPTER 1 FUNCTIONAL COMMUNICATION TRAINING...............................................................11 Function-Based Treatment......................................................................................................11 Applications of Functional Communication Training............................................................12 Component Analyses of FCT.................................................................................................13 2 MAINTENANCE FOLLOWING FCT..................................................................................15 The Natural E nvironment.......................................................................................................15 Delay Fading................................................................................................................... ........15 Fixed-Interval Schedule Thinning..........................................................................................17 3 SIGNALS........................................................................................................................ .......18 4 STUDY 1: COMPARISON OF UNSIGNALED VERSUS SIGNALED VI SCHEDULES...................................................................................................................... ...24 Method......................................................................................................................... ...........24 Participants and Setting...................................................................................................24 Response Measurement and Reliability..........................................................................24 Stimulus Preference Assessment.....................................................................................25 Procedure...................................................................................................................... ...25 Results........................................................................................................................ .............26 5 STUDY 2: CLINICAL EVALUATION OF SIGNALED VI SC HEDULE THINNING.....30 Method......................................................................................................................... ...........30 Participants and Setting...................................................................................................30 Response Measurement and Reliability..........................................................................31 Safety Precautions...........................................................................................................33 Procedure...................................................................................................................... ...33 Functional analysis (baseline)..................................................................................33 Functional communication training (FCT)..............................................................35 Signal discrimination assessment.............................................................................36

PAGE 6

6 Signaled VI schedule thinning.........................................................................................36 Results........................................................................................................................ .............37 6 DISCUSSION..................................................................................................................... ....44 LIST OF REFERENCES............................................................................................................. ..51 BIOGRAPHICAL SKETCH.........................................................................................................55

PAGE 7

7 LIST OF TABLES Table page 5-1 Operational definitions.................................................................................................... ...43

PAGE 8

8 LIST OF FIGURES Figure page 4-1 Responses per minute on the arbitrary ta sk under the unsignaled and signaled VI-30" schedules...................................................................................................................... ......29 5-1 Rate of problem behavior and comm unication during baseline (CR v. NCR), FCT, and during signaled VI thinni ng for Amanda and Jennifer................................................40 5-2 Rate of problem behavior and comm unication during baseline (CR v. NCR), FCT, and during signaled VI thinni ng for Robert and Shelly.....................................................41 5-3 Rate of problem behavior and comm unication during baseline (CR v. NCR), FCT, and during signaled VI thi nning for Mandy and Elaine....................................................42

PAGE 9

9 Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy EFFECTS OF SIGNALS ON THE MAINTEN ANCE OF ALTERNATIVE BEHAVIOR UNDER INTERMITTENT REINFORCEMENT By Carrie Melissa Dempsey August 2007 Chair: Brian A. Iwata Major: Psychology A common approach to the treatme nt of problem behavior is to strengthen an alternative, appropriate behavior. Although continuous reinforcement is required during acquisition, the practicalities of maintenance require intermittent reinforcement. Previous methods for thinning reinforcement for alternative behavior (i.e., de lay fading), however, have resulted in response deterioration. Results of several basic studies have shown that extero ceptive stimuli (signals) may be used to control the rate and temporal distribution of responding when reinforcement is intermittent or delayed; however, only one app lied study has examined the use of signals for maintaining alternative behavior du ring reinforcement-schedule thinning. The purpose of this study was to evaluate patterns of responding when reinforcer availability was signaled by visual stimuli and delivered according to a variable-interval (VI) schedule. Study 1 involved a basic demonstratio n of the effects of signaled and unsignaled reinforcer availability under a VI schedule. Stud y 2 involved a clinical ev aluation of signaled, VI reinforcement as a maintenance procedure for altern ative behavior in the context of treatment for severe problem behavior. Results of Study 1 revealed that the signaled-VI schedule was associated with moderate rates of respondi ng, whereas the unsignaled schedule tended to produce either high or zero rate s of responding. Results of Study 2 replicated and extended the

PAGE 10

10 results of Study 1,showing that th e signaled-VI schedule produced low rates of problem behavior and moderate rates of alternative behavior as the reinforcement schedule was thinned.

PAGE 11

11 CHAPTER 1 FUNCTIONAL COMMUNICATION TRAINING Function-Based Treatment The development of methods for identifying be havioral function, such as the functional analysis (Iwata, Dorsey, Slifer, Bauman, & Richman, 1982), spurred a new generation of treatment for problem behavior, commonly referred to as function-based treatment. In functionbased treatment, the variables that maintain proble m behavior are identified and then are used to inform the course of intervention. Function-based treatment may be applied in several ways. One application involves extinction (EXT ), in which the functional rein forcer is eliminated following problem behavior. For example, Iwata, Kalsher, Cowdery, and Cataldo (1990) showed that the self-injurious behavior (SIB) of 7 participants with developmental disabilities was maintained by escape from task demands. Extinction was implem ented by terminating escape following SIB, and resulted in decreases in SIB to near zero levels. Another application of func tion-based treatment involves noncontingent reinforcement (NCR), in which the functional reinforcer is delivered on a time-bas ed schedule that is independent of responding. For example, Vollmer Iwata, Zarcone, Smith, and Mazaleski (1993) identified that the SIB of 3 participants with mental retardation was maintained by access to attention. During NCR, attention was delivered initially on a dens e, fixed-time schedule, which was subsequently thinned. All participants s howed immediate reductions in behavior, which were sustained over time. A third application of function-based treatmen t involves differential reinforcement of an alternative behavior (DRA), in which the func tional reinforcer is withheld following problem behavior, and used to establis h an alternative, competing re sponse. DRA has been used to increase compliance in participants with escap e-maintained problem behavior (e.g., Lalli et al.,

PAGE 12

12 1999) and leisure item engagement in participan ts who engage in automatically reinforced behavior (Vollmer, Marcus, & LeBlanc, 1994). In addition, and of particular relevance to the current study, DRA has been used to establish appr opriate communication (e.g., Wacker et al., 1990). In the context of communication, a co mmon term given to DRA is Functional Communication Training (FCT; Carr & Durand, 1985), although the basic structure of the intervention remains unchanged. The following sectio ns review topics rele vant to the use and effectiveness of FCT. Applications of Functional Communication Training Carr and Durand (1985) conducted the first appl ication of FCT with 4 participants who engaged in disruption in a classroom setting. Ante cedent variables were manipulated to generate information regarding the events that reinfo rced problem behavior. Once reinforcers were identified, participants were taught to emit two statements in response to an experimenter prompt: a relevant statement that produced access to the reinforcer, and an irrelevant statement that produced access to an arbitrary consequen ce. Throughout the evaluation, problem behavior resulted in EXT. When relevant statements were reinforced, decreases in problem behavior and increases in communication were observed for a ll participants. However, when irrelevant statements produced access to an arbitrary cons equence, treatment effects were not observed. Since its initial app lication, the effectiveness of FC T has been demonstrated across several behaviors including aggression (Thom pson, Fisher, Piazza, & Kuhn, 1998), SIB (Wacker et al. 1990), elopement (Tarbox, Wallace, & W illiams, 2003), inappropriate sexual behavior (Fyffe, Kahng, Fittro, & Russell, 2004), ster eotypy (e.g., Durand & Carr, 1987), breath holding (Kern, Mauk, Marder, & Mace, 1995), and off-task behavior (Meyer, 1999). Given the positive treatment outcomes associated with FCT, studies have examined the components of FCT that contribute to its effectiveness.

PAGE 13

13 Component Analyses of FCT Carr and Durand (1985) and Wacker et al. (1990) suggested th at FCT differs from other differential reinforcement procedures (e.g., DRO), in that it allows the part icipant to control the delivery of reinforcement. That is, the particip ant may engage in communication to produce the reinforcer at any time. Kahng, Iwata, Deleon, an d Worsdell (1997) investig ated the contributing role of control over reinforcement by compari ng responding under two procedures, one in which participants' behavior influenced reinforcer delivery (FCT), and one in which participants' behavior did not influence re inforcer delivery (NCR). The amount of reinforcement was yoked across the two procedures. Both procedures pr oduced comparable decreases in behavior, indicating that control ov er reinforcement by a participant did not influence treatment effects. In addition, investigators have examined the role of EXT in FCT. For example, Hagopian, Fisher, Sullivan, Acqui sto, and LeBlanc (1988) conducte d a large-scale analysis of FCT and found that, in the absence of EXT, pr oblem behavior was not reduced by 90% for any of 11 participants. Worsdell, Iwata, Hanley, Thompson, and Kahng (2000) evaluated the role of EXT during the acquisition of alternative behavior for 5 participants whose problem behavior was maintained by social positive reinforcement. When participants were exposed to an initial phase in which both problem behavior and al ternative behavior were reinforced, only 1 participant acquired the altern ative response. The schedule of reinforcement for problem behavior was made more intermittent for the remaining 4 participants to determine the point at which more responding would be al located to alternative behavior than to problem behavior. Switching to alternative behavior did not occu r until reinforcement for its occurrence was 2, 3, and even 20 times more likely than for problem behavior. Shirley, Iwata, Kahng, Mazeleski, and Lerman (1997) examined the influence of EXT during both the acquisition and maintenance of manual signs during FCT for 3 participants.

PAGE 14

14 When SIB continued to produce reinforcemen t, SIB remained at baseline rates, and communication was not acquired. Th e addition of EXT, however, wa s associated with decreases in SIB and increases in signing for all participants. Collectively, these results suggest that EX T is a critical component of FCT; however, some research suggests that FCT may also be conducted with high integr ity in the absence of EXT if the establishing operation for behavior (e.g., Smith, Iwata, G oh, & Shore, 1995), or parameters of reinforcement for appropriate and inappropriate responses are manipulated (Horner et al., 1980; Vollmer et al., 1999). Aside from examining the relative effects of FCT components, a small number of studies have in vestigated methods for maintaining treatment effects following FCT. These methods ar e reviewed briefly in Chapter 2.

PAGE 15

15 CHAPTER 2 MAINTENANCE FOLLOWING FCT The Natural Environment During FCT sessions, communicatio n is typically reinforced on a fixed-ratio 1 (FR-1) schedule to facilitate response acquisition, which tends to enge nder high and stable rates of communication (Carr & Durand, 1985). However, reinforcement for communication may be delayed or intermittent in the natural environment. Some examples: A busy parent may have to direct attention to other individua ls or activities, some tangible items (e.g., edibles) are unhealthy if provided in large quantities, and some demands (e.g., hygiene tasks) require compliance. If communication occurs freque ntly but fails to contact re inforcement, it may extinguish. Subsequently, other behavior that has produced reinforcement in th e past (i.e., problem behavior) may re-emerge. In support of this, Dura nd and Carr (1991) observed a breakdown in communication and re-emergence in problem behavi or when treatment was implemented in the natural environment for 1 of 3 participants. Direct observations, which were conducted during follow up, revealed that the teacher was not attending to the participant's reque sts for assistance. Relatively few strategies have been designed to maintain communication under conditions of intermittent or delayed reinforcement. Two such strategies ar e delay fading (Fisher, Thompson, Hagopian, Bowman, & Krug, 2000) and fixed-interval (FI) schedule thinning (Hanley, Iwata, & Thompson, 2001). Delay Fading Delay fading involves gradually lengthen ing the delay between communication and reinforcement. Typically, interv al increases are determined ba sed on levels of responding during preceding sessions. Three applied studies have examined the effects of delay fading on responding following FCT and have obtained comparable results. Fi sher et al. (2000) evaluated

PAGE 16

16 progressively increasing delays to reinforcem ent for communication and found that additional treatments (e.g., providing tasks during delay) were required to maintain treatment effects once the delay interval was increased to 30 s. As part of a larger study, Hanley et al. (2001) examined delay fading by systematically increasing the de lay to reinforcement for communication from 1 s to 25 s, and found that communication decrease d to low rates under the 25-s delay, at which point problem behavior increased. Hagopian et al. (1998) conducted a large-scale study to assess the outcome of FCT alone or in combination with other procedures (e.g., EXT) with 21 clients who engaged in problem behavior such as SIB, aggre ssion, and property destruction. Func tional analyses revealed that problem behavior was maintained by access to attention for 9 participants, by escape from demands for 7 participants, by access to tangible items for 1 participant, and by multiple sources of control for 4 participants. Demand fading or delay-to-reinforcement fading was conducted in 12 of 25 applications of FCT w ith EXT, but reductions in problem behavior were maintained after fading in only 5 applications Further, attempts to regain control over the behavior in the remaining 7 cases (e.g., reducing delay interval or number of demands ) were unsuccessful. The results of these studies suggest that a newly acquired response may deteriorate under delayed reinforcement, a conclusion that is also supported by basic resear ch (Azzi, Fix, Keller, & Rocha E Silva, 1964; Schaal & Branch, 1988; Sizemore & Lattal, 1978). Decrements in responding under delay may reflect a decrease in contingency strength, which is determined by the temporal contiguity between response and reinforcement (Baum, 1973) and the probability of reinforcement for a response relative to the response-independent probability of reinforcement (Hammond, 1980; Lattal, 1995). Under conditions of delay, the temporal contiguity between a response and reinforcer is eroded. For exampl e, Sizemore and Lattal (1978) showed that

PAGE 17

17 response rates decrease systematical ly with obtained (actual) delays to reinforcement, in addition to nominal (programmed) delays. Fixed-Interval Schedule Thinning Under an FI reinforcement schedule, a reinforc er is delivered for the first response that occurs after an interval of time has elapsed since the last reinforced response (Ferster & Skinner, 1957). FI schedule thinning occurs by graduall y lengthening the time interval across sessions, based on levels of responding during precedi ng sessions (Hanley et al., 2001). In this arrangement, the contingency between a respons e and reinforcer may be preserved given the high probability of immediate re inforcement for communication. Hanley et al. (2001) implemented FI thinning by gradually increasing the interval length from 1 s to 58 s. Results showed that probl em behavior remained low throughout thinning. However, communication gradually increased to rates that were higher than those observed during initial training, and only a small proportion of co mmunication responses resulted in reinforcement. Elevated rates of communicati on during FI schedule thinning suggest that the contingency was preserved. On the other hand, a large propor tion of communication responses occurred when reinforcement was not available. Persistence of high rates of unreinfor ced responding over time may degrade the contingency between the respon se and reinforcement, causing a break down in responding (Hammond, 1980). The results of studies on delay fading and FI schedule thinning illustrate that communication may not occur at appropriate levels under intermittent reinforcement if reinforcement is delayed or difficult to discriminate As such, the identifica tion of strategies that enhance reinforcer immediacy and discriminatio n is warranted. Research relevant to the development of one strategy is addressed in the next chapter

PAGE 18

18 CHAPTER 3 SIGNALS In basic and applied research, exteroceptive stim uli, or signals, have been used to control the occurrence of responding under conditions in which reinforcement is delayed (Kelley, 2003; Lattal, 1984; Morgan, 1972; Pi erce, Hanford, & Zimmerman, 1972; Richards, 1981; Schaal & Branch, 1988, 1990; Williams, 1976) or intermittent (Fisher, Kuhn, & Thompson, 1998; Hanley et al., 2001). Several basic studies have shown that response rates may be maintained under delayed reinforcement when the response that begins the delay also produces a change in visual stimulation, such as a blackout or flashing light. Fo r example, Azzi et al. (1964) trained three rats on a schedule of continuous reinforcement befo re exposing them to unsignaled reinforcement delays of 1, 3, 5, 7.5, 10, 15, and 20 s. Results showed that response rates decreased systematically for all 3 subjects. Subjects were then exposed to a reinforcement delay of 20 s and then 30 s for several days. Under each reinforcem ent delay, half of the sessions involved an unsignaled reinforcement delay, and half of the sessions involved signaled reinforcement delay, in which responses produced a blackout pe riod (chamber light off) that endured until reinforcement was delivered. The signaled delay wa s associated with high response rates at both interval values, whereas the unsignaled delay was associated with low response rates. Richards (1981) examined the influence of si gnals using two reinforcement schedules that were expected to produce similar response ra tes but disparate reinforcer rates under reinforcement delay: a VI 60-s schedule and a di fferential-reinforcement-of-low-rate (DRL) 20-s schedule. Four pigeons were trained on the VI-6 0 s schedule, and 5 pigeons were trained on the DRL schedule, following which all subjects we re exposed to 48, 1-hour sessions with the following reinforcement delays: 10 s, 5 s, 2.5 s, 1 s, .5 s. Half of the subjects were exposed to

PAGE 19

19 signaled delays (blackout) in the first half of sessions and unsignale d delays in the second half of sessions. The other subjects were exposed to the reverse order. Although higher rates of responding were observed under unsignaled rather th an signaled delays at short delay values (e.g., 1 s), large decreases in response rates unde r unsignaled delays and little reductio n relative to baseline under signaled delays wa s observed at longer delay values. Schaal and Branch (1988) compared the eff ects of briefly signaled versus continuously signaled delay intervals with three pigeons trai ned on a VI-60 s schedule. Under briefly signaled delays, a .5 s change in key light color from re d to green occurred immediately following a key peck that initiated the delay. U nder continuously signaled delays, the key light color changed to green, but remained green throughout the delay and during delivery of the reinforcer. Brief signals maintained response rates under delays up to 9 s; however, continuous signals were required to maintain responding at delays of 27 s. In a follow-up study, Schaal and Branch (1990) showed that key-pecking rate was an incr easing function of the duration of the key light that signaled a 27-s delay. Kelley (2003) extended findings on signaled re inforcement delay from basic research to application with 3 participants diagnosed w ith developmental disabi lities who were taught communicative responses. During signaled delays, a visual (timer that tracked the delay) or auditory (shaking a can of money) stimulus was presented continuously during the delay interval. During unsignaled delays, no stimuli were presented. Communication maintained under progressively longer delays when the delay peri od was signaled but not when it was unsignaled. Vollmer, Borrero, Lalli, and Da niel (1999) investigated the effects of signals within a choice paradigm in which, following FCT, probl em behavior produced access to a smaller, immediate reinforcer (e.g., one chip immedi ately after aggression), whereas communication

PAGE 20

20 produced access to a larger, delayed reinforcer (e .g. three chips after 10 s). When the delay to the large reinforcer was unsignaled, participants made impulsive choices (chose the smaller, immediate reinforcer); when the delay to the large reinforcer was signaled (e.g., placement of timer in front of the participant), participants engaged in self-control (c hose the larger, delayed reinforcer). The mechanism(s) by which signals maintain responding during reinforcement delay has not been identified definitively. However, it is often speculated that response maintenance may be a function of either conditioned reinforcemen t or discrimination (Richards, 1981; Schaal & Branch, 1990). From a conditioning standpoint, a signa l may acquire reinforcing value due to its close temporal relationship with the delivery of food reinforcem ent. From the standpoint of discrimination, nonresponding comes under discrimina tive control of the delay signal, and is therefore less likely to influence responding at other points in time. In applied research, signals have been pr ogrammed during periods of reinforcement and EXT to facilitate discrimination of changing conditions. Specifically, signals function as discriminative stimuli due to a differential correla tion of reinforcement in their presence relative to their absence (Michael, 1982). Fi sher et al. (1998), for example, trained participants to use communication responses to access two different s ources of positive reinforcement (e.g., toys) in the context of FCT. During trai ning, each response was paired with a distinct stimulus (e.g., picture of a person playing with toys). Follo wing training, reinforcem ent contingencies and associated stimuli were alternated every 30 min, and participants displaye d only the response that would be reinforced during that period, suggestin g that the signals had become discriminative for each response.

PAGE 21

21 Hanley et al. (2001) compared the effects of a multiple schedule, in which different stimuli signaled periods of reinforcement a nd EXT, to those of a mixed schedule, which contained similar, but unsignaled periods, to evaluate the effect s of thinning reinforcement for communication following FCT. Under the multiple schedule, a white card was present when the reinforcement component was in effect and co mmunication produced 10-s access to a preferred item on a FR-1 schedule, whereas a red card was present when the EXT component was in effect and communication produced no programmed reinfo rcement. Initially, the component durations were 45 s and 15 s for FR 1 and EXT, respectivel y. The reinforcement schedule was then thinned to 1 min, whereas the EXT schedule was eventu ally increased to 4 min. Schedule thinning progressed when problem behavior remained at or below 85% of the baseline mean for two consecutive sessions. Throughout the assessment, problem behavior produced no programmed consequences. Under the mixed schedule, indiscriminate responding (i.e., high across both schedule components) occurred across components for both participants, and problem behavior occurred at high rates for 1 participant. Under the mu ltiple schedule, differential responding occurred across schedule components (i.e., low rates duri ng EXT and moderately hi gh rates under FR-1), and problem behavior was maintained at low ra tes for both participants The authors concluded that discriminative stimuli, which were not pr esent in the mixed schedule, contributed to differential responding observed under the multiple schedule. Collectively, this research suggests that sign aled procedures may f acilitate control over the rate and temporal distribution of respondi ng under conditions of de layed or intermittent reinforcement. Only one study (Hanley et al., 20 01) has examined the influence of signals on responding during intermittent reinforcement following FCT. However, signals may be

PAGE 22

22 particularly important in main taining communication under such c onditions. Specifically, signals may increase the discrimination of reinforcer availa bility so that communication is more likely to occur when reinforcement is available, and therefor e, more likely to contact reinforcement. If the majority of communication responses result in reinforcement, the contingency between communication and reinforcement is likely to re main strong, resulting in the maintenance of communication over time. Moreover, signals are common co mponents of the natural environment. For example, the green light at the intersection indi cates that accelera ting through the intersection is likely to result in obtaining safe passage to the other side. Walk ing through a door labeled "exit" is correlated with successfully vacating a building. Further, th e presence of a child's mother is likely to engender a number of attention-seeking behavior s, due to the probability by which her mother has delivered attention in the past. As such, th e programming of signals to control appropriate communication following FCT appears to be a pragmatic approach to maintaining communication under intermittent reinforcement. Interval reinforcement schedules, whether fixe d or variable, often ch aracterize the way in which adaptive behavior is reinforced in the na tural environment. As such, the use of interval schedules in maintaining communication followi ng FCT seems logical. Hanley et al. (2001) showed that a FI schedule of reinforcement wa s associated with excessively high rates of communication; however, more moderate rate s of communication might have occurred if reinforcer availability was signaled. No previous studies have investigated th e use of a VI schedule during maintenance following FCT. Nevertheless, VI schedules re flect the manner in which several important

PAGE 23

23 reinforcers are provided to individuals. For exam ple, caregivers often pr ovide attention following varying durations of time gi ven the influence of othe r schedule requirements. In the current study, the e ffectiveness of a signaled VI schedule in maintaining appropriate responding under intermittent reinfor cement was examined. Study 1 involved a basic operant comparison of performance under two VI-30 s schedules: one in which the availability of reinforcement was signaled, and one in which the availability of reinfo rcement was unsignaled. Demonstrating the effects of the VI schedul e using arbitrary responses within a basic arrangement prior to using the schedule clinica lly carries two potential advantages. First, the responses selected for study coul d be made comparable so that differences obtained across the schedules would be more likely to reflect the role of the signal, as opposed to other sources of uncontrolled variability. Second, a ba sic demonstration would allow for the assessment of the VI schedule with the signal presen t and absent, so that a clini cal evaluation of responding under both types of schedule would be unnecessary. As su ch, the risk of exposing participants to an ineffective or even counter-therapeutic procedur e might be minimized. Once a basis for using the signaled VI schedule was established in Study 1, Study 2 was conducted, which involved a clinical evaluation of signaled reinforcer av ailability under a VI schedule following response acquisition with FCT.

PAGE 24

24 CHAPTER 4 STUDY 1: COMPARISON OF UNSIGNALED VERSUS SIGNALED VI SCHEDULES Method Participants and Setting Four individuals who attended a local specia l education school participated. Katy was a 21-year-old female who had been diagnosed w ith mental retardation and who had a limited vocabulary. Charlotte was a 19-year-old nonverbal female who had been diagnosed with mental retardation and who communicated through gestur es. Lesley was a 40-year-old male, who was diagnosed with mental retardation and communi cated through vocalizatio ns and gestures. All participants could follow one-step instructions and were capable of performing simple tasks. George was a 15-year-old male diagnosed with Down's syndrome, who engaged in speech at home but not at school (i.e., was selectively mute). Sessions were conducted in isolated areas of a treatment room located at the school. Session areas were equipped with a table, two chairs, and target task materials (see below). Sessions lasted for five min and were conducted 2 to 5 times per day, 3 to 5 days per week. Response Measurement and Reliability Target behaviors consisted of arbitrary re sponses that were selected based on each participant's level of adaptive functioning. The ta sk for Katy and Lesley consisted of pressing a switch, which was defined by the onset of a light. Ch arlottes task involved placing an index card into a slotted bin, which was defined as the mo ment at which the card left Charlotte's hand. George's task was touching his nos e, and was scored when his inde x finger contacted any part of his nose. Data were collected by trained obser vers on laptop computers during continuous 10-s intervals, and were converted to responses per minute. A s econd observer simultaneously and independently collected data during at least 27% of sessions for each participant. Interobserver

PAGE 25

25 agreement was calculated based on interval-by-inte rval comparisons of observers records. The smaller number of responses scored in each interval was divided by the larger number of responses; these fractions were then average and multiplied by 100%. Mean agreement for the arbitrary task was 89.6% (range, 85%-100% ), 93.5% (range, 83.3%-100%), 90.7% (range, 71.7%-100%), 91.6% (range, 70%-100%) for Katy, Ch arlotte, George, and Lesley, respectively. Stimulus Preference Assessment Paired-stimulus preference assessments (Fishe r et al., 1992) were conducted to identify a highly preferred edible item to use as a reinforcer for task co mpletion. Preference for 9 edible items was assessed. Prior to the assessment, participants were allowed to sample a small portion of each item. During the assessment, items were presented in pairs, and the participant was prompted to select one item. Each item was pair ed once with every other item, in a quasi-random fashion, for a total of 36 trials. The item selected most often, and on at least 80% of trials, was selected to serve as the reinforcer. Procedure The effects of unsignaled and signaled VI schedules on response rates were evaluated using a combined multiple baseline and reversal design. No-reinforcement baseline. An initial baseline was conducted to ensure that participants did not engage in the target task in the absence of reinforcement. Participants were seated at a table facing the therapist with the task positioned dire ctly in front of them. Immediately prior to session, the participant was prompt ed to engage in the response using a three-step prompting procedures consisting of verbal, model, and phys ical prompts, and no programmed consequence occurred. During the session, target respons es resulted in no programmed consequences. Fixed-ratio 1(FR-1) baseline The FR-1 baseline served as an acquisition condition to ensure that the item identified in the preference asse ssment functioned as a reinforcer for target

PAGE 26

26 responses and to establish a base line of responding for compari ng the effects of unsignaled and signaled VI schedules. This condition was iden tical to baseline except that the pre-session prompt resulted in the delivery of the reinforcer and target responses dur ing the session resulted in the delivery of the reinforcer on an FR-1 schedule. Unsignaled VI schedule During this condition, the seating arrangement and task position were identical to the FR-1 baseline; however, target re sponses resulted in the de livery of the reinforcer according to a VI-30 s schedule. The first respons e in a session was reinforced. Thereafter, a reinforcer was delivered for the first response that occurred after a pre-de termined interval of time had elapsed since the last response was re inforced. The programmed time intervals varied around an average value of 30 s and ranged from 15 s to 45 s. The VI 30-s schedule was generated using Microsoft Excel RAND function, a nd a different schedule was used for each session. Signaled VI schedule. This condition was identical to the unsignaled VI schedule condition, except that a signal (e.g., colored card or light) was presented in front of the participant as soon as a time interval had elapsed, and remained pr esent until the participant engaged in a target response. Results Figure 4-1 shows the rate of responding on the arbitrary task during the comparison of the unsignaled and signaled VI schedules. Katy (top panel) emitted no responses during the noreinforcement baseline but high rates of res ponding during the first FR-1 condition. Higher and more variable rates of responding were observed under the unsignaled VI schedule, sometimes as high as 140 rpm. Response rates stabilized during the second FR-1 phase. When the availability of reinforcement was signaled during the sec ond VI 30-s phase, res ponse rates gradually decreased until they became low and stable. An additional measure, the percentage of target

PAGE 27

27 responses that resulted in reinforcement duri ng the last ten sessions of the unsignaled and signaled VI schedules, was calculated to determine the extent to which the signal exerted control over responding. This measure was calculated by di viding the number of reinforcers earned in each session by the number of responses that occurred in each session and multiplying by 100%. The mean percentage of responses resulting in reinforcement for Katy was 5.9% in the unsignaled VI condition and 56.7% in the signaled VI condition. Charlotte (second panel) emitted no responses during the no-reinforcement baseline and moderate rates of responding dur ing the initial FR-1 phase. Her responding increased initially under the unsignaled VI schedule but then fell to near zero rates and ceased completely during the final nine sessions of the condition. The seco nd FR-1 phase was associated with moderate, variable rates of responding. During the Si gnaled VI Schedule condition, her responding increased initially but then d ecreased. In contrast to th e unsignaled VI condition, however, Charlottes responding decreased, but maintained at a low rate as the condition progressed. The mean percentage of responses resulting in re inforcement was 10% in the unsignaled condition and 98.6% in the signaled condition. Lesley (third panel) emitted very few res ponses during the no-reinforcement baseline. His responding increased during the first FR-1 phase and then decreased immediately with the introduction of the unsignaled VI schedule, but gradually increased ac ross the condition and continued to increase during th e second FR-1 phase. Lesleys responding dropped to moderate, stable, rates during the signaled VI condition. The mean percentage of responses resulting in reinforcement was 53.7% in the unsignaled co ndition and 100% in the signaled condition. George (bottom panel) emitted very few responses during the no-reinforcement baseline. His responding increased steeply and remained at high rates during both FR-1 phases.

PAGE 28

28 Decreased rates of responding were observe d under both the unsigna led and signaled VI schedules; however, response rates were somewhat lower under the signaled VI schedule relative to the unsignaled VI schedule. Georges mean per centage of responses resulting in reinforcement was 40.1% in the unsignaled condition a nd 67.4% in the signaled condition. In summary, idiosyncratic patterns of responding were observed across participants during the unsignaled VI schedul e condition, and none of the part icipants engaged in stable, moderate rates of responding. Further, a relative ly small percentage of responses resulted in reinforcement under the unsignaled VI schedul e. By contrast, response patterns became moderate and stable for all pa rticipants, and a large percenta ge of responses resulted in reinforcement, under the signaled VI schedule condition.

PAGE 29

29 0 25 50 75 100 125 150 Katy BLFR-1 Unsignaled VI 30" Signaled VI 30" FR-1 0 5 10 15 20 Charlotte 0 5 10 15 20 Lesley 5 10 15 20 25 30 35 40 45 50 55 60 0 5 10 15 20 GeorgeSESSIONS Figure 4-1. Responses per minute on the arbitrary task under the unsignaled and signaled VI-30" schedules

PAGE 30

30 CHAPTER 5 STUDY 2: CLINICAL EVALUATION OF SIGNALED VI SCHEDULE THINNING Method The results of Study 1 indicated that the signa led VI schedule may be an effective way to maintain desirable response patterns (low rates) under conditions of intermittent reinforcement. Given these positive results in the context of esta blishing and maintaining an arbitrary response, a clinical extension seemed warranted. Thus, Study 2 involved an evaluation of signaled VIschedule thinning as a maintenance procedure for adaptive behavior following its initial acquisition as a replacement for problem behavior. Participants and Setting Six individuals who attended eith er a special education school or outpatient clinic and who were referred for the assessment and treatment of problem behavior participated. Amanda was a verbal 18-year-old female who was diagnosed with Prader-Willi syndrome. Her problem behavior consisted of aggression, which staff reported to occur mo re frequently when preferred items were removed or denied. Jennifer was a ve rbal 15-year-old female, also diagnosed with Prader-Willi Syndrome, who engaged in severe SI B, aggression, and disruption. In several cases, Jennifer's SIB produced open wounds and bleeding that resulted in her being temporarily removed from the school. Robert was a verbal 15year-old male who was diagnosed with autism and who engaged in severe aggression directed toward teachers and other students. Shelly was a verbal 13-year-old female who engaged in seve ral topographies of severe SIB, aggression, and disruption, which had caused her to be permanen tly removed from five previous educational settings. Mandy was a verbal 10-year-old fe male who was diagnosed with Pervasive Developmental Disorder. Her pr oblem behaviors consisted of aggression and disruption, which were reported to occur more often when aske d to do something she did not like to do (e.g.,

PAGE 31

31 schoolwork). Elaine was a nonverbal 10-year-old female with Down's syndrome who had a history of aggression toward adults during wo rk situations. Sessions were conducted in an isolated room at the special education school or in a room equipped with a one-way mirror at the outpatient clinic. Sessions were 10 min in lengt h and were conducted 3 to 6 times a day, 1 to 5 times per week. Response Measurement and Reliability Data were collected on participants probl em behaviors, alternative communication responses, and the presentation of a signal. Amanda's aggression was defined as hitting. Jennifer's SIB consisted of skin picking, scratchi ng, self-biting, and tampering with bandages that covered existing wounds. Jennifer's aggressi on was defined as hitting, kicking, biting, and throwing objects at others. Robert's aggression wa s defined as hitting, kicking, choking, and head butting Like Jennifer, Shellys topographies of SIB incl uded skin picking, scratching, selfbiting, and bandage tampering; however, Shelly also engaged in head banging, head hitting, teeth banging, and nose picking Shelly's aggression consisted of hitting, kicking, biting the body or clothing of others, and throwi ng objects at others. Her disr uption included throwing, breaking, hitting, and kicking objects, and overturning furniture. Mandy's aggression was defined as hitting, kicking, and biting, a nd her disruption was defined as throwing or breaking objects, hitting or kicking surfaces, and drawing on walls. Elaine's aggression was de fined as hair pulling, scratching, and pinching. See Table 5-1 for a detailed listing of operational definitions for problem behaviors. Alternative responses were se lected for participants base d on their level of adaptive functioning. The verbal response, "Can I have my toys back, please?" was selected for Amanda, Robert, and Shelly. The verbal resp onse "Can we talk?" was select ed for Jennifer, and the verbal response "Break please?" was selected for Ma ndy. A voice-output microswitch, which produced

PAGE 32

32 Break please, was selected as the alternative response for Elaine. All verbal responses were scored immediately following emission. Signals were identified individually for pa rticipants based on a signal discrimination assessment (see details in procedure section). Si gnal presentation was de fined as the introduction of the signal into the participant's immediate e nvironment. Signals consisting of colored cards were presented either directly in front of a participant (Mandy) or on a wall (using Velcro) to the participant's immediate left (Amanda, Robert). Colored badges were presented by attaching the badge to the therapist's shirt (Jennifer, Shelly). The light panel, which was affixed to a wall to the participant's immediate left, was presented by ac tivating a switch that turned on the light. Data were collected by trained observers on laptop computers during continuous 10-s intervals. Data on problem behavior and commu nication responses were converted to a rate measure; communication responses also were summa rized as the percentage of responses that occured in the presence of the signal. Relia bility was assessed by having a second observer simultaneously and independently collect data during at least 27% of sessions for each participant. Agreement percentage s were calculated base d on interval-by-inte rval comparisons of observers records. The smaller number of re sponses was divided by the larger number of responses in each interval; these fractions we re then averaged and multiplied by 100%. Mean agreement for problem behavior was 99.6% (range, 95%-100%), 99% (range, 90%-100%), 99.3% (range, 88.1%-100%), 99.5% (range, 97.2%100%), 98.8% (range, 93.3%-100%), and 98.6% (range, 92.5%-100%) for Amanda, Jennife r, Robert, Shelly, Mandy, and Elaine, respectively. Mean re liability for alternative communi cation responses was 98.8% (range, 96.7%-100%), 97.1% (range, 91.6%-100%), 99% (range, 96.7%-100%), 97.8% (range, 86.7%-

PAGE 33

33 100%), 98.1% (range, 93.3%-100%), and 98.9% ( 96.7%-100%) for Amanda, Jennifer, Robert, Shelly, Mandy, and Elaine, respectively. Safety Precautions Certain measures were taken to prevent pa rticipants who engaged in SIB from harm throughout the course of the study. First, attempts at severe SIB were physically blocked by the therapist in the session. Second, criteria for (a) terminating sessions, and (b) seeking additional medical assistance were set fo r each participant prior to ad mission into the study. If the participant met the session term ination criterion (e.g., presence of blood, bruising, or swelling) during the study, he or she was immediately re moved from the session area and first aid was administered (e.g., compress or cold pack applied) If first aid did not result in a decrease in injury, as assessed by the experimenter, the partic ipant was brought to the school or clinic nurse for additional medical assistance. Procedure A functional analysis, FCT intervention, and ev aluation of maintenan ce under a signaled VI schedule were conducted for each participant. Responding during the functional analysis was evaluated within a multielement design; the effects of FCT and VI reinforcement were evaluated in a multiple baseline design across participants. Functional analysis (baseline) The functional analysis was conducted using a pairwise format (Iwata, Duncan, Lerman, Iwata, & Shore, 1994) during which responding in one test condition and one control condition was compared and contrasted. In th e test condition, the reinforcer presumed to maintain behavior was provided contingent on the occurrence of pr oblem behavior (contingent reinforcement; CR). Specifically, test conditions consis ted of contingent access to pref erred items (tangible condition) for Andrea, Robert, and Shelly, a ttention (attention condition) for Je nnifer, and escape from task

PAGE 34

34 demands (demand condition) for Mandy and Elaine During the attention condition, the therapist provided moderately preferred toys to the participant, and instructed the participant to play with the toys while the therapist completed work. The therapist then sat down and ignored the participant unless the participant engaged in pr oblem behavior. Contingent on problem behavior, the therapist provided approximately 10-15 s of at tention in the form of response interruption, reprimands (e.g., dont do that) and statemen ts of concern. During the demand condition, the therapist instructed the participant to perform a variety of educational and vocational tasks using a three-step graduated promp ting procedure (i.e., verbal, m odel, physical). Contingent on problem behavior, the therapist terminated task instructions, moved away from the participant, and did not interact with the participant for 30 s. During the tangible condition, the participant was given 2 min of continuous access to preferre d tangible items (e.g., toys) prior to the session, after which, the therapist removed the preferre d items, and the session began. The therapist returned the items to the participant for 30 s fo llowing each occurrence of problem behavior, and ignored all other behavior. In the control condition, the same reinforcer was available conti nuously and irrespective of the occurrence of problem behavior (noncont ingent reinforcement; NCR). Each condition was conducted in a separate area of a classroom (s chool) or a different treatment room (clinic), and was associated with the therapist wearing a di stinctly colored t-shirt. The functional analysis continued until a pattern of responding emerged in which high rates of problem behavior during the CR condition relative to the NCR condition ve rified that the presumed reinforcer was responsible for behavioral maintenance. The CR condition of the pairwise FA served as a baseline for evaluating treatment effects during the FCT and signa led VI schedule conditions.

PAGE 35

35 Functional communication training (FCT) During the FCT condition, the contingency was reversed from baseline so that appropriate communication, as opposed to problem behavior, produced access to the reinforcer identified for each participant. At the outset of session, the therapist associated with the CR condition implemented antecedent events in a manner identical to the CR condition. That is, preferred items were removed from Andrea, Robert, and Sh elly, Mandy and Elaine were instructed to engage in tasks, and Jennifer was "ignored". Du ring session, each communication response resulted in 30-s access to the reinforcer identified during baseline, whereas, problem behavior resulted in extinction (no reinforcemen t). To facilitate the emission of communication, a second therapist verbally prompted the par ticipant to appropriately communicate for the reinforcer at the beginning of each session. If the participant appropriately requested the reinforcer, no additional prompting occurred. If the participant did not request the reinforcer independently following the initial verbal prompt, additional prompts were given (initially at an interval 25% less than the interresponse time fo r problem behavior during baseline) and then were eliminated using a prompt-f ading procedure, in which the prompt interval was increased by 50%. Additional prompting was required for the fi rst session with Amanda and Robert, and for the first 6 sessions with Elaine. Fading was accomp lished by gradually increasing the latency to the verbal prompt over time. The initial delay to prompt the communicati on response was set at 25% below the average inter-response time between occurrences of problem behavior during the last three CR sessions for Amanda and Robert, a nd 25% below the average latency to engage in problem behavior following an instruction during the last 3 baseline CR sessions for Elaine. The prompt delay was increased by 50% following two se ssions at which problem behavior was at or below a 90% reduction from baseline. The FCT condition was terminated when a participant

PAGE 36

36 engaged in (a) five consecutive sessions with problem behavior at or below a 90% reduction from baseline, and (b) independent communication. Signal discrimination assessment An assessment was conducted to identify a sign al that was likely to be discriminated by each participant. During the assessment, visual stimuli were presented in random locations near the participant, with each presentation constitu ting a trial, and attendi ng behavior (i.e., eyes directed toward stimulus) within 10 s of stimulus presentation was measured. Each stimulus was presented on 10 separate trials. A colored card or badge represented the first stimulus assessed for all participants, with the assumption that th e presentation of a colored card would be easy to use in the natural environment. If the participan t attended to the card on 80% of the trials, the colored card was selected as the signal. If the pa rticipant did not attend to the card on at least 80% of trials, additional stimuli we re assessed. Five of the 6 partic ipants attended to the colored card (Amanda, Robert, Maggie) or badge (Jennifer, Shelly) during the assessment; Elaine required assessment with 5 additional stimuli before a tungsten light panel was identified. Signaled VI schedule thinning During this condition, communication conti nued to produce access to the functional reinforcer, and problem behavior produ ced no programmed consequences; however, communication was reinforced according to a VI schedule that was thinned gradually over time. According to the VI schedule in effect, reinfo rcement was delivered for the first communication response that occurred after an interval of time had elapsed since the last communication response had been reinforced. Within the VI sc hedule, the availability of reinforcement for communication was signaled by th e presentation of the stimul us identified dur ing the signal assessment. The signal was presented to the partic ipant at the precise sec ond that the VI interval elapsed and remained present until the particip ant engaged in communication and received the

PAGE 37

37 reinforcer. Once the signal was removed, the ne xt interval began. Throughout this condition, each session began with the immedi ate presentation of the signal. VI schedules were generated and randomized for each session using the Microsoft Excel RAND function, and schedule values ranged from 50% below to 50% above the VI schedule value used within that session. For example, if communication was reinforced according to a VI 70-s schedule, 50% of the VI values included in terval lengths between 35-70 s, and 50% of the VI values would include interval lengths betwee n 70 and 105 s. The initial VI value was set at 25% below (a) the average inte rresponse time between communicati on responses during the last three FCT sessions (Amanda, Jennife r, Robert, and Shelly) or (b) the average latency to engage in communication following an instruction dur ing the last three FCT sessions (Mandy and Elaine). The VI schedule was thinned by increasing the interval value by 50% following two consecutive sessions in which (a) the rate of problem beha vior did not increase and (b) communication maintained. Thinning was terminat ed following five consecutive sessions in which the rate of problem behavior was at or below a 90% reduction from baseline, communication continued to occu r, and a large propor tion of communication responses occurred in the presence of the signal under a terminal schedule of VI 240 s. Results Figures 5-1 and 5-2 depict the re sults of the clinical evaluation of the signaled VI schedule for the participants whose problem behavior was maintained by social positive reinforcement. During baseline, Amanda (Figure 5-1, top panel) engaged in aggression during the CR condition but not during the NCR condition, indicating that her aggression was maintained by positive (tangible) reinforcement. During FCT, when requests ("Can I have my toys back please") produced access to toys, and aggression was placed on EXT, Amanda immediately began

PAGE 38

38 making requests, while her aggression decrease d to zero. During thinning, when the schedule of reinforcement for requests was increased from VI-30 s to VI-240 s, her requests gradually decreased and maintained at low rates, with no observed increase in problem behavior. The percentage of requests that resulted in reinforcement during thinning was 80%. Jennifer (Figure 5-1, bottom panel) engaged in problem behavior (SIB, aggression, and disruption) during the CR conditi on in baseline but not during th e NCR condition, indicating that these problem behaviors were maintained by positive reinforcement (access to attention). The implementation of FCT resulted in an immediat e increase in requests for attention ("Can we talk?"), and an immediate decrease in problem behavior. Thinning reinforcement from a VI-30 s to a VI-240 s was associated with an immediate decrease in reque sts, which decreased further as the condition progressed. With the exception of a few sessions, thinning was associated with low rates of problem behavior relative to the ba seline CR condition. In addition, 98% of Jennifer's communication responses resulted in access to reinforcement. Robert (Figure 5-2, top panel) engaged in aggression durin g the CR condition in baseline but not during the NCR condition, indicating that his aggression was maintained by positive (tangible) reinforcement. During FCT, when re inforcement was delivere d following appropriate requests ("Can I have my toys back, please?") but not following aggression, requests increased immediately, and aggression decr eased to zero. Thinning the re inforcement schedule to the terminal value of VI-240 s was accomplished with gradual reductions in requests while maintained zero rates of aggression. In additi on, 90% of Roberts requests resulted in reinforcement during thinning. Shelly (Figure 5-2, bottom panel) engaged in SIB and aggression during the CR condition in baseline but not during the N CR condition of baseline, indicati ng that these behaviors were

PAGE 39

39 maintained by positive (tangible) reinforcement. FC T was associated with an increase in requests ("Can I have my toys back please?") and a decr ease in problem behavior to zero. Signaled VI thinning was associated with an immediate decr ease in requests and fu rther decreases throughout the condition. Problem behavior incr eased temporarily when the sche dule was first increased to a VI 30-s but subsequently decrea sed to zero for the remainder of the condition except for one session. Further, 100% of her requests resu lted in reinforcement during thinning. Figure 5-3 shows data for the 2 participants whose problem behavior was maintained by social negative reinforcement (escape from task demands). Mandy (top panel) and Elaine (bottom panel) both engaged in problem beha vior under the CR condition but not in the NCR condition of baseline, verifying that problem behavior was maintained by escape. When FCT was implemented for Mandy, appropriate requests (" Break please") increased to variable rates, whereas aggression and property destruction showed noticeable variability initially, before decreasing to zero. Schedule thinning produced an immediate and then more gradual decrease in requests and low to zero rates of problem behavior. For Mandy, 97% of communication responses resulted in reinforcement during th inning. During Elaines FCT condition, voiceoutput microswitch presses emerged somewhat slow ly but then increased steadily. Her problem behavior showed some variability throughout the first part of the FCT condition but eventually decreased to zero. During sche dule thinning, Elaines microsw itch press decreased gradually throughout the condition, and probl em behavior remained at or near zero except for three sessions. In addition, 97% of Elai nes microswitch press resulted in reinforcement during the thinning phase.

PAGE 40

40 0 1 2 3 AmandaAGG (CR) AGG (NCR) BLFCTSIGNALED VI THINNING Communication AGG 30 45 70 105 160 240 0.0 0.5 1.0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 0 5 10 15 20 0 1 2 3 4 Jennifer30 45 70 105 160 240 Combined Problem Behavior SESSIONS Figure 5-1. Rate of problem behavior and co mmunication during baselin e (CR v. NCR), FCT, and during signaled VI thinni ng for Amanda and Jennifer

PAGE 41

41 0 1 2 3 0.0 0.5 1.0 RobertBLFCTSIGNALED VI THINNING Communication 30 45 70 105 160 240AGG (CR) AGG (NCR) AGG 5 10 15 20 25 30 35 40 45 50 55 60 0 1 2 3 4 5 Shelly20 30 70 105 160 240 45 0 1 2 3 4 5 Combined Problem BehaviorSESSIONS Figure 5-2. Rate of problem behavior and co mmunication during baselin e (CR v. NCR), FCT, and during signaled VI thinni ng for Robert and Shelly

PAGE 42

42 0 1 2 3 4 5 6 MandyAGG (CR) AGG (NCR) BL FCT SIGNALED VI THINNING Communication Problem Behavior 70 105 160 240 0 1 2 3 5 10 15 20 25 30 35 40 45 50 55 60 65 70 0 1 2 0 1 2 3 Elaine15 10 70 105 160 240 20 45 30 SESSIONS Figure 5-3. Rate of problem behavior and co mmunication during baselin e (CR v. NCR), FCT, and during signaled VI thi nning for Mandy and Elaine

PAGE 43

43Table 5-1. Operational definitions Participant Target Operational Definitions Amanda Aggression Hitting: Forceful contact between hand and the body of another individual. Scoring required hand to travel at least 6" and to produce a n audible sound upon contact SIB Skin picking: fingernail to skin contact with movement associated with skin depression or lifting Scratching: movement of a fingernail across the skin Self-biting: closure of teeth on any part of participant's body Bandage tampering: finger to bandage contact with movement of bandage Aggression Hitting : forceful contact between hand and body of another individual. Scoring required hand to travel at least 6" and to produce an a udible sound upon contact Kicking: forceful contact between foot and body of another individual. Same scoring criteria as with aggression Biting: closure of teach on any part of another individual's body Throwing objects at others : object must pass within 2' of individual Jennifer Disruption Ripping/tearing objects : ripping at least 1 or breaking objects such that tear or break in material is visible Throwing objects: object must travel at least 2' Robert Aggression Hitting: Forceful contact between hand and the body of another individual. Scoring required hand to travel at least 6" and to produce a n audible sound upon contact Kicking: forceful contact between foot and body of another individual. Same scoring criteria as with aggression Head butting : Forceful contact between head and the body of another individual Choking: placement around neck of another individual with pressure directed inward SIB Skin picking: fingernail to skin contact with movement associated with skin depression or lifting Scratching: movement of a fingernail across the skin Self-biting: closure of teeth on any part of participant's body Bandage tampering: finger to bandage contact with movement of bandage Head banging: forceful contact between the head and any surface from a distance of 6" or greater with audible sound. Head hitting : forceful contact between the hand and any part of the head from a distance of 6" or greater with audible sound Teeth banging : forceful contact between teeth and any surface Nose picking : insertion of a finger into the nose Aggression Hitting : forceful contact between hand and body of another individual. Scoring required hand to travel at least 6" and to produce an a udible sound upon contact Kicking: forceful contact between foot and body of another individual. Same scoring criteria as with aggression Biting: closure of teach on any part of another individual's body or clothes Throwing objects at others : object must pass within 2' of individual Shelly Disruption Throwing objects : object must travel at least 2' Breaking objects : tear or break in material must be visible Hitting/Kicking objects : forceful contact between hand/foot and any surface. Body part must travel at least 6" and produce audible sound Overturning furniture : furniture must be displaced 45 degrees from upright Aggression Hitting : forceful contact between hand and body of another individual. Scoring required hand to travel at least 6" and to produce an a udible sound upon contact Kicking: forceful contact between foot and body of another individual. Same scoring criteria as with aggression Biting: closure of teach on any part of another individual's body or clothes Mandy Property Destruction Breaking objects : tear or break in material must be visible Hitting/Kicking objects : forceful contact between hand/foot and any surface. Body part must travel at least 6" and produce audible sound Writing on walls Elaine Aggression Hair pulling: grasping hair with any part of hand and pulling in direction away from head Scratching: movement of fingernail across skin Pinching: closure of at least two fingers around the skin with pressure directed toward the skin

PAGE 44

44 CHAPTER 6 DISCUSSION The present studies provide both a basic dem onstration and a clini cal extension of the effects of signaled VI reinforcement on responding. Data from both studies indicated that this type of schedule may facilitate maintenan ce when relatively low rates of responding are desirable following initial response acquisition. In Study 1, response rates were compared under two VI-30 s schedules that were identical except that the end of the VI interval (i.e., the availability of reinforcement) was unsignaled in one schedule but signaled by the pres entation of a visual stimulus in the other. Under the unsignaled VI-30 s schedule, only 1 of the 4 participants engaged in moderate response rates (George), and a small percentage of responses resulted in reinforcement across participants. By contrast, responding became modera te and efficient for all participants under the Signaled VI 30-s schedule, occu rring at approximately 2.0 rpm at the end of the condition. Further, a large proportion of responses resu lted in reinforcement across participants. As the unsignaled and signaled conditions diffe red only with respect to the presence and absence of the signal, differences in respondi ng observed between the two conditions can be attributed to the effects of the si gnal. In the absence of the signal, it is likely that participants could not discriminate the availability of reinfo rcement (the end of the interval), particularly given the variable nature of the schedule. Hence, other factor s, such as those discussed below, may have contributed to the vari ability in responding observed. The presence of the signal may have incr eased the likelihood that participants could discriminate the availability of reinforcement, such that responding occurred primarily in the presence of the signal, when it was likely to co ntact reinforcement. Given the high probability of

PAGE 45

45 reinforcement for responding, the response-reinforcer contingenc y was likely to remain strong, resulting in low but persis tent rates of behavior. The results of the signaled VI condition are consistent with those reported by Azrin (1958), who examined the effects of an auditory stim ulus, noise, on the FI observing responses of 80 soldiers. A meter needle was deflected every 3 min, and soldiers were trained to release a switch to bring the needle back to its normal position. The room was kept dark, and presses on a button illuminated the needle for 0.1 s; the sight of the needle deflecting was considered the reinforcer. During baseline, responses occurred rarely during the fi rst 2 min of the FI schedule; however, high response rates occurred throughout the 3rd min of the FI until the needle was detected and restored. When a 110-decibel white noise was co ntinuously present throughout the first 165 s of the interval but was discontinued for the 15 s pr eceding a needle deflection, observing responses occurred only during the 15-s pe riod of quiet and never during the initial 165 s of noise. The temporal relation of noise to the target was then reversed so that 165 s of quiet preceded a 15-s period of noise. Soldiers re sponded by making observing responses during the 165 s of quiet and no observing responses durin g the 15 s of noise. Recovery occurred around interval 14, and soldiers began to respond only during the 15 s of noise At that point, the temporal relation of noise to the target was agai n reversed to the initial conditions, and soldiers again responded exclusively duri ng periods of quiet. Azrin (1958) concluded that the noise (or quiet) had acquired discriminative control over observing responses to the exclusion of control by the temporal properties of the FI schedule. Previous research with nonhuman subjects has shown that unsignaled VI schedules tend to produce low to moderate, stable response rates (Ferster & Skinner, 1957). In the current study, however, only 1 participant enga ged in moderate response ra tes throughout the unsignaled

PAGE 46

46 condition (George), and the responding of the ot her participants was characterized by a high degree of variability. High response rates, which stabilized toward the end of the condition, were observed for 1 participant (Katy) whereas modera te, but increasing respon se rates were observed for Lesley. Initially high response rates, that d ecreased to low to zero ra tes after four sessions, were observed for Charlotte. Responding that is inconsistent with a curr ent reinforcement schedule, such as that observed with Katy, Charlotte, and Lesley, may refl ect the influence of historical variables such as prior experience with a different reinforcem ent schedule (Weiner, 1969). Freeman and Lattal (1992), for example, trained three pigeons on an FR-1 schedule before exposing them to a VI schedule, while discriminative stimuli were he ld constant across the schedules. Following training, all pigeons engaged in high response rates during the initial 8 sessions of the VI condition, and one pigeon engaged in high response rates across all sessions of the VI condition. Historical control, such as that identified in the study by Freeman and Lattal (1992) may have accounted for the persistent high rate re sponding observed under the VI schedule for Katy and Lesley. In Charlotte's case, however, high response rates may have interacted with the temporal delivery of reinforcement under th e VI schedule to produce the steep decline in responding after four sessions. Mo re specifically, if high response rates were emitted under the VI schedule, and reinforcement was not delivere d for an average of 30 s, several responses would have been unreinforced, resulting in EXT of the response. In support of this, Charlotte's data under the VI schedule are characteristic of responding during EXT. That is, a slight increase in responding resembling an extin ction-induced burst was observed immediately after Charlotte switched schedules, followed by systematic decreases in response rate to zero.

PAGE 47

47 The generally decreasing trend observed acro ss the unsignaled and signaled VI conditions for George may suggest the occurrence of an order effect. That is, responding under the unsignaled VI schedule may have influenced responding under the signaled VI schedule. However, the greater percentage of responses that produced reinforcement under the signaled VI schedule (67.4%) rela tive to the unsignaled VI schedule ( 40.1%) suggested that the signal, as opposed to a previous history with the VI sche dule, accounted for the moderate and efficient responding observed under the signa led VI schedule for George. To summarize, the signaled VI schedule wa s effective in maintaining moderate and efficient response rates, which provided a basis fo r evaluating the schedule in an applied context in Study 2. Study 2 involved a clinical evaluation of the signaled VI schedule as a maintenance procedure following acquisition of an alternativ e response to problem behavior. A pairwise functional analysis similar to that used by Ha nley et al. (2001) id entified a maintaining contingency for each participants problem behavior. Although the pairwise assessment included only one test condition and did not rule out othe r environmental influenc es on problem behavior, it was sufficient to develop an intervention strategy based on the id entified function. The differential reinforcement procedures implemented as FCT were effective in decreasing problem behavior and increasing communica tion for all participants. These results were consistent with those from a large body of resear ch supporting the use of FCT as treatment for problem behavior maintained by social consequences (e.g., Hagopi an et al., 1998; Wacker et al. 1990). The vast majority of studies on FCT involve continuous schedules of reinforcement for alternative behavior (e.g., Carr & Durand, 1985) Although continuous reinforcement may be advantageous during response acquisition, a compar able reinforcement schedule is not likely to

PAGE 48

48 exist in the natural environment, where severa l factors preclude the immediate delivery of reinforcement (e.g., parent atte nding to another child). To pr eserve the integrity of FCT outcomes under naturalistic conditions, stra tegies for maintaining communication under conditions of intermittent or delayed reinforcement are necessary. A survey of the basic and applied literature re vealed that response maintenance is related to preservation of the response-reinforcer conti ngency (Hanley et al., 2001; Sizemore & Lattal, 1978), and that contingency preservation may be f acilitated through the use of signals that are correlated with the availability of reinfor cement (Hanley et al., 2001). In the current study, signaled VI thinning, in which al ternative behavior was reinforced less frequently over time according to an increasing VI schedule that includ ed a signal when the interval terminated, was used as the maintenance strate gy. The schedule was effective in maintaining treatment effects under increasingly intermittent reinforcement. That is, as the reinforcement schedule was thinned to the terminal value of 240 s, problem behavi or remained low, and alternative behavior decreased to rates that reflected th e availability of reinforcement. Data indicating that a high per centage of alterna tive responses resulted in reinforcement during thinning provides evidence th at the signal had acquired stimul us control over appropriate behavior. At any given interval value duri ng thinning, there are limited opportunities for reinforcement. Under a VI-70 s schedule, for example, there are approximately 8 potential reinforcer deliveries within a 10-min session. Fo r a high percentage of responses to contact reinforcement, a participant would have to respond when, and only when reinforcement was available. Across participants, 93.7% of communi cation responses resulte d in reinforcement, indicating that participants di scriminated the availability of reinforcement and responded accordingly.

PAGE 49

49 Results of the current study ar e consistent with those of Ha nley et al. (2001), showing that signaled reinforcer availability produces moderate rates of comm unication that occur at appropriate points in time (i.e., when reinforcement was available) and low rates of problem behavior. The differences between the multiple-sch edule used by Hanley et al. (2001) and the VI schedule in the current study suggest different ty pes of application. When only one response is sufficient for reinforcement, but reinforcement w ill be delayed or available only once in awhile, the signaled VI schedule may be preferred. Most relevant to the use of the signaled VI schedule are instructional or other demand -related situations in which only one request for a break or task termination is required and a ppropriate; however, completion of the task is mandatory (e.g., brushing teeth). By contrast, when higher rates of be havior are either accept able or desirable, but reinforcement will be unavailable for periods of time, the multiple-schedule may be advantageous. For example, repeated requests fo r attention are consider ed appropriate during periods of conversation or play, and may be accommodated nicely within a multiple schedule. During reinforcement thinning, several sess ions were conducted before alternative behavior became moderate and efficient, whic h suggests that discrimi native control over responding may require time to develop. Future investigations may examine procedures for facilitating the establishment of stimulus contro l prior to exposing newl y acquired behavior to conditions of intermittent or delayed reinforcement. For example, presenting a stimulus during the initial reinforcement phase (FR-1) would esta blish a longer history of reinforcement in the presence of the signal and may be an efficient way to generate stimulus control over responding prior to using the signal clinically. The structure of the signaled VI procedur e is potentially susceptible to producing inadvertent reinforcement of problem behavior. More specifically, the onset of the signal is time-

PAGE 50

50 based, and if problem behavior occurs temporally close to the onset of the signal, it may become associated with the signal, and the subsequent production of reinfo rcement. The terminal result of this process is that the signal may become discriminative for the occurrence of problem behavior (as opposed to alternative behavior), en gendering high rates of problem behavior in its presence. Although not a problem in the current inve stigation, therapeutic use of the signaled VI procedure may incorporate a changeover dela y contingent on the occurrence of problem behavior. A changeover delay invo lves a brief delay to reinfor cement (or onset of the signal) following problem behavior, such that inadverten t reinforcement of problem behavior can be reduced. Future research may also examine additional therapeutic uses of signals. In some situations, it is not possible or acceptable to use extinction during the treatment of problem behavior, and discriminative control over appr opriate responding may decrease the likelihood that problem behavior will occur and inadvertently contact reinforcement. For example, severe SIB may be maintained by attention; however, th e delivery of attention following self-inflicted injury may be unavoidable. If, however, appropri ate requests for attent ion are brought under the control of signals, appropriate requests may o ccur to the exclusion of problem behavior.

PAGE 51

51 LIST OF REFERENCES Azrin, N. H. (1958). Some effect s of noise on human behavior. Journal of the Experimental Analysis of Behavior, 1, 183-200. Azzi, R., Fix, D. S. R., Keller, F. S., & Rocha E Silva, M. I. (1964). Exteroceptive control of response under delayed reinforcement. Journal of the Experimental Analysis of Behavior, 7, 159-162. Baum, W. M. (1973). The corre lation-based law of effect Journal of the Expe rimental Analysis of Behavior, 20, 137-153. Carr, E. G., & Durand, V. M. (1985). Reduc ing behavior problems through functional communication training. Journal of Applied Behavior Analysis, 18, 111-126. Durand, V. M., & Carr, E. G. (1987) Social influences on "self-stim ulatory" behavior: Analysis and treatment application. Journal of Applied Behavior Analysis, 20, 119-132. Durand, V. M., & Carr, E. G. (1991). Functiona l communication training to reduce challenging behavior: Maintenance and a pplication in new settings. Journal of Applied Behavior Analysis, 24, 251-264. Ferster, C. B., & Skinner, B. F. (1957). Schedules of reinforcement. New York: AppletonCentury-Crofts. Fisher, W., Piazza, C. C., Bowman, L. G., Hagopi an, L. P., Owens, J. C., Slevin I. (1992) A comparison of two approaches for identifying reinforcers for persons with severe and profound disabilities. Journal of Applied Behavior Analysis, 5, 491 Fisher, W. W., Thompson, R. H., Hagopian, L. P ., Bowman, L. G., Krug, A. (2000). Facilitating tolerance of delayed reinforcement during Functional Communication Training. Journal of the Experimental An alysis of Behavior, 24 3-29. Freeman, T. J., & Lattal, K. A. (1992). S timulus control of behavioral history. Journal of the Experimental Analysis of Behavior, 57, 5-15. Fyffe, C. C., Kahng, S., Fittro, E., & Russell, D. (2004). Functional analysis and treatment of inappropriate sexual behavior. Journal of Applied Behavior Analysis, 37 401-404. Hagopian, L. P., Fisher, W. W., Sullivan, M. T., Acquisto, J., & LeBlanc, L. A. (1998). Effectiveness of functional communication training with and w ithout extinction and punishment: A summary of 21 inpatient cases. Journal of Applied Behavior Analysis, 31, 211-235. Hammond, L. J. (1980). The effect of conti ngency upon the appetitive conditioning of freeoperant behavior. Journal of the Experimental Analysis of Behavior, 34, 297-304.

PAGE 52

52 Hanley, G. P., Iwata, B. A., & Thompson, R. H. (2001). Reinforcement schedule thinning following treatment with functi onal communication training. J ournal of Applied Behavior Analysis, 34 17-37. Horner, R. D., & Day, H. M. (1980). The eff ects of response efficiency on functionally equivalent competing behaviors. Journal of Applied Behavior Analysis, 24, 719-732. Iwata, B. A., Dorsey, M. F., Slifer, K. J., Ba uman, K. E., & Richman, G. S. (1982). Toward a functional analysis of self-injury. Analysis and Intervention in Developmental Disabilities, 192 193-120. Iwata, B. A., Pace, G. M., Kalsher, M. J., Cowdery, G. E., & Cataldo, M. F. (1990). Experimental analysis and extinction of self-injurious escape behavior. Journal of Applied Behavior Analysis, 23, 11-27. Kahng, S., Iwata, B. A., DeLeon, I. G., & Worsdell, A. S. (1997). Evaluation of the "control over reinforcement" component in functional communication training. Journal of Applied Behavior Analysis, 30, 267-277. Kelley, M. E. (2003). The effects of signals on responding during delayed reinforcement. (Doctoral Dissertation, Louisi ana State University, 2003). Dissertation Abstracts International, 64, 1478. Kern, L., Mauk, J. E., Marder, T. J., & Mace, F. C. (1995). Functional analysis and intervention for breath holding. Journal of Applied Behavior Analysis, 28, 339-340. Lalli, J. S., Vollmer, T. R., Progar, P. R., Wr ight, C., Borrero, J., Daniel, D., et al. (1999). Competition between positive and negative reinforcement in the treatment of escape behavior. Journal of Applied Behavior Analysis, 32, 285-296. Lattal, K. A. (1984). Signal func tions in delayed reinforcement. Journal of the Experimental Analysis of Behavior, 42 239-253. Lattal, K. A. (1995). Continge ncy and behavior analysis. The Behavior Analyst, 18, 204. Meyer, K. A. (1999). Functional analysis and treatment of problem behavior exhibited by elementary school children. Journal of Applied Behavior Analysis, 32, 229-232. Michael, J. (1982). Distinguishing between discri minative and motivational functions of stimuli. Journal of the Experimental Analysis of Behavior, 37, 149-155. Pierce, C.H., Hanford, P.V., & Zimmerman, J. (1972) Effects of different delay of reinforcement procedures on variab le-interval responding. Journal of the Experi mental Analysis of Behavior, 18, 141-146. Richards, R.W. (1981). A comparison of signale d and unsignaled delay of reinforcement. Journal of the Experimental Analysis of Behavior, 35 145-152.

PAGE 53

53 Schaal, D. W. & Branch, M. N. (1988). Re sponding of pigeons under variable-interval schedules of unsignaled, briefly signaled, and completely signaled delays to reinforcement. Journal of the Experimental Analysis of Behavior, 50 33-54. Shirley, M. J., Iwata, B. A., Kahng, S., Mazal eski, J. L., & Lerman, D. C. (1997). Does functional communication training compete with ongoing contingencies of reinforcement? An analysis during response acquisition and maintenance. Journal of Applied Behavior Analysis, 30, 93-104. Sizemore, O.J., & Lattal, K.A. (1978). Unsignaled delay of reinforcement in variable-interval schedules. Journal of the Experimental Analysis of Behavior, 30, 169-175. Smith, R. G., Iwata, B. A., Goh, H., & Shore, B. A. (1995). Analysis of establishing operations for self-injury maintained by escape. Journal of Applied Behavior Analysis, 28, 515-535. Tarbox, R. S. F., Wallace, M. D., & Williams L. (2003). Assessment and treatment of elopement: A replication and extension. Journal of Applied Behavior Analysis, 36, 239244. Thompson, R. H., Fisher, W. W., Piazza, C. C., & Kuhn, D. E. (1998). The evaluation and treatment of aggression maintained by attention and automatic reinforcement. Journal of Applied Behavior Analysis, 31 103-116. Vollmer, T. R., Borrero, J. C., Lalli, J. S., & Daniel, D. (1999). Evaluation self-control and impulsivity in children with severe behavior disorders. Journal of Applied Behavior Analysis, 32, 451-466. Vollmer, T. R., Iwata, B. A., Za rcone, J. R., Smith, R. G., & Mazal eski, J. L. (1993). The role of attention in the treatment of attention-mainta ined self-injurious behavior: Noncontingent reinforcement and differential reinforcement of other behavior. Journal of Applied Behavior Analysis, 26, 9-21. Vollmer, T. R., Marcus, B. A., & LeBlanc, L. (1994). Treatment of self-injury and hand mouthing following inconclusi ve functional analyses. Journal of Applied Behavior Analysis, 27 331-344. Vollmer, T. R., Roane, H. S., & Ringdahl, J. E ., & Marcus, B. A. (1994). Evaluation treatment challenges with differential reinfor cement of alternative behavior. Journal of Applied Behavior Analysis, 32 9-23. Wacker, D. P., Steege, M. W., Northrup, J., Sass o, G., Berg, W., Reimers, T., et al. (1990). A component analysis of func tional communication training ac ross three topographies of severe behavior problems. Journal of Applied Behavior Analysis, 23 417-429.

PAGE 54

54 Williams, B.A. (1976). The effects of unsignaled delayed reinforcement. Journal of the Experimental Analysis of Behavior, 26 441-449. Worsdell, A. S., Iwata, B. A., Hanley, G. P ., Thompson, R. H., & Kahng, S. (2000). Effects of continuous and intermittent reinforcement for problem behavior during functional communication training. Journal of Applied Behavior Analysis, 33, 167-179.

PAGE 55

55 BIOGRAPHICAL SKETCH Carrie Dempsey was born in Santa Rosa, Calif ornia (CA), in 1975. She graduated with a Bachelor of Arts degree in ps ychology and Latin American & Iber ian studies from the University of California-Santa Barbara in 1995. Following graduation, Carrie pursue d higher education at the University of the Pacific in Stockton, CA where she was first exposed to behavioral psychology and became inspired to develop as a be havior analyst. Carrie entered graduate school at the University of Florida in 2002 under the supervision of Brian A. Iwata. Upon graduation, she intends to return to CA to establish a research-oriented treatment center for severe problem behavior. In her spare time, Carrie enjoys be ing in the outdoors, ka yaking, traveling, and learning about various aspect s of computer technology.


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 E20101113_AAAAHD INGEST_TIME 2010-11-13T08:47:17Z PACKAGE UFE0021351_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES
FILE SIZE 25582 DFID F20101113_AAAXUE ORIGIN DEPOSITOR PATH dempsey_c_Page_38.QC.jpg GLOBAL false PRESERVATION BIT MESSAGE_DIGEST ALGORITHM MD5
f7582b233df68d7553943ae390b9af13
SHA-1
dcf174dea591437dbe42d0ca7120abb93a65fb98
1053954 F20101113_AAAXPH dempsey_c_Page_14.tif
b3f0418edec18925d23034f277b29698
9cec1fb353bbf49bcc02aad19f027cab87268a1d
7019 F20101113_AAAXKK dempsey_c_Page_35thm.jpg
0a93cf8085d2abeebc682645200a547a
e1938af0a7c230fa6f2d59d04cccee2f19c6df06
4258 F20101113_AAAXUF dempsey_c_Page_29thm.jpg
1c0638a1086db3de995ed01491411427
a080ac667e6be8fbf131f05ed775e16cb5bb8010
F20101113_AAAXPI dempsey_c_Page_18.tif
ca5ade45231840818bf5037fbc8e5326
de6103c53e90830ef52e30c7f6b26fa552531c5a
2180 F20101113_AAAXKL dempsey_c_Page_16.txt
d9f02222bdf17b2ca1529dc9dbb1d8e6
f90ef693a6bea671e302ea9c413cc38d00c769da
24293 F20101113_AAAXUG dempsey_c_Page_37.QC.jpg
a463d9cc39a248ed59de4bcd3c28e6c3
642718ba761d82952f9547a4334aff8307156ba1
F20101113_AAAXPJ dempsey_c_Page_19.tif
82c4f9d9365bb452f9406f9b371ccd6a
ed9591791f12746708843616dfa031f1a1e7417d
52458 F20101113_AAAXKM dempsey_c_Page_18.pro
a62c77db3fda789516c338d3f8080163
197c74511a6adfdf007b29246985805b8940d345
6849 F20101113_AAAXUH dempsey_c_Page_11thm.jpg
e4668509834a23688f95527963f6cf3f
37928b83ef5651055a9345f49bc29619d42d3047
F20101113_AAAXPK dempsey_c_Page_20.tif
df9d95b62fb9e21bd1920667e2002d89
bdd081175d2fb1bd4041e5d4fd93c8d5be6c0650
2324 F20101113_AAAXKN dempsey_c_Page_01thm.jpg
144adc02536e062e0dcb2cd759367417
fe6182baeb64099c48adcd09d53974f95ccd1b6e
F20101113_AAAXPL dempsey_c_Page_23.tif
b453b1614c9fd23724825eb4852cb43b
8d82b35614217005b8340c7688b5fc4ce9e83b35
74122 F20101113_AAAXKO dempsey_c_Page_37.jpg
43ca006cbe53c1d8fe7cc0afc6344f28
f44eeeac40d1c10140c6f89cdc386d18f19400dd
25782 F20101113_AAAXUI dempsey_c_Page_16.QC.jpg
671b28dfacc6459fbde7361ad144109f
606c8078b8fbfcfca632326206099f27eccc60ab
F20101113_AAAXPM dempsey_c_Page_25.tif
e611842c03bda68dec1c2b50c8d2c240
d027257ed353c118596260f8d0c56770f6cc3222
109932 F20101113_AAAXKP dempsey_c_Page_24.jp2
49613f0d35042ebdaf924f45985c78e7
00e1e77df4b8f13545335f831ddd4892d41a6e13
6586 F20101113_AAAXUJ dempsey_c_Page_18thm.jpg
dc132905abe93276280c69c5a70af2aa
7873114971fe87798411a3c0f36cd55690590df6
25271604 F20101113_AAAXPN dempsey_c_Page_26.tif
925c79f2804fea222edd128eb1a6b746
f30a1dfabde7fc8f018e23e4a8d1d7fa139c4304
67643 F20101113_AAAXKQ dempsey_c_Page_05.pro
01190fda1867e14d06421c17cbd82b6a
6f033c9ca9debc45804f45063c1cc31ccaf80a57
3054 F20101113_AAAXUK dempsey_c_Page_08thm.jpg
a1482cac0cffe6b89fb0e874b0a19633
c6bf1404cfffb214fb2b181238e3e454c377c9ea
F20101113_AAAXPO dempsey_c_Page_27.tif
b350886f41a700652a33bd4003ca9973
7dc1cef749a293b364464c4a0795e6332e4488c2
50193 F20101113_AAAXKR dempsey_c_Page_17.pro
32e18d59c3dc9e27a0257e8e2cc44797
37a0df5c1a78f279aadfa9aeb18fc921f8858bb9
6853 F20101113_AAAXUL dempsey_c_Page_45thm.jpg
c498ff3df825c596cb45eee3d83541ba
fa7c93003baebe4acca2337c0cd1e748d39b5128
F20101113_AAAXPP dempsey_c_Page_30.tif
d8a5abe5f9ff6fe25760d8fb1fd31af8
07ed0b66021a75562f3d5e5c20fe69172b2e255f
19670 F20101113_AAAXKS dempsey_c_Page_04.pro
f80effd25a46e87d15d707daea88925c
82deb1c55caeb6d8ec60f938b04dee0cb6b8e41e
24983 F20101113_AAAXUM dempsey_c_Page_27.QC.jpg
50b3f70638192dfd872ccc2ebc121ea6
f17ed86bca2a3760d46278671dc125c59717c002
F20101113_AAAXPQ dempsey_c_Page_31.tif
57be7dbb5426f76125d5e8688ad9d1a5
0039f8557560ce6416179a7d6f4cc9b3a8ca5c94
24765 F20101113_AAAXKT dempsey_c_Page_34.QC.jpg
bacdb2135e4456c2d35a0c3cd84b18a5
9575c482da4bf98adc995c91d454e917908e8a81
25468 F20101113_AAAXUN dempsey_c_Page_45.QC.jpg
02f875a04ebf8643002057b611417572
5ac165881dd2a1f74856162f7b192ccce7296300
F20101113_AAAXPR dempsey_c_Page_32.tif
b37341f9f900d6f306293154f5107cef
b33feed42bf4cf990a969cc345f8d0c0e0cf81b6
12486 F20101113_AAAXKU dempsey_c_Page_40.QC.jpg
4a3c73751dbbd9093bebf40faa0a7817
5514fce0f678669e4262cb55e3c99098055ce1c9
7625 F20101113_AAAXUO dempsey_c_Page_01.QC.jpg
c9aabee4440d2a429e0185f06bd19851
2e6ec491c911c1585c0610d16a3941573d332977
F20101113_AAAXPS dempsey_c_Page_33.tif
5495bee833f520356405671baeb10565
078ff86f98c3eddd61cb0d07e9a5fddb6af64a27
106352 F20101113_AAAXKV dempsey_c_Page_22.jp2
816d423354e39ddb9a5c35b3610db5fc
52e99631c942f89318ff4648c77c6c1d9a1f8b08
4185 F20101113_AAAXUP dempsey_c_Page_41thm.jpg
25f13690f67efdb454d5a80bd53f0bf4
3fbf84b2b7bcce142b80da609437b3e864011a4f
F20101113_AAAXPT dempsey_c_Page_37.tif
c014c797c032504a82a6ae9be160b488
00ed8c18ae0f4b5d62b79ad51e105d7fc962859e
1051947 F20101113_AAAXKW dempsey_c_Page_05.jp2
e44abcf61f700020cda296af8e516e36
aa18c58e81764140f3c8d16f7d202267e188b655
6583 F20101113_AAAXUQ dempsey_c_Page_30thm.jpg
0c38f5ebf95da8e4d8ddc3e74cbbba86
fc0ddd5fcd4082ae4653e5037602c9a55f082271
F20101113_AAAXPU dempsey_c_Page_38.tif
f58fe77dbe73fbca0e7f7060d79f0738
1e8e13fc12a985858e4c1f957174dc0fc49e349c
F20101113_AAAXKX dempsey_c_Page_01.tif
9ce9a05b798ef91ed7a09809b9ada569
8bfc78fe146fd239969637d6f3602a6494b72f1c
6827 F20101113_AAAXUR dempsey_c_Page_37thm.jpg
378e4a006f60ee27d1dfd6b57d04dd8c
91f3976950f414b3c280beda18202225d0e29270
F20101113_AAAXPV dempsey_c_Page_39.tif
751623fdbf6b5f302706fa95a3211425
ea01bd03e8e35dc62a67722df38b7d7751f1264e
F20101113_AAAXKY dempsey_c_Page_21.tif
f1db7948e3e03520017cd423c917e39e
882062de93ff911fc39db6a40ada6aaaa5da03bd
26520 F20101113_AAAXUS dempsey_c_Page_52.QC.jpg
855324f8a6944ec0f3592de2dab2b16b
204486590906f25e366a83d668ac0ff3c8ccc3b6
F20101113_AAAXPW dempsey_c_Page_40.tif
d36c21ba13ec610f9760a1ac7be771ff
c63dd0fa23db3aa0496084380ef7664e63829778
6862 F20101113_AAAXIA dempsey_c_Page_16thm.jpg
dff206c590220216c9226d872ebcd888
7cddf5b686f96f0e6e899d33dd8d6fb336e3caad
24498 F20101113_AAAXUT dempsey_c_Page_25.QC.jpg
f9cacf1d785ee6beb6b604e2b5e1443f
cfcd7e4d0e5467bdceb29502ce5a79093c0c6865
F20101113_AAAXPX dempsey_c_Page_44.tif
c56835ceaeec792711782ef76dec0002
d9f6fdd711706c681d7b163e8cdd0ecbffb87d19
11035 F20101113_AAAXIB dempsey_c_Page_06.pro
a1acdf217a4f4964558d3ee91ddc4ea4
0931f8baac664d1f62bf3b36cb0866a7aba06350
38275 F20101113_AAAXKZ dempsey_c_Page_41.jpg
b0924ee2cbc52da8cf9cea48beb9698a
975747d0b06c00279171760b15d032207bc58d33
2129 F20101113_AAAXUU dempsey_c_Page_54thm.jpg
d757fa7b0bae10c7b5c6a2c11f4b98b3
b176ccf9c8febb87ce6c4831f06013c1635f2faf
F20101113_AAAXPY dempsey_c_Page_45.tif
596e8ee970375ed031658ca54021acfa
b1bfa4287cc7415160b60fdefdb7bb90876fa392
55317 F20101113_AAAXIC dempsey_c_Page_45.pro
274df922e7edc2fe41a63050cef421a8
680e92e2108783fdb4dda71cd5ac24f532c3d9de
1812 F20101113_AAAXUV dempsey_c_Page_10thm.jpg
ae33c8ba4fc70c9c2960be65d2587323
78d1d20808784ff8c1c7b98c8583b79fa9a57e99
F20101113_AAAXPZ dempsey_c_Page_46.tif
03b1fafef2bb35a3bec8ad390719b249
d7acb7c3316014390c38733ae72761ae163904c8
6647 F20101113_AAAXID dempsey_c_Page_21thm.jpg
e4ce9a2ca07bfa619f3ee7e4cca64916
6038a32704b13bda8bc6f74389580f2a4d4ea50d
6495 F20101113_AAAXUW dempsey_c_Page_33thm.jpg
f9f499dbc2b9e77bf0b2a35e91178513
6537f54328554f60c39ad931b139f55fb0437a37
37345 F20101113_AAAXNA dempsey_c_Page_29.jpg
a2f7dbedf7f039024180e1abb4237da6
146dee0040a38beb7b0f422fafba30e7bda1f176
1380 F20101113_AAAXIE dempsey_c_Page_03thm.jpg
086c49507cf2f8f8edaecfd230f5ad3b
920a05130cad9bc6c3bc953d4d47caa85a96a014
25711 F20101113_AAAXUX dempsey_c_Page_49.QC.jpg
8c97837e119b3afa1a9923a9eaa9e76f
41b0c066b791d76cea0af24a756f540ea1838067
74876 F20101113_AAAXNB dempsey_c_Page_30.jpg
8f80770e1c158983e2e11937c795a682
11f46a6ca8ee7dd8b9a5e9c2ae7d1794c6f7e2e1
132406 F20101113_AAAXIF dempsey_c_Page_51.jp2
b426e4daf2f2c167d4ccaa6baad2320d
a7ef48d04ca37df1ab292b47af4c55d913ff84d2
1577 F20101113_AAAXSA dempsey_c_Page_23.txt
f34a827c679baeaf558b24e59155f10e
44b1d8d986023fb9f03c7af8a3248d3441a114dd
85360 F20101113_AAAXUY UFE0021351_00001.xml FULL
b7a896f8ae214d516b4c341f175e2700
ef6b64d425b4e72c829f65ad43fe99d1a6165f32
71930 F20101113_AAAXNC dempsey_c_Page_33.jpg
91acc9d4ea81f647c660bde87deac7ae
dc866b7fc77a15a4bbb18d248ff65f49e25fd612
377518 F20101113_AAAXIG dempsey_c_Page_29.jp2
617cc8d060eaa3f3083a68b510c21778
2da75037998bd18ccaddf25470340746659d537b
1351 F20101113_AAAXUZ dempsey_c_Page_02thm.jpg
fd41497700324cf70b8b28225e67282e
f4d84895193061784133c1f81fe7938db4a8b956
79811 F20101113_AAAXND dempsey_c_Page_35.jpg
dca847916028624b3a4c5bb66cecdc83
ff76d1de656c620ef7f5a6e776eb34735f7ea609
446980 F20101113_AAAXIH dempsey_c_Page_40.jp2
13f66ca23b596db41623df8ad1d99807
19356bda0c776a72e954c713d562dff31368f297
2026 F20101113_AAAXSB dempsey_c_Page_24.txt
ccd210767af02a0dc5244aa3229e06ff
c7ebda09b60bcfe78c88e9a8975d07064ecab107
77661 F20101113_AAAXNE dempsey_c_Page_36.jpg
481a84a36747b18bae5b4e0129c2948f
3565bfe6457dfd9090af33a934f3781663e893db
109348 F20101113_AAAXII dempsey_c_Page_20.jp2
e2c2af35a58e13e26c7deff620adefc7
f49a44eb91406bbb7dc2dee753275624ad8e8c0d
2088 F20101113_AAAXSC dempsey_c_Page_25.txt
d8b0d125ba7bcedc04c6c9aa817f4ba8
dba3159ae1087e1f68f183a459a3c04e42b5b1b2
77335 F20101113_AAAXNF dempsey_c_Page_38.jpg
b48e9f2053cc1bdcef4aba5a53870e90
ee7efd03bc00c81293399b3ffed258f13a5beeb4
F20101113_AAAXIJ dempsey_c_Page_17.tif
d647da65a77bb071214bc0071860fe89
e8895108be5d8cc38a34449d9c540749c983d933
2125 F20101113_AAAXSD dempsey_c_Page_27.txt
c0f818ee4aaf769f744c0c7e5c26f13d
8d34f7fe2bb099d1047a3b49d31a94cdc0df207c
71371 F20101113_AAAXNG dempsey_c_Page_39.jpg
6bf7b833c61fdc589c7d8b128173b67c
c89d47ff37d50bf083d9ad99734c7332f0631c63
12349 F20101113_AAAXIK dempsey_c_Page_14.QC.jpg
1535a645c2769588abae59bb85a36e33
b7892b4f268595b14d1e9893d12700e7d7768171
905 F20101113_AAAXSE dempsey_c_Page_28.txt
209a47b07007518ecf99624be9b905de
5501de0083e1381d5a46fa5035743246fdb332ca
40392 F20101113_AAAXNH dempsey_c_Page_40.jpg
e138dbcebd0bfa7d226207e5dbc08b0d
87d94057ae9f23a28de95c95ca6649a25a0755af
5305 F20101113_AAAXIL dempsey_c_Page_43.txt
e3dc23fe2e7a0e49bd63fa8fb051fbf8
727d8931f70181d7e14830eb885d902f31d02423
2074 F20101113_AAAXSF dempsey_c_Page_30.txt
977d5232430be51aa318b1cf5bd80aa6
abb5a931e7695e3f2e80a38c890cacbc6368ebc3
46094 F20101113_AAAXNI dempsey_c_Page_43.jpg
c234380aedfe9627e780f12dfbaf7cac
c9d79b3476624b840765ea09af6131178b7fbc38
7375 F20101113_AAAXIM dempsey_c_Page_26thm.jpg
a5c6c25362ee14af6e63cc604479ef55
a55a954fef39e68d0a9ba2378403c7783fc87b43
2135 F20101113_AAAXSG dempsey_c_Page_31.txt
e864fa8a848365dfe3f1ce6bb13fb389
c7e2e0a4e75160ebccb843776d5a0bf5f9d508a2
71477 F20101113_AAAXNJ dempsey_c_Page_44.jpg
00c25588c44a69955c9f5ae8fe143fc8
cc8a234453d5a0024cb05088fd17d8a98a283c78
3147 F20101113_AAAXIN dempsey_c_Page_02.QC.jpg
1f88279fcd99bfe4f0f06f58e64df45b
4111efc93c4f09ed58bf9ac13945160359b5c1ef
2047 F20101113_AAAXSH dempsey_c_Page_32.txt
fd7429714578e425fcd5dd910a72e901
66e72ad55d3c69b299e2343d77b320c7fc884e12
77228 F20101113_AAAXNK dempsey_c_Page_45.jpg
671694e04f285a8d4deb9f7eada4bd6f
c1b24c44aeac3ae962fbb08cfd0adbdc8f4c5d2f
5386 F20101113_AAAXIO dempsey_c_Page_23thm.jpg
c9e8144e7940de1d340cc9dd5acb3d44
648f31d5a4c210bface05c75e7f33959accb6afa
2101 F20101113_AAAXSI dempsey_c_Page_34.txt
2a8aa5811caeb429702a475da46defc6
8ce48c1ad0e7a432e3095f74ce44451737626223
75352 F20101113_AAAXNL dempsey_c_Page_46.jpg
b1316a56e9222092a78c959a0b0d82d5
1d5390a3b9812ab68b0b44805c5f19604b1f31fd
12594 F20101113_AAAXIP dempsey_c_Page_29.QC.jpg
04762feeb06ee6e46ccd18c486862f42
b20a55bbb772804c57e43497873a053d7d793ad5
2172 F20101113_AAAXSJ dempsey_c_Page_35.txt
03a6463e4676e693af1c57037cd79533
ce67c1bf04fb3b6e31651f8baadb9708b8953c35
76157 F20101113_AAAXNM dempsey_c_Page_47.jpg
bcaede6fb658369aed32ee5c2b24a5d0
970443643c9adba0fafc824490c76ae9e10cbefc
55294 F20101113_AAAXIQ dempsey_c_Page_36.pro
e404b910e0f2af10c90730d488cad49a
a9309052b79790274ff7d231dd914478e26ddb36
2096 F20101113_AAAXSK dempsey_c_Page_37.txt
392ba526d1d81c4c2b7a736529f70ceb
1e0f3a1cf2d4f97bd2aa9522efa3514b722d2b32
75018 F20101113_AAAXNN dempsey_c_Page_48.jpg
5ac215a1495f01934cd7ec63f4dc9f5c
cd37693d067e365051838405c5b415d8dfaf0184
1473 F20101113_AAAXIR dempsey_c_Page_50.txt
7971e735e0338ec72ee5504f81f7b4a8
172f4ec75c786b0874805e1a7b451760a0495296
2113 F20101113_AAAXSL dempsey_c_Page_38.txt
4f0f2ee46153e68feb63866b50351f0a
661956b28113b6ec7623de867b6822898655fcde
54851 F20101113_AAAXNO dempsey_c_Page_50.jpg
af3050acdc381fcdf4828d516cbd8358
28e09990441bc4c55c7fbcacbfed936eea26d268
24952 F20101113_AAAXIS dempsey_c_Page_21.QC.jpg
1391396e2015b8858c06b2216a9ab97a
8a8cd59b1bf755826ae9d5aaa8ecb434b5954709
1926 F20101113_AAAXSM dempsey_c_Page_39.txt
83d6a6648ac2ae4ce2ac91265c6f1f27
605c48e2f44ee1627eaab82af31a49b41b6f82a4
88328 F20101113_AAAXNP dempsey_c_Page_51.jpg
c15f589a73273b4f1e4c7fb5de290ea8
0e66aa0878b8facc9351e46eb9bf2bac7cf1fe36
2532 F20101113_AAAXIT dempsey_c_Page_51.txt
14637c52ee110dab74b422b7be6f73f1
12e970b3c04d5cbec7048549f21ef4115d5a1063
672 F20101113_AAAXSN dempsey_c_Page_41.txt
46fe7e0ccff3cbe1a9a18f3885a2a408
8a97e9cd6deffca3b04db5b5c2f8ab124206dd11
91499 F20101113_AAAXNQ dempsey_c_Page_52.jpg
19fd0d865dee3d4632f52847e142d96c
c5317a92fc16ce679e080e64bc1f862e4931cd89
58707 F20101113_AAAXIU dempsey_c_Page_23.jpg
8664aedd446adbcfb8c6559732b44f2d
b0f91287257bfe2fe7d14bfaca78c0f9c511fe85
763 F20101113_AAAXSO dempsey_c_Page_42.txt
7012f0e1eec4665dae324e6607387226
277af60f85e06883a32e045edfbaac0c51a7d50c
37656 F20101113_AAAXNR dempsey_c_Page_55.jpg
1fdd1d2f869e93fcb32854f0c5f7fdfe
6ca0fa9b4a66aafceb279e4ab0a9a4b0229d39ca
F20101113_AAAXIV dempsey_c_Page_24.tif
f03e8aa06120dec9efe4e43a0baf1571
a132a99dbdf1e03d5cd9fc43fa0707c01d5bb05b
2181 F20101113_AAAXSP dempsey_c_Page_45.txt
4eba9967d0ad5f04a4d5025a2059a2cb
de392d65bf20a183e37974fdbaaac4e369881454
26108 F20101113_AAAXNS dempsey_c_Page_01.jp2
24060cefabce777342aa2ae185e52d0f
e770f222ba96a1a464a841c989c97e142618ecdf
F20101113_AAAXIW dempsey_c_Page_13.tif
d0ecd4d1aa35125ecd45bc923694ab99
6cbd5f5105b498fc19c6e60da79a64c0796f8b60
2100 F20101113_AAAXSQ dempsey_c_Page_47.txt
5a506d0cbaead454c5ed81c03791878d
6ac1258b2a072086abc05fcb0d25518388a2f5ab
5940 F20101113_AAAXNT dempsey_c_Page_02.jp2
3e18651c7754c0c9e6d20e653c9fecf6
85fa743fccb385e8eb0ba75c9e404a53f1b5ba55
2091 F20101113_AAAXSR dempsey_c_Page_48.txt
3e79e7d441618f2aa5cd1898602f80dc
530946e95fcf33e3b5e1980440dd61b196bee5c3
6430 F20101113_AAAXNU dempsey_c_Page_03.jp2
e2d8adfa1133761e16887c600f6dbfac
9ffb8abf1325604fbd14feff1bf11a0fe8516b2c
2186 F20101113_AAAXSS dempsey_c_Page_49.txt
1690d37176a0a22bcf18032e19d09c3f
7ec1121a753b5f1f5d916019861d226b43444d72
45737 F20101113_AAAXNV dempsey_c_Page_04.jp2
40715e2e178538240c25c9a04350ed4a
e24e5b9c73819605d5f5bbd5342da52fbe4e685c
F20101113_AAAXIX dempsey_c_Page_42.tif
6e95c79f13cab7894e108f681403a278
cc4a38e42aa2e889d1925eb0499aaa3e79ff7d0f
2697 F20101113_AAAXST dempsey_c_Page_52.txt
1c4638d364faba31e217f2b92495176a
e268767fa399235da32a2315b29beff82d62f348
350462 F20101113_AAAXNW dempsey_c_Page_06.jp2
a809c56179c8a7fc572e9eb9541e0727
f80090ecac2691d8c407942aec300b3d2aae8418
531 F20101113_AAAXIY dempsey_c_Page_29.txt
d81991835086eea925828bdd3ac8c986
9402b9e53c2e402bf74102117bb71580b6544d7b
2643 F20101113_AAAXSU dempsey_c_Page_53.txt
c89e0cdce0415c06fb3c520e18793e5f
03a39d1262bbb71c25b02b28e54a4d2a4b94f3b2
105428 F20101113_AAAXNX dempsey_c_Page_07.jp2
b1661b5ea0d36c31f83c2982155a7ed4
a99ce431fcfffb2e3e038eddaf0464dd452cf42d
2168 F20101113_AAAXIZ dempsey_c_Page_36.txt
8b447bea12d051591da87ac05bfe5452
e981022a4ccf0ce4e7582ed1af9c06e7b1aac015
428 F20101113_AAAXSV dempsey_c_Page_54.txt
dc252a1a6d6a2dacc1f850793d0213f3
17eadec93c43bbf458d25837ba19bf6859ef88e2
105830 F20101113_AAAXNY dempsey_c_Page_09.jp2
2389e637d884df74f1190b47d9dfbbbc
004e03719be465d70a8f724343c02caa1aaf9186
899 F20101113_AAAXSW dempsey_c_Page_55.txt
d6b3e8a2ea71bd10aa4815dd2ab8a0c7
898613bf85c8e747a8cbb628428afe582e6d590e
790 F20101113_AAAXLA dempsey_c_Page_08.txt
9bf2d6c88ee0b461caeb48f28c86d684
69861836cdb363e422cec7409c12a3f3f3af4b21
110787 F20101113_AAAXNZ dempsey_c_Page_11.jp2
7ca8b4dac3f59997a5a44a1f8b503dcd
14f5b0766f79f16aad97fe1dddc44d6c5512d658
278319 F20101113_AAAXSX dempsey_c.pdf
b058616d57fe8a1c9a71a3e9b8f7d18e
e0fc8f3c3be116300756a0646f5f431073b06b5c
77166 F20101113_AAAXLB dempsey_c_Page_05.jpg
bfe12bfabd54fa4d0fcfd4fa522c3343
51bd9509a8c1f811a672cabff2bba22ecc26c82c
24298 F20101113_AAAXSY dempsey_c_Page_48.QC.jpg
d0ddd8f48c99e096e8c47724cc8969c2
eebd59a2d710162d733ab6f56eb6b112fe7a021c
74913 F20101113_AAAXLC dempsey_c_Page_32.jpg
8ce0962d7c2a3eeacd5d6bed99826581
564b0dfac6bf6ac240cf91937f24020ac65a8836
5962 F20101113_AAAXSZ dempsey_c_Page_06.QC.jpg
50d3397b4a91aae2b3382116daff7321
c5c1359f3176dd3158d889f4a5f9336f95742507
77320 F20101113_AAAXLD dempsey_c_Page_31.jpg
9c3d6706304cc2a7093c046f591de3d3
6fc3768ec6e0b52b95a10af0702e02378eca8ad6
F20101113_AAAXQA dempsey_c_Page_47.tif
315fef9a5dae097db4df8e7e488ec308
776da2d83b2e363c98e90640d572e9df2d836182
9967 F20101113_AAAXLE dempsey_c_Page_08.QC.jpg
32a15a5c00aff0d148e87c970432f103
7ab47e491cf4b524573df14a10688ab7d06d334e
F20101113_AAAXQB dempsey_c_Page_48.tif
51d6b33765b6465a8e4dd11d065dcb1e
88b8ff93e732222f37274211f4b12d5128da64ef
1635 F20101113_AAAXLF dempsey_c_Page_07thm.jpg
01ea01dcdf0da12a42a83e95906d68be
c1409745df13bbdbf6faaf1ee7f322afa6478307
21189 F20101113_AAAXVA dempsey_c_Page_05.QC.jpg
b6ffa1b6a8ed3b55a1c7f43d9684d0a5
a74058ed66f084ee78bb6390168e92ce933e69a8
F20101113_AAAXQC dempsey_c_Page_49.tif
c49f31c9551e18706db081d3edb41894
886598bf69724ee190f9dff8ff8e44d75dd9976d
20811 F20101113_AAAXLG dempsey_c_Page_54.jpg
83d27eab1a011754d1a0fa3a025221cf
89311c6ecd59ae7d59a2b0a56f94aacbc0ea027f
6468 F20101113_AAAXVB dempsey_c_Page_09thm.jpg
cd4396b1c84baa03b416894df5cbf01e
34f3e71a3a5c4df17df127d0240a719bf7909f38
F20101113_AAAXQD dempsey_c_Page_50.tif
abc5bfb74d61f8ce3009a842a890843f
57368004383db084c3b36b6616f21d5d835735d0
118957 F20101113_AAAXLH dempsey_c_Page_36.jp2
96de51fff8aed4959094568b95826196
438bc91e47c4867f2cf206e0b2c101c2f8745863
25924 F20101113_AAAXVC dempsey_c_Page_12.QC.jpg
19a6a3ecb03bc08aa466c6d3b83cd997
e1e1bd12d102b6628a33e97ed1d303007eb9866e
F20101113_AAAXQE dempsey_c_Page_52.tif
a208971eee0389c4d465afa7ba4cc823
ea24705ada07c795421b1c0f7b1b3f201ea5501e
21687 F20101113_AAAXLI dempsey_c_Page_55.pro
a92993f7d37e99a6fd068241e8e0c742
a82a31cbbe45991b08e1bafc8e6abb1f082bba37
23896 F20101113_AAAXVD dempsey_c_Page_15.QC.jpg
38675ad442d7658b3a1948868d42d78e
b78321941641a39e36449ad189b739d8499db279
F20101113_AAAXQF dempsey_c_Page_53.tif
37bdea177edcf56fd7a54d18de064742
1ac6e8110d249881113ac352a91d7b1c070a3085
14198 F20101113_AAAXLJ dempsey_c_Page_10.jp2
66a7b4e177209747577a25dba8b240ce
9008767018891f7188bb37046f2b44b775540300
F20101113_AAAXQG dempsey_c_Page_54.tif
236a70514b1d3023eab2c657cb1e00df
be923c8ab42fe1d752c8e85b398fe2c815b1cd6c
53070 F20101113_AAAXLK dempsey_c_Page_47.pro
282503e9cd03d018ec6ee6642a8eb217
cc999bfdb55e7b6ff79efce7c74e3ebb19fe7476
23187 F20101113_AAAXVE dempsey_c_Page_17.QC.jpg
2ef3ba156062568a3cc5b220421f8b02
42c9acbde1107d0e0aca31263fad53731f783322
8513 F20101113_AAAXQH dempsey_c_Page_01.pro
24abdab0f7738e7934c7214217907571
5137da1e8de646ba7f5fe74e1ba68a1ba7f414b0
3620 F20101113_AAAXLL dempsey_c_Page_07.pro
8e819b51ca9242cd842dfb67b9f5d3fd
3798fb6f272a53892b5a5811b6e3721e972e2221
24586 F20101113_AAAXVF dempsey_c_Page_18.QC.jpg
cd10f97f941e1ed11ac79357330ba0f5
804d5510a7112219c4d3b9d888fe211bb8b064cf
1402 F20101113_AAAXQI dempsey_c_Page_03.pro
9fe3b04410f4f064cc15a267f54d8055
4c1342951faecf3457da981b020d70c119015da1
F20101113_AAAXLM dempsey_c_Page_36.tif
5a44007ff1addbeaf27de957e80db4d4
80e8af881c2cf54cdd6ecbcf1e8ce242b14174d6
23474 F20101113_AAAXVG dempsey_c_Page_20.QC.jpg
f2b098d7812ea24a2d6392f4f6087b76
72ee10de5aa63245566c992a9a1e6a43b6663085
19029 F20101113_AAAXQJ dempsey_c_Page_08.pro
962142e2c973af54bf32ecffdb8a6404
f1bc6671d143e8169c7e9675904f5826b1a5c4a0
6973 F20101113_AAAXLN dempsey_c_Page_31thm.jpg
e2b3f7003f97a84aa6616914dd574bf8
6717f65145f16f51795fcecbcc0db7becc98ce8d
6341 F20101113_AAAXVH dempsey_c_Page_22thm.jpg
5e8b35f946c007568c1bb4f623df9211
0a8e50bb62e291bd401c259998d7787b556f7551
52510 F20101113_AAAXGQ dempsey_c_Page_26.pro
752e122442d610da34c77a5b62c88a3a
fb4f58faa0c100ca4318e2e7ff6c3d07202936e8
48452 F20101113_AAAXQK dempsey_c_Page_09.pro
9f539f9430f88475ca54e452c86b6d71
0017517699fac384355bf0ad0c91caf0864b2911
2144 F20101113_AAAXLO dempsey_c_Page_18.txt
4a51f698c8b754f687ec3f13735ec225
da789678aa6f377d69a82a66317950c1afb45981
27083 F20101113_AAAXVI dempsey_c_Page_26.QC.jpg
a55215f911d60b7cac5f15f744d9867e
587450f8c97cb65496212019aae089c215c883c2
6930 F20101113_AAAXGR dempsey_c_Page_53thm.jpg
7f079f591b25bf5eabf46504471e6a8c
041564ec2acdf2e7907c02ad28729776ff682c13
5005 F20101113_AAAXQL dempsey_c_Page_10.pro
7f38c794f6ce2fec34f915819c16f9d5
2f54e94128ce6a4e98633377a1b5526ec4f50e3e
6473 F20101113_AAAXLP dempsey_c_Page_17thm.jpg
f681b4dc6de518a53353cb14991ff7dc
6434240c91703e0f230b1c86a8fb2366bdb98971
12282 F20101113_AAAXVJ dempsey_c_Page_28.QC.jpg
c94699b3f7bd124909ca5e0baf19b5aa
ed7c4d7c3285565655215d40b3fde3f295d3fd8b
117202 F20101113_AAAXGS dempsey_c_Page_27.jp2
67914748e8e8fafaf42288a6cafaa4b5
c146b443425b0fca02f6ee0e62945d69e5a874ba
50842 F20101113_AAAXQM dempsey_c_Page_11.pro
4e69e027842df2a44d6e25615b5fb1b2
0e308f6b5e7b1c000c83e06bba7a9938b3d890d4
25744 F20101113_AAAXLQ dempsey_c_Page_36.QC.jpg
a4b1b914065232eb76c94feb114dc529
be54d9576324ce48a0068263e56679fc2b574934
3741 F20101113_AAAXVK dempsey_c_Page_28thm.jpg
552806bf3e3a068b127d5e81790622a6
c5ad4e1ad9725b472e58627396c33199b00deafa
24557 F20101113_AAAXGT dempsey_c_Page_11.QC.jpg
2ec62d358b5b39f151dc92f0390ab9dd
b807856893d8cd48d5653de64ffe722ef1f79137
55826 F20101113_AAAXQN dempsey_c_Page_13.pro
243bafa308726f5df3a27f7184ec7a1f
83c01cb06334d104e4de6bc14b495160a56e950f
2134 F20101113_AAAXLR dempsey_c_Page_12.txt
6e1bf470f230643c1cd3901c313c9c93
58869a22ff24d36c3af80dd859345b7fbf6e366f
22973 F20101113_AAAXVL dempsey_c_Page_44.QC.jpg
6f367fee90a4874bb45ab3038f6c485d
22dcafa9714e2f06bd229fb7d559c662c1dc4455
72870 F20101113_AAAXGU dempsey_c_Page_24.jpg
ddea020bc8c7757c15cfdf6d055d8657
4c39e78710f5e82b28c96c4464707b024206dcb8
22259 F20101113_AAAXQO dempsey_c_Page_14.pro
f3a2d867326b78a63710e4c95ade569f
a852d79923a99d92150eece2de98caf076dfa5eb
51091 F20101113_AAAXLS dempsey_c_Page_55.jp2
d1c1164d16d5273eeee88d3b95e7a236
fb62e973fff11289611c6abee9c51e52eb80b31a
6879 F20101113_AAAXVM dempsey_c_Page_46thm.jpg
41810088b26b01ee82abf980f99bb209
95feaf989e570c6352cf477b583fad82359e5a94
48948 F20101113_AAAXQP dempsey_c_Page_15.pro
0bb81be80ad496858f3f6ef8c31b04b8
6d652f438f9b7640030e5d439c8c0d55185225e2
20539 F20101113_AAAXLT dempsey_c_Page_06.jpg
0005a412a256b870d8242acb35780e2b
045204aef30cc153a069ebd37847c1cc793d6aa1
7010 F20101113_AAAXVN dempsey_c_Page_51thm.jpg
d4c2ac1edfe8a820c23ee2a9a07e4116
1d3010003e6cc1be8d762157cd14634dff549caf
19486 F20101113_AAAXGV dempsey_c_Page_23.QC.jpg
6d36765232faae4c48b3578dc7988dc9
a03da060d98ff4e08c69cc622a187f95624b73e4
55422 F20101113_AAAXQQ dempsey_c_Page_16.pro
f97cd05cbcda5b887f65f6092eb1f2ec
041ab89fa08db2dbaad1172318ac53f2d4b6388c
7227 F20101113_AAAXVO dempsey_c_Page_52thm.jpg
35035a358e9f277f5c2cfcde0f65091a
5e5facbc18bf3bd54183ab9d81f9356a2b90b0de
F20101113_AAAXGW dempsey_c_Page_12.tif
7ab40e84fb72600223503bafc6582778
9093773579ceddf281172a490e8bfd19522e3116
54376 F20101113_AAAXQR dempsey_c_Page_19.pro
8683906066159fce73791c95f0c32a0b
03cbb23c5adff2e80e263de1f2df5ab80775e18a
109947 F20101113_AAAXLU dempsey_c_Page_33.jp2
3d27a537576d7c388f75f527a6819d98
0e947eafe1ceb598e7a142285f7c6531adea9697
6466 F20101113_AAAXVP dempsey_c_Page_54.QC.jpg
bd8302c036c8470d5b730cb7b7232675
ee504c002dfadf123c412435fedf1b71d0617a97
F20101113_AAAXGX dempsey_c_Page_15.tif
c7f04d37b57d908849151f0fae086ba4
f29f41ce3414ac48c0a17f189672f08ad6cd5168
50258 F20101113_AAAXQS dempsey_c_Page_20.pro
f324e04c60d1ab8a7be7c98a36414774
6d8e8f1e5387bb79a8040e8dff0b4bf8c164ecfd
6901 F20101113_AAAXLV dempsey_c_Page_47thm.jpg
07da2b7af3423fdfd540229f9246b4b7
a16ae74d5487dba2a03b2843881c2c716e36d002
1986 F20101113_AAAXGY dempsey_c_Page_33.txt
229fd69155eeab7238f9e275e136422b
60e84a8ed86a68cfc279c6a7d7ebe1c1441133e7
52809 F20101113_AAAXQT dempsey_c_Page_21.pro
2dac5f7374a59f479c2d9466b7d2db2f
13c9c77e982726e1a0fb8379a6db8f2a862d6488
F20101113_AAAXLW dempsey_c_Page_28.tif
c368ca189142e168fcad2083693e3a16
463386a537d95081e136244c9ae844f64858c30d
2099 F20101113_AAAXGZ dempsey_c_Page_26.txt
d46ec311ab97a8afc09f8f35ad6dcdf1
54c6fd2367c22ce43d7b93786cab06243b2571a9
49075 F20101113_AAAXQU dempsey_c_Page_22.pro
dd0f107791e8dd4edb78e894498e157c
2ad53cb13f0272c8cd23a112105db58437b4e60b
3247 F20101113_AAAXLX dempsey_c_Page_03.QC.jpg
2937f750866b454f3276521c664844d5
276f6175945e56b9295d0b0101f4eaa9e7bfc0f1
49357 F20101113_AAAXQV dempsey_c_Page_24.pro
e00512f9c788a559d4f1328c7b898d8c
001ebc19bcd6628432e767f532e43f3338f95f6f
F20101113_AAAXLY dempsey_c_Page_51.tif
b26747e8c5bec80bcb4c5e2e6ef197d0
2a1dd4c3945fce2d48d78a423eee4ce3a72c3daf
53145 F20101113_AAAXQW dempsey_c_Page_25.pro
38eddf2c6b2f60c9c6f3425b093553cf
f83136d966d978e3486b5b4e1d0c08a8b99200f9
149 F20101113_AAAXJA dempsey_c_Page_07.txt
1af8c0813aa7a3b05f0b5f9f7d47f653
48f3dfd5c7daf360989d008a361d5ecadb113276
76991 F20101113_AAAXLZ dempsey_c_Page_34.jpg
1aef9d40f77e4b32962c86342ade46cb
08f5092e4ed603576744450746486e7f7d99214b
53866 F20101113_AAAXQX dempsey_c_Page_27.pro
c4bb3ba296e144f2a786a97a936108de
49c0c3146141a80e9e4c0f71f640310683418225
24337 F20101113_AAAXJB dempsey_c_Page_46.QC.jpg
bc2ad74173fce5b846c195d85096b751
e0ebf3dcbd03446a549b0ddf8dd08ea000529458
22624 F20101113_AAAXQY dempsey_c_Page_28.pro
626164c36bbd69d17bda60147eaec162
ba0f2622e77031ee022176d03898cdf135e1391c
12607 F20101113_AAAXJC dempsey_c_Page_07.jpg
d3a14b56edd51ea2eadb071a5ed99d00
29be3514ada9c2375a41c0917cf87a3c382d5d8b
118480 F20101113_AAAXOA dempsey_c_Page_12.jp2
e3682e1023543658625294c9830e2f30
59ca4963f313fa89c64b0f064c72b78f7c552f5f
8126 F20101113_AAAXQZ dempsey_c_Page_29.pro
b210e7f79bb86b8de89d50462513440a
5cf3eaed6f4bc315d40e9275f2fdc14d7e36a5c8
78055 F20101113_AAAXJD dempsey_c_Page_49.jpg
745fdcda5b5b3e12e02c54fd26c7cfea
f7e20530124feb1e7c1047d69c918937120e07e2
120518 F20101113_AAAXOB dempsey_c_Page_13.jp2
1b3d3cb3d996b619468a902dd24ce3b4
7ff0e6cb60166f358b817ec0dc3a13e6a03f5d12
4635 F20101113_AAAXJE dempsey_c_Page_42thm.jpg
85918bd74630ad0afc55bf7c9dfb6548
7dbb4262569dbd98a12fdf56efc9731c9d7b6fd5
107649 F20101113_AAAXOC dempsey_c_Page_15.jp2
ea8f6c9ac5ddba209682778a776b2908
29ddc6d8ec4b0703fe0502ae8db19ce04225d511
43126 F20101113_AAAXJF dempsey_c_Page_42.jpg
bfcf6014c8973b363926ade35d553b69
50fbf777a9ead4dc2c2c18dd8c3c80920f478d3a
24864 F20101113_AAAXTA dempsey_c_Page_53.QC.jpg
a126e6e693151f6bdff7a7a5a577f89b
b0d085654afe202ac2aed4c4539a360b94c0e3be
120698 F20101113_AAAXOD dempsey_c_Page_16.jp2
d50803cc896239bdbb586bacf891e325
235a1d181e419eb5a2deafe796aab8c86d02ef59
493 F20101113_AAAXJG dempsey_c_Page_01.txt
b1349a01670deb297ce5bb02f5803e7a
18e1fe5c7c6e8c75575008b70066c7ff52ae720e
3732 F20101113_AAAXTB dempsey_c_Page_14thm.jpg
59b76cd65c989b069d48cf4cefee29ad
633c76194d985bc23fa24667b029b8bd5914229a
109340 F20101113_AAAXOE dempsey_c_Page_17.jp2
0ac18219db137bc7464d528d83841723
4e0f825dd2eb1387a38bd92fa383f6f0f6943157
117147 F20101113_AAAXJH dempsey_c_Page_38.jp2
37a2514cd6970a9fb583eddf83baa412
37c6a4949191cd28cbadfc773ca1df58af7bd582
113455 F20101113_AAAXOF dempsey_c_Page_18.jp2
a5a02b7f2cea29f4cd13adf96baf6b49
f0d46c920f8a816aa05b4441fd8271e87a3229d7
114753 F20101113_AAAXJI dempsey_c_Page_21.jp2
76c5d83db70b4c04e8eb909182e34e7b
936d3bcf859c6e3b7a67ce141afd69576b00dcde
7057 F20101113_AAAXTC dempsey_c_Page_36thm.jpg
2a67facc84ee128f3db7fb95c73c55c6
d3c37f029e920b6d7eabb51f94de3779cb23844d
118924 F20101113_AAAXOG dempsey_c_Page_19.jp2
51b8149df700898041c0506aec9d2859
ec009c0ea0fdd2f9dd3968901f2279c13fd10e5c
107547 F20101113_AAAXJJ dempsey_c_Page_39.jp2
a1a6dc9e4fd6b0658d65c0922cf74841
9b5f1355dc5ef5a0af62f305f7f6bffb306b7f9d
12189 F20101113_AAAXTD dempsey_c_Page_43.QC.jpg
05dba51ddc5768abcff575963b42d746
89ec742ef3cf21de05642bb5d6fb47efbefe3f75
87641 F20101113_AAAXOH dempsey_c_Page_23.jp2
0c69d327a70c730a0b0a18a56668b340
86549f60682ed819fb2cc045295dfc5379405fe2
6943 F20101113_AAAXJK dempsey_c_Page_49thm.jpg
3a91ed1a1d97b32fae6ca8783429567b
625331553f1a1d3c8a539df7373dd0648cbed687
7100 F20101113_AAAXTE dempsey_c_Page_13thm.jpg
179113c5c9fb815e5026505a5b08845f
d906564bdd2173ab112986823f6407671a0852bc
116528 F20101113_AAAXOI dempsey_c_Page_25.jp2
86ad8159de8285a6067a3be27655fcce
52d0215455d724e72c4432e5c2cefda6a3be9211
15089 F20101113_AAAXJL dempsey_c_Page_10.jpg
d1977ac1edf65f0ea1990a5b7ca1398a
6776e812af76d260a6bcf8d0059124b628a37f37
23586 F20101113_AAAXTF dempsey_c_Page_39.QC.jpg
7f0ee4ea24ef6f759e1785c7ae3f7af6
be130434e0954e4fc9f61c5e1201682e70fed05b
1051958 F20101113_AAAXOJ dempsey_c_Page_26.jp2
80232f3eb688feb28990e6bc419dd89b
7f554887aa8f77fb155c70a74e50a2ffdcf3c10d
53689 F20101113_AAAXJM dempsey_c_Page_12.pro
e69ae62cb58edc2c62cb8482d36aaabb
c18a80086bd55d6f0ee995a1da6ce7e214bc1e85
6852 F20101113_AAAXTG dempsey_c_Page_38thm.jpg
e2470c529b95a113d1e7da91e04f0c78
0f0daee5500ca8e7aab25b61acabad84dcacf029
52807 F20101113_AAAXOK dempsey_c_Page_28.jp2
a1278fbda81fa243f8e029a82652829b
cd0e34066710e33036c21b981a3f9abfec48ff21
F20101113_AAAXJN dempsey_c_Page_05.tif
fa4d5781415587b00a4fa269bd429adf
17125f9f4bd851b05b9094c23f1601dc9167942e
6837 F20101113_AAAXTH dempsey_c_Page_34thm.jpg
aaf6df6ae3b6527171c3e1ea440b46a6
6720a4047349dac36abe106c9dc74f80bbd8bbf1
118530 F20101113_AAAXOL dempsey_c_Page_31.jp2
0e51dd5ebfdb933968c899e4bd41a35a
04e52b8954ad89d30171bad354932aecdd851118
12084 F20101113_AAAXJO dempsey_c_Page_55.QC.jpg
5bf5212908f3d31ca186804d912a9f64
35cec8a4bba12313e085707acbb4a3c9996b6505
5289 F20101113_AAAXTI dempsey_c_Page_50thm.jpg
b43ba68e1fdfbc033dbb7bbeb4da43b2
2a688faccf39280309be14190c99ffda704f48da
117273 F20101113_AAAXOM dempsey_c_Page_32.jp2
ae59d2b290bb369d0b9e8c1eb6333a2e
2ec6e7203639813f43a13e74cc93a7bed79fc2bc
112720 F20101113_AAAXJP dempsey_c_Page_30.jp2
717dcf769fc10295494488ee4398f303
049f5ed9658a9f053bd180c355e6e915e8b279dd
25708 F20101113_AAAXTJ dempsey_c_Page_51.QC.jpg
bbec4c6376e4dc53b71e9495e1494c08
8971cd9128b7e7700cc210aa0a70df9ddf7b46ec
114404 F20101113_AAAXON dempsey_c_Page_34.jp2
07318ce1fe91ac9c188fd2ea6518d6e0
0a656474b726181f73491e1662b5e7a97368bb9c
F20101113_AAAXJQ dempsey_c_Page_02.tif
9573a265a2c87323a303f990ae95023b
bc4e6a4682c298d0afa001abffbc3cc8a3862655
3266 F20101113_AAAXTK dempsey_c_Page_43thm.jpg
b09b170b7f976cfc5d795d4061aac247
ecd01f1afd3b55c905d5007eab18a922efacf25f
120025 F20101113_AAAXOO dempsey_c_Page_35.jp2
249cedcc25ffbdc271ecff8132da7919
3e0ff5ca233cfe2f0a1e3721b04df5d79f7fcf55
6513 F20101113_AAAXJR dempsey_c_Page_24thm.jpg
941bcbddeda2736d106f1bf005184224
1851c3150c512daa9dfe0891ff3c793f468a3b80
22584 F20101113_AAAXTL dempsey_c_Page_09.QC.jpg
a41a1d55e6a429a49c0dd4c80323bc12
ef0202063a2913c03fde80f4693341b65ba0fb47
114943 F20101113_AAAXOP dempsey_c_Page_37.jp2
a17c094940c2ce5f3a9dc20514f11d48
5ebbb85d36751779bbeff57a24f2867a3b1f433e
26330 F20101113_AAAXJS dempsey_c_Page_13.QC.jpg
3f3e4a5ac60019929706c54f617bd521
1e36f175516e62777c4f18a78d078e452767fd85
22707 F20101113_AAAXTM dempsey_c_Page_22.QC.jpg
8fdd037a0e5b3859af6ac8c0d84523bd
2d6bfaa3b595a2664d782cd029668394eb5bf396
6823 F20101113_AAAXJT dempsey_c_Page_12thm.jpg
505500044e579a852f4db0d1ec333db4
8c91ae917716e3dfb1254a300744bc980eed9cb2
24254 F20101113_AAAXTN dempsey_c_Page_32.QC.jpg
46e5bf6f13634817970a3182e59c3151
b6d46a0ed48cde26b85fed3191981e8fd1b439e7
408547 F20101113_AAAXOQ dempsey_c_Page_41.jp2
4237b20f65f7bf799f861a210270063b
228fe11609038e402ef8558126bf44af56a527af
1038 F20101113_AAAXJU dempsey_c_Page_02.pro
6bd69157b513db026380301b9d9b82a1
f55a654d595e9d69d11908b462e2c8e8a7df3950
6567 F20101113_AAAXTO dempsey_c_Page_20thm.jpg
aa4c7fbd6e201a23d122238a06759e4c
aee19f8d61439327995e43b0884010506fa0414c
128620 F20101113_AAAXOR dempsey_c_Page_43.jp2
6a4b3a92bbf2e7f49ee974ecba3881b8
e536ec7cd739f746387f23a34277890532755e4c
134490 F20101113_AAAXJV dempsey_c_Page_53.jp2
0c6bbcf0ee4abc2eddf17f91ac667038
933694cd70ffae0c5c77d5c775fca959f566136d
6812 F20101113_AAAXTP dempsey_c_Page_19thm.jpg
34b2c224c856f374d4d710bc29278aef
e01fe5eb15c0be9b463002d6eda880100d80ce76
107947 F20101113_AAAXOS dempsey_c_Page_44.jp2
68349989747aacb6304e85c24e06de74
938929f9bd00913698e66914ed3ed306542f5bc0
25580 F20101113_AAAXJW dempsey_c_Page_31.QC.jpg
9cee508e60818754c8e2ecee132063dc
019cc2e36fe8b80ba34dc301e0d1fc8d87b6c8d4
6833 F20101113_AAAXTQ dempsey_c_Page_25thm.jpg
0acea6c716ee29b9dcb8fb4d7ded765d
af488cbf929ad80b6d5d0aab6e1b63455f7a2198
117372 F20101113_AAAXOT dempsey_c_Page_45.jp2
f8019cdecbf2676f451bdda654c9af7e
44827750cded9d2236299e74762d3f3718cbb368
F20101113_AAAXJX dempsey_c_Page_35.tif
f8a37481edd66c9cf2adad027dd47113
49f0fb6ef530cfa6ab75a78cb6f6fa61ed44917e
26215 F20101113_AAAXTR dempsey_c_Page_35.QC.jpg
c30f9ee0017b4cdf369ccaf246275219
eee01e3bd7acd51f9c6647971cd732a8f4e57f33
114160 F20101113_AAAXOU dempsey_c_Page_46.jp2
86498e1f613a2ec8444ec53f1b06fd23
481094f753a3d2c383f294293804841952f7bb1d
11056 F20101113_AAAXTS dempsey_c_Page_04.QC.jpg
45510b7a9e3023461106d738d94d5b51
94e92449e91d899c47622370072edb1506f48eab
113281 F20101113_AAAXOV dempsey_c_Page_48.jp2
41afc748cf0d4bda6e91979acd089048
50e580f29b3e20f7554a8991f56474a223f3912a
F20101113_AAAXHA dempsey_c_Page_29.tif
bc5f3aaa4c0deabba3e02a16bcb038f2
46ac5d1121ad0b510350a9674f31238664b2cda5
F20101113_AAAXJY dempsey_c_Page_41.tif
6f240d536f56d729b065e04df5ca60a6
790cf946999f0b2cd87056b41b5b8632ba3b0951
23785 F20101113_AAAXTT dempsey_c_Page_33.QC.jpg
5cc2b933f46d47d8c3bcd7117664b5ee
357ba5622a23f6c447f0109dc54135d0e18caf71
118189 F20101113_AAAXOW dempsey_c_Page_49.jp2
c7f1a79cf3f959ba3db21e53387b53f2
60b8567cb193763cd9e82863cfa85d3434de2399
2234 F20101113_AAAXHB dempsey_c_Page_13.txt
bf9c8af7373f9f4970b56d301ace4e94
2b61d34aece80042a776bccc5cdab1b273ea9b00
F20101113_AAAXJZ dempsey_c_Page_16.tif
b99b454efab6fc2d6e22163d8d46f99a
fc0386aac8e52f4fdbfdb9e44574b90c1cc86427
3374 F20101113_AAAXTU dempsey_c_Page_04thm.jpg
9566b69d24048008482ad7bf3767209b
7700b9b93ffc089ab72122a5fc09375f3d6290a8
82239 F20101113_AAAXOX dempsey_c_Page_50.jp2
a98827fd469d4f23bb4c6ede8a8190c9
a3a803d4e1a692b641685b565743c18c491eb9e2
24229 F20101113_AAAXHC dempsey_c_Page_24.QC.jpg
bed5019f18d5e255fdf672fae3789b3f
6a6cda319d24ab97a0921bdb875c9904646d8756
25315 F20101113_AAAXTV dempsey_c_Page_19.QC.jpg
ac7b56b442bceeea46d4f37ea5c737a1
6ba1dfc702f76fb06cc56d9f17785d31757f73c7
140214 F20101113_AAAXOY dempsey_c_Page_52.jp2
316b00b38ef8538ebbf50f9e7131a950
f1613756134eab1823df272c94ad820a47a34173
24890 F20101113_AAAXTW dempsey_c_Page_47.QC.jpg
cface790c24696ab733531ac09e7ce7d
fc9ce29b4e268c33bbfd4993437fe4eb34255cfd
24516 F20101113_AAAXMA dempsey_c_Page_30.QC.jpg
f9874d3a4376647aa468a43afe5b72fc
a3cad3a6c2e67b5c3526be5ac5911d77a34e9b80
25839 F20101113_AAAXOZ dempsey_c_Page_54.jp2
255054cefaad6030d5d4335131b7ee24
71d97ee507bd38e3d5bafac91a82b07f385e6074
18666 F20101113_AAAXHD dempsey_c_Page_50.QC.jpg
e2313022e1f837c59cd61af553c3ca7f
dd2e497c81006e815a8deb0aebe76db00fbd1823
2155 F20101113_AAAXTX dempsey_c_Page_06thm.jpg
9be887a866f95ec947e951aa702b8efc
175dea2d69a56125fe6f156bb38f7e034f44731b
94 F20101113_AAAXMB dempsey_c_Page_02.txt
437dce4a0881abaa2a12279638f85959
e74c8a6eaa77144b334aa5c0247ba7d63dade8a3
6374 F20101113_AAAXHE dempsey_c_Page_39thm.jpg
78c88a5965ed3b24ddbfbfac61440800
4c8bc739990918f45dd317d1feee367f144c33fa
6891 F20101113_AAAXTY dempsey_c_Page_27thm.jpg
a06b4b4d474633ebbc18b1f706576488
b0261b83acface3ed660a85cd75a38c94d8ed03e
478132 F20101113_AAAXMC dempsey_c_Page_42.jp2
3ea71d1f073375a2b7b16ff7ca45d413
363321efda0aca8c7e1873920a4ef563da85960e
5231 F20101113_AAAXHF dempsey_c_Page_10.QC.jpg
377985af558e4b14e894eabc8b954a78
775057fe176c927bdf62b50109cd2521b544d7f1
50632 F20101113_AAAXRA dempsey_c_Page_30.pro
cdf640a3f5f630a8a0def982cde8a145
15ec35234fc96b563f02214bddc0743260387b34
13654 F20101113_AAAXTZ dempsey_c_Page_42.QC.jpg
17db7847f8ba59c27635be45764fedd5
e890f3f884f7914eb2931e4ebfd5eac4f79ff519
71958 F20101113_AAAXMD dempsey_c_Page_09.jpg
b0d8b088db1fb05ba9add53b0ecab0ac
930f882f09551de10c16851c4a4c42b7448602d6
52769 F20101113_AAAXHG dempsey_c_Page_46.pro
4fe1c80fa9550f1aef0978456fd23de6
64609286c751ae5128befd84255e071bbe9a0ab6
54351 F20101113_AAAXRB dempsey_c_Page_31.pro
ff3e669bf8a2336e71421c479232fcc2
621f3f5370c3a8d86a8dc3a046ecfa4244ea53d1
66301 F20101113_AAAXME UFE0021351_00001.mets
0a4f3fc9a10af87e9f057d3166e4475e
3afeef832948ded92c5de34b81916360e67b75a7
827604 F20101113_AAAXHH dempsey_c_Page_08.jp2
548af0e908e3461db01eafc4b0d89a43
848c966ad3487df0ea0f3579b5d3a5aae70a471c
51916 F20101113_AAAXRC dempsey_c_Page_32.pro
9a02ed8bd13a1e59b7d245693f460e6f
e5ae6688e24db7bd886a4b0d6a042d7734e7c958
2017 F20101113_AAAXHI dempsey_c_Page_17.txt
aba33fe4533473568ed4aa9ff9d2efd5
324d325025b0229cd7050d3c91c0d47ce8a59ded
50363 F20101113_AAAXRD dempsey_c_Page_33.pro
b2a76bce4c1fac6188f2c9c19a692979
234521f731922db63b8bfefc15fc43fa34b6f97f
1054428 F20101113_AAAXHJ dempsey_c_Page_43.tif
47afc96c90bc53f9a75f97c64c0e5e12
c5892c0937549578a4ec246d161ea70d7423f059
53469 F20101113_AAAXRE dempsey_c_Page_34.pro
d7ff83ade45b64ee9b1a15f6bdec0dca
fbc2ae12f3d7c8548c4926a9a313a1554fe6d260
25528 F20101113_AAAXMH dempsey_c_Page_01.jpg
8b1945a4416908dc0ea51bc34333840a
83f4bc46a8b2448e52d7b80cb8a125547715823c
F20101113_AAAXHK dempsey_c_Page_34.tif
1c69dacc20a9cd143a7bc5ec23d5efa0
7b4dd35682bf0c4cb39ff34a461d97d6d16fe33d
55411 F20101113_AAAXRF dempsey_c_Page_35.pro
cc80afc66080127601db3fdd8cb33d27
e31d39f0721f9b25be748f2ba1fd087648667c3d
10655 F20101113_AAAXMI dempsey_c_Page_03.jpg
f917102e2a7f3835dde7ce6a9bb4def1
aaec93d90a4e7bec50b27465232752d5e3ee415d
10228 F20101113_AAAXHL dempsey_c_Page_02.jpg
f3569e4bbe03d0be39a26a42b8ea16ba
0f14d3c86db51e038559617addd7798b06d3e236
53499 F20101113_AAAXRG dempsey_c_Page_38.pro
e8003f02396050be274b758d9ed82fea
380ce94fa1204375940d65ebdd6b782bce5f117e
34299 F20101113_AAAXMJ dempsey_c_Page_04.jpg
6735fe61a47a0bb17db906812b4e3a7d
8e377cf2c7267cf9c7a3482b7f5a65703c1bdb0f
72071 F20101113_AAAXHM dempsey_c_Page_15.jpg
1f5a4038c85118a4a27bd259bc264956
464247906093942c305f81f1209a31adea71242b
48689 F20101113_AAAXRH dempsey_c_Page_39.pro
b537c454a2eecfbeebcc204f922b197b
6bd97526bbbc670a24992bbf6fe2edc99376398b
35368 F20101113_AAAXMK dempsey_c_Page_08.jpg
02f0afb674387839adc3e10eaf4c3052
5c74123ab5a55bcb0df512e14aa967c9192b8a92
51542 F20101113_AAAXHN dempsey_c_Page_14.jp2
657c043cf171056e7c69bf5788585091
c77e509f1a732debdb2dbb87607b1af326a3b46b
10792 F20101113_AAAXRI dempsey_c_Page_42.pro
4acb683d6b4d92a0421a6e5c1a218f54
0e202c71cce88f0d6a271daa97d59528cf507d0b
74840 F20101113_AAAXML dempsey_c_Page_11.jpg
cac7c3d388e228374333624d5c86e0d7
9a5f059c5bcfccda9592d3135ce9012120c927b5
62013 F20101113_AAAXHO dempsey_c_Page_51.pro
e7c5ad81f807107627e9459d63dc02ac
777a868f73d0eab247f79941ac1bea5715b75332
97796 F20101113_AAAXRJ dempsey_c_Page_43.pro
2f2197f11be1353ff095209111847749
5d5f212c4a1eb2b6f8df5792bf5e5cd4fdf2563c
77704 F20101113_AAAXMM dempsey_c_Page_12.jpg
aac5b05f8a0fdec2e9ea87a4000867c2
28f45490cbbf5a57320e306e50feb5b36e31d2f6
10153 F20101113_AAAXHP dempsey_c_Page_40.pro
f701be0b7c847786bd9cd6f68970ede9
dbe61582ad7e30b73ccb88171f6e3297d20f29c1
50259 F20101113_AAAXRK dempsey_c_Page_44.pro
7043f03ebac1133220ac8f5b41bf14ca
3b033a5d8918dc23559b7628b0712b3ebe3c6c9a
80617 F20101113_AAAXMN dempsey_c_Page_13.jpg
5160cd0da539ae563f495f5ac1266a34
7248360b3dc408c1557aeb9a7c99af02c033e009
3581 F20101113_AAAXHQ dempsey_c_Page_55thm.jpg
eeb52fc3f02cb36572454da80153757e
4571780499dae95b00137557b5070f0b466f381f
53057 F20101113_AAAXRL dempsey_c_Page_48.pro
2596e3d790d672fae30b1ec08534df2d
14c624b638b96d5dd34cbdb8b90761345e2c4b8b
37628 F20101113_AAAXMO dempsey_c_Page_14.jpg
299c29081bb47f79e7801a6cb307fbba
4c30aeae8bfe5752db113aba719d9de2d0459260
115944 F20101113_AAAXHR dempsey_c_Page_47.jp2
a2ff540bf6f885327b3db3fa0fe89da2
2fdacc2f1c9fc4be265b3cecd35baf0a8d3e3385
78471 F20101113_AAAXMP dempsey_c_Page_16.jpg
d9598702f29b4b3755ef951249bae181
8785c5b38f6330a2e8c09071a9f4a2f6030a0a24
4388 F20101113_AAAXHS dempsey_c_Page_40thm.jpg
6d63da9a8416e0a3b3d45065c9dcfed6
ed19273a6ec81fab2de92a473f7c7e36955edcf4
55351 F20101113_AAAXRM dempsey_c_Page_49.pro
ec72613aed561f1d65b7205a896eafa5
6d55eb96c7e66fe88059962319b79d2e26bf479d
71737 F20101113_AAAXMQ dempsey_c_Page_17.jpg
bbf5240eaabdca66e80b4045a62f5490
c023dd1d54e6314be4c87edef112149017356a8c
2079 F20101113_AAAXHT dempsey_c_Page_46.txt
672787baad0493e535dd310e350027d6
9416f403873e940d6d064d014233dbb3ada5283a
36958 F20101113_AAAXRN dempsey_c_Page_50.pro
4dff362d4c85748b48a215f43a143bf4
9beb18ab2c8f7acec61c06a381fe06152a0a97ab
75589 F20101113_AAAXMR dempsey_c_Page_18.jpg
56ce539ed49d7d97477add96c1b6f68c
a20f37449b887369e0c6c3df655f4c191a9f133d
F20101113_AAAXHU dempsey_c_Page_06.tif
032d2918dd666b8c2bcf12dc6698b3f7
160c21c1bb8d59f35dbc7b9466eefb323aa83b21
66265 F20101113_AAAXRO dempsey_c_Page_52.pro
d56af480dc1d52015af95410efa477f6
9c16c26804287597552764173dec09e50e847a8a
78419 F20101113_AAAXMS dempsey_c_Page_19.jpg
de460d07d7ab1ec5842a665023d2c6f1
352b44f72da780681a000b5620588ecf8f490f9a
39652 F20101113_AAAXHV dempsey_c_Page_23.pro
5542b07dd98af128d477a8cf9482e57e
ac7f3c8217ff3a5da9580335c9bce4309f4ec27a
10275 F20101113_AAAXRP dempsey_c_Page_54.pro
d0fb5a5e7e13b708422f66dd6c922f18
b3691b683601584944f6be8ac8983e598335df21
72959 F20101113_AAAXMT dempsey_c_Page_20.jpg
2452a746802fd284f5c6147d4a2e4f16
c514f6ebbfb22483ecc217451006a7fd060ab3a9
110 F20101113_AAAXRQ dempsey_c_Page_03.txt
85173d9b3e8711ccd00673175674f6f8
9a2d9c1e9b87c97e4c8725fe75e2af6eac704270
76285 F20101113_AAAXMU dempsey_c_Page_21.jpg
1fdee4042fc426b270fb9e6f044b91c2
fda992296c63e2ed2c0808b8680b121a5d8a303c
11507 F20101113_AAAXHW dempsey_c_Page_41.pro
1d770fbc10b840acdc1f50ed43102c67
b732ca10f08255e6b31e5e17700d20f209f2f417
819 F20101113_AAAXRR dempsey_c_Page_04.txt
37ab10296109cfcbd74535f845fd5f0a
4b1f3826703f51782c53a09a536f08014b7f9e67
70236 F20101113_AAAXMV dempsey_c_Page_22.jpg
9ac6544876aaa08bda2ef79ced965c88
cee4445aba9392d043cc509309eacc08ce2f3942
6739 F20101113_AAAXHX dempsey_c_Page_15thm.jpg
8ccabbb9f6fce726193bb58d1c8193e3
8ce1b0c3b127dbcc8b97c0f1607e5a45b0a021c9
2857 F20101113_AAAXRS dempsey_c_Page_05.txt
d6f4cbbcb89b6ca57c2373a95ad906a3
0a1761f31332659fd63fcf8dc55119c88eda5725
73620 F20101113_AAAXMW dempsey_c_Page_25.jpg
6f2d0925cddef587fe888c8a02833d98
083da9e92ac00e7ff4044c0c74acc666cdeecb80
F20101113_AAAXHY dempsey_c_Page_55.tif
d3c5d197141d4b75c38d5c291469023d
c3a0b81e732108c952ddbdabe3ab3238b905cb30
455 F20101113_AAAXRT dempsey_c_Page_06.txt
9b9a8d9ee3594f1f395828b6fc125ff9
1b3b221a9424f1859e59f67bf94877e860c4bd21
82575 F20101113_AAAXMX dempsey_c_Page_26.jpg
4f8c7c2e4dd714811a7f48aeea5116d7
c0802b6b97d151fa3d77af00bbc99c8655c308f6
F20101113_AAAXHZ dempsey_c_Page_11.txt
726302bdae8ea97ee4bc7883f7301e81
ea9abb8a70332e480a9dc9fdcafe8563f8254067
F20101113_AAAXRU dempsey_c_Page_09.txt
d49adc3d62e43de3b7c76b4d1495b5d5
ea0f9e71dbe3770fea211230e4d9d980adee622f
76700 F20101113_AAAXMY dempsey_c_Page_27.jpg
4186d103daa30856545f330994e76378
871c51fb764e1fc2ae134432597183fb17ae3d7f
203 F20101113_AAAXRV dempsey_c_Page_10.txt
fab1c3eb2665d172fa808dd1072de245
09f87c1028c84c980d0e8ece06484a353649395a
37489 F20101113_AAAXMZ dempsey_c_Page_28.jpg
e0db9434d9ed9436b79fd3703b7fa8b5
a01abb7cd8ccdca15062ed5aa03f3f90e20c2b5c
2075 F20101113_AAAXRW dempsey_c_Page_15.txt
7dc8d831b8d879c631b352233dc74066
85bda9b1de30af3cf491833c93f5c86fa0ef478b
661 F20101113_AAAXKA dempsey_c_Page_40.txt
31c7b9e90e88199b680326e593515d18
e8692c24bd359f18bb267dabacbaeafa81f0ff66
2142 F20101113_AAAXRX dempsey_c_Page_19.txt
82a2da0b6a975475815e93195c752ef9
6d71c77f4153eebefaa86aa88f79164779056eac
3975 F20101113_AAAXKB dempsey_c_Page_07.QC.jpg
c80b31b4f0233c4a2bcd12088217c977
0f27fb54e32d46d1c29151237e248af7719f88a9
2085 F20101113_AAAXRY dempsey_c_Page_21.txt
c3ff0ec93774df7daf9ae394afbd0340
48cb7508a00584c665cb59e4f752f27e379cd3e1
64236 F20101113_AAAXKC dempsey_c_Page_53.pro
2599f284016286a8d95639149879fc16
f38fa7c7f6708a17029155a7dbc5b456369017f8
F20101113_AAAXPA dempsey_c_Page_03.tif
21f6f3c72f948118e03b94787c5afb93
4f23d0f5c66bf5b62fd137b60c80130cd601dd18
1952 F20101113_AAAXRZ dempsey_c_Page_22.txt
5aaf3623b374167c91c1e75d45eb840e
19b10d5cae103dcac49f5f2422c3182709ab45ca
F20101113_AAAXKD dempsey_c_Page_44.txt
de1b9db42144b1b5bf3f9ea2901a6027
eecf38f96eadf8806cfcdd563022c8473f6f96ea
F20101113_AAAXPB dempsey_c_Page_04.tif
0d6abb35319a441f9455514bf3854555
e8e34624214f14369d3d77628097d0995ff14dbd
52276 F20101113_AAAXKE dempsey_c_Page_37.pro
71275cdff8cd29c4994587e27ac3fea5
ce802064dd67813a04ea74765ee1383369e1e0bc
6762 F20101113_AAAXUA dempsey_c_Page_48thm.jpg
2a79ce02e4e112bf90b70afeba38a69c
79549fd45226fa0151ef06bbae0205630701e7ac
F20101113_AAAXPC dempsey_c_Page_07.tif
0c8084b716d33e04b6647037e5062df5
2acac7a0571fdcb1220afdc537ac73d8f4bf9061
88409 F20101113_AAAXKF dempsey_c_Page_53.jpg
cb0fcd4febf23e2d78f9adaeada321d1
fe6b748022010be578dd2258f43ed69cd120ec1d
6640 F20101113_AAAXUB dempsey_c_Page_32thm.jpg
d215d1f5fd0f002c1869f07ea3103e57
ed2109399a8dd3c9bc15d5143b6ee2c76ee9949a
F20101113_AAAXPD dempsey_c_Page_08.tif
eb03d88bc699da9893f26a9e26a30f02
8ab857a701a33fffb52581e0509798ce29436074
891 F20101113_AAAXKG dempsey_c_Page_14.txt
52375d21c392df4db81ac3a0c85ea5fc
ef070ff35bf79616c1bd35c8c52d1093df846489
12263 F20101113_AAAXUC dempsey_c_Page_41.QC.jpg
b5688f671b6d926626acae7391a8d0f0
e0510dfa14e7fbf7d0c67ffad2fe15e691a6a44d
F20101113_AAAXPE dempsey_c_Page_09.tif
02672348eb7e34d02ef0b847fcc6f61b
f904290f10f244024c1587046871c995c4ee9ae1
F20101113_AAAXKH dempsey_c_Page_22.tif
90aa374c938cca95ab3cfe6d94709c6a
0eea6bd9b53b9d84e8b017d884ced1ecd792df20
F20101113_AAAXPF dempsey_c_Page_10.tif
a611f04620b0295e2de973b2f7532340
35d58dde312cd2a77eb5a80e18612dbbf85d4fc6
1989 F20101113_AAAXKI dempsey_c_Page_20.txt
d5b37b913a927c66e37706fb31fb6a5e
debc1f6a8815163b76e5ec4d582028973d2cc97f
6575 F20101113_AAAXUD dempsey_c_Page_44thm.jpg
a097800d4937c6b147ef5f69b96f2a4c
9bb59fce115d3e13f9ea39111392f3d1d987a873
F20101113_AAAXPG dempsey_c_Page_11.tif
d82c53b85cd04eb02b09c8474ebf6eb7
27f804b9df05ad5198e71b6fa795697aa0389dbf
5792 F20101113_AAAXKJ dempsey_c_Page_05thm.jpg
e11ac81120bf4f28d899c3b50cd55b8a
4b42e3dc1d7febb25f0beed659c6b86765b2774a